Boundary Delimitation —
English
 

Boundary Delimitation

Boundary delimitation usually refers to the process of drawing electoral district boundaries. However, it can also be used to denote the process of drawing voting areas (also called polling areas, districts or election precincts) for the purposes of assigning voters to polling places. The periodic delimitation of electoral boundaries is necessary in any representative system where single-member districts or uniformly small multi-member districts are used. If electoral boundaries are not periodically adjusted, population inequities develop across districts. 

Countries have adopted various methods for delimiting districts. Countries that delimit districts must establish a formal structure and a set of rules for carrying out the delimitation process. In some, the choice of methods is simply a matter of historical tradition, but the delimitation can also be based on a variety of factors that may include the geographic size of the country, its physical features, or its financial resources. Different sets of districts can produce different election outcomes, and even if the underlying vote patterns remain constant, the choice of delimitation practices is important. 

The Boundary Delimitation Encyclopedia topic discusses the administrative and cost implications of delimiting electoral districts for representation purposes and considers the social and political context in which delimitation takes place. It also covers the boundary setting process associated with voting areas. 

 

Encyclopaedia Content

Overview

The term "boundary delimitation" usually refers to the process of drawing electoral district boundaries. However, it can also be used to denote the process of drawing voting areas (also called polling areas, districts or election precincts) for the purposes of assigning voters to polling places. The term has been employed to describe the process of demarcating administrative boundaries such as state/province, county or municipality lines as well.

This section focuses only on the delimitation of electoral districts and voting areas. Furthermore, because the delimitation of electoral districts is far more complicated and much more controversial than the delimitation of voting areas, the vast majority of this topic area will be devoted to the delimitation of electoral district boundaries.

Delimiting Electoral Districts

The periodic delimitation of electoral boundaries, or redistricting or redistribution, is necessary in any representative system where single-member districts or uniformly small multimember districts are used. If electoral boundaries are not periodically adjusted, population inequities develop across districts.

Adjusting district boundaries can have major consequences not only for the legislators who represent the districts, but also for the individual and community constituents of the districts. Ultimately, election results and the composition of the legislature are affected by the selection of district boundaries. But the importance of the redistricting process is seldom recognised outside of political circles.

Countries have adopted various methods for delimiting districts. In some, the choice of methods is simply a matter of historical tradition. In others, methods for delimiting districts have been borrowed from a colonial power or an influential neighbouring country. In still other countries, delimitation is based on a variety of factors that may include geographic size of the country, its physical features, or its financial resources. Recently, countries have taken their political and social context into account when making decisions on which redistricting practices to adopt. Clearly, there is a broad range of possibilities. Informed decision-making is the best approach to selecting or reforming a redistricting process.

Electoral Systems that Delimit Electoral Districts

The delimitation of electoral districts is most commonly associated with plurality or majority electoral systems. Both systems rely heavily, if not exclusively, on single-member districts. These districts must be redrawn periodically to reflect changes in the population.

Plurality and majority systems, however, are not the only types of electoral systems that require the periodic delimitation of electoral districts. Some mixed and proportional representation systems, including the single transferable vote system, also must occasionally delimit electoral districts.

The importance of the delimitation process varies, depending on the type of electoral system. Because plurality and majority systems can, and do, produce election outcomes that are disproportional with regard to the ratio of legislative seats to partisan votes, the delimitation process is very important. It is less important in mixed systems or proportional representation systems.

Structure and Rules for Delimiting Electoral Districts

Countries that delimit districts must establish a formal structure and a set of rules for carrying out the delimitation process. Because different sets of districts can produce different election outcomes, even if the underlying vote patterns remain constant, the choice of delimitation practices is important. Electoral legislation outlining the formal structure and rules for delimitation should address the following issues:

  • Who will draw the district lines or boundaries? And who will have the ultimate responsibility for selecting the final districting plan?
  • Should the persons who draw the districts be independent from the legislature?
  • Should the boundary authority be politically neutral?
  • Should the legislature have any formal role at all in the process?
  • Should some mechanism exist for public input into to the process?
  • Should criteria be adopted for the line drawers to follow? If so, what should these delimitation criteria be?
  • How often should districts be redrawn and how long should the redistricting process take?
  • Under which criteria should districts be redrawn? 
  •  

Delimitation practices vary markedly across countries. In the United States, for example, legislators are usually responsible for drawing electoral district lines. Partisan politics and the protection of incumbent legislators play a large role in the redistricting process. By contrast, politicians in many Commonwealth countries have opted out of the delimitation process. Districting is left to independent commissions with neutral delimitation criteria for guidance. The reasons for these differences are best explained by the social, political and cultural norms.

Tasks Involved in Drawing Electoral District Boundaries

Although the rules for delimitation vary markedly across countries, the tasks involved in drawing districts are generally very similar. Drawing district boundaries entails:

 

  • allocating seats to sub-regions of the country, such as states or provinces;
  • creating a database minimally composed of maps and population data;
  • assigning geographic units to districts until all geographic units within the territory have been assigned;
  • summarising and evaluating the districting plan.
  •  

This can be a complex, time-consuming and expensive process.

Delimiting Voting Areas

Most countries, regardless of the type of election system employed, delimit voting areas for the administration of elections. Voting areas are contiguous geographic areas in which all the voters are assigned to the same voting (or polling) station. Since voting areas are used for election administration purposes only, the boundaries of these areas tend to be less controversial, and the delimitation of these areas is usually left to the discretion of election administrators. However, because the data required and the tasks involved in the delimitation of voting areas are similar to those involved in drawing electoral district lines, the delimitation of voting areas is discussed under the Boundary Delimitation section of ACE.

Conclusion

The Boundary Delimitation topic area discusses the types of electoral systems that require periodic electoral district delimitation and considers the advantages and disadvantages of various districting alternatives (see Delimiting Electoral Districts). It considers the formal structure and rules that countries use to conduct electoral district delimitation, or redistricting (see Structure and Rules for Delimiting Electoral Districts). It outlines each step of the district drawing process, from the creation of a redistricting database to describing and evaluating districting plans (see Tasks Involved in Drawing Electoral District Boundaries). In addition, this topic area covers delimitation of voting areas for election administration purposes (see Special Considerations: Delimiting Voting Areas).

We hope this discussion will better prepare countries to make informed decisions on whether to delimit electoral districts and, if so, which boundary delimitation practices to adopt.

 

Context

The decision as to whether to delimit electoral districts, and by what means, should depend on country-specific conditions that take into consideration the political and social context and available administrative and financial resources. But this is often not the case – in many countries delimitation practices have simply been a matter of historical tradition. In other countries, district delimitation methods have been borrowed from a colonial power or an influential neighbouring country. Delimitation practices that work well in some countries will not work well in others. Electoral districts were developed as a means of providing fairer representation. Under certain circumstances, they may still provide a good option for representation, but this is very much dependent on the country context.

Historical Background [1]

At the beginning of the nineteenth century, representation was based on communities, not on the number of individuals or voters. Constituencies varied greatly in population. In England, for example, each county, borough, and university, regardless of size, had two representatives in the House of Commons. The composition of the House of Commons, and legislatures throughout Europe, reflected the view that communities or distinct categories of society – for example, the clergy and the nobility – should be represented, not individuals or voters.

Beginning in the latter part of the eighteenth century, citizens began to demand a broadening of their franchise and fairer representation in the legislature. Single-member districts for the election of legislators were embraced by this movement as a means towards greater democracy. Not only would single-member districts ensure more equal representation for citizens, it was believed that they would produce a more representative legislature – increasing representation for the urban, working class and other traditionally under-represented groups.

In the eighteenth century, single-member districts were first adopted in the British colonies that later became the United States. During the nineteenth century, many European countries gradually followed suit. Denmark adopted single-member districts for elections to its lower chamber in 1849. A newly unified Italy chose single-member districts to elect representatives to the national legislature in 1861. The North German Confederation adopted single-member districts in 1867; imperial Germany, in 1871. France has used single-member districts intermittently since 1875. Britain adopted single-member districts in 1885, and the Netherlands followed suit in 1887. Norway, one of the last European countries to adopt single-member districts, did so in 1905 when the country obtained full independence.

Most of the remaining European countries that continued to use communities, rather than specially delimited electoral districts, for the election of representatives were the less progressive countries, such as Portugal or the Balkan states. Of the more modern European countries, only Switzerland and Belgium used multimember districts throughout the nineteenth century. Belgium used its nine historic provinces as boundaries for its multimember districts, but it recognised the principle of equal representation by varying the number of representatives assigned to province. When proportional representation was adopted in Belgium in 1899, multimember districts based on provincial boundaries were retained.

In Switzerland, prior to 1848, each canton sent a single representative to the federal assembly. The constitution of 1848 provided for representation based on population; so in 1850, 49 electoral districts, or constituencies, were created. The boundaries of these constituencies went unchanged for 70 years, however. In 1919, Switzerland adopted proportional representation. Since 1919, representatives have been elected from multimember districts that correspond to the canton boundaries.

The trend towards single-member districts in Europe ebbed with the advent of proportional representation in the late nineteenth century. Between 1899, when Belgium adopted proportional representation, and 1921, when Norway moved to proportional representation, most of the continental European countries adopted one form of proportional representation or another.

Because multimember districts are used with systems of proportional representation, countries that have adopted such systems are no longer required to periodically redraw district boundaries. Instead, administrative divisions such as states, provinces, or counties can be used to elect representatives. Equality of population is achieved by varying the numbers of legislators elected from a district, rather than redrawing district lines. Today, the norm in Europe is multimember districts that correspond to administrative divisions and do not require periodic delimitation.

Social and Political Context

The delimitation of single-member districts has several advantages. The three most often cited are simplicity, stability and strong links between elected representatives and their constituents. Each of these advantages may be important ones, depending on the social and political context in which the districts are adopted.

Elections held in single-member districts tend to be quite easy for voters to understand, especially in conjunction with plurality or majority voting rules. Simplicity may be a significant advantage for countries with high illiteracy rates.

Single-member districts promote stability by facilitating strong, single party government. This is because single-member districts tend to produce election outcomes in which the majority party is over-represented. This may be an important advantage in countries that have reason to fear or have actually experienced a proliferation of small extremist parties or coalition governments that have frequently fallen.

Single-member districts provide voters with strong constituency representation. Voters have a single, easily identifiable district representative to whom they can appeal for constituency service. Voters also have a single district representative whom they can hold accountable for protecting constituency interests. This may have a positive affect on voters’ feelings of political efficacy, which may, in turn, increase voter turnout. Political efficacy and turnout are both important ingredients for system legitimacy, which may be important to newly emerging democracies.

Single-member districts have one very serious drawback: they tend to over-represent the majority political party at the expense of the other political parties. Countries that delimit single-member districts must be willing and able to accept disproportional election outcomes. Although it is possible to devise a fair and non-partisan redistricting process, it is not possible to guarantee an unbiased election outcome with single-member districts unless there are provisions for a second, party vote (as is the case with a mixed electoral system).

Disproportional election results may be difficult to accept in a country with many political parties representing widely disparate interests. The results will be virtually impossible to accept if deep cleavages exist in the society. For example, if there is a relatively large, politically cohesive ethnic, racial, or religious minority group that has consistently been denied what it perceives as fair representation, elections could lead to conflict, possibly even violence and instability.

Taking the Politics Out of Redistricting Traditionally, legislatures have been responsible for drawing their own districts. Electoral abuses such as malapportioned districts (districts that vary substantially in population) or “gerrymandered” districts (districts intentionally drawn to advantage one political party or group at the expense of the others) were not uncommon. These abuses led a number of countries to adopt reforms designed to remove “politics” from the redistricting process. In these countries, non-partisan commissions draw district boundaries following a set of neutral redistricting criteria. The public is encouraged to participate through a public inquiry process. And the legislature is permitted only a limited role, if any role at all, in the redistricting process.

These reforms have been adopted by many Commonwealth countries, where the reforms appear to have been quite successful. Redistricting is rarely viewed as “partisan,” even when the outcome of an election clearly favours one party at the expense of the other parties. Despite their success in the countries that have adopted them, reforms of the redistricting process have not been embraced everywhere. For example, legislatures still draw congressional districts in most states in the United States.

In the United States, the political system and political institutions were designed on the premise of competing factions, with special interests and parochial concerns vying for limited resources within the legislature. As a result, Americans tend to assume that “politics” and the pursuit of political self-interest are inevitable. Reforms of the redistricting process are unlikely to be adopted in the near future because many Americans believe it is as impossible to divorce politics from the redistricting process as it is to divorce politics from the legislative process in general.

Only countries that are less pluralistic, more politically ideological, and more public-regarding accept the proposition that politics can be removed from the redistricting process. Furthermore, strong party organisations and centralised legislative authority are needed to enforce sanctions against legislators who attempt to influence the redistricting process. While these conditions appear to be met in most Commonwealth countries, they are not present everywhere.

But countries that adopt independent boundary commissions and neutral redistricting criteria can only prevent partisan interests from controlling the process; these countries cannot alleviate disproportional election outcomes. If political fairness is defined by outcome, rather than by process – and specifically by a proportional outcome for political parties and/or minority groups – then single-member systems will fail the fairness test more often than not, no matter who draws the districts. Countries that value proportionality over all else – perhaps because of the need to ensure equitable representation to deeply divided groups within the society – are wise to choose some form of proportional representation, which may or may not include provisions for single-member districts. If stability in the form of strong, single party government is more important, however, delimiting single-member districts is a good choice.

Administrative and Cost Considerations

Delimiting electoral districts can be a complex, expensive, and time-consuming process. Countries that choose to delimit electoral districts must be prepared to administer what may be a very complicated procedure and must be willing to spend the necessary funds to complete the process.

Administering the process will be complicated. Information from a wide variety of sources must be collected, verified, and synthesised. Districts must be drawn, and then must be evaluated – procedures that could involve political party and other stakeholders input. And after a final set of districts is adopted, they must be implemented. This usually requires changes to voter registration lists, and voter notification of new district assignments. In addition, election officials may have to redraw voting areas and relocate polling places. Ultimately, implementation may require a great deal of co-ordination with local and regional governments as well as election officials at the local, regional, and federal levels.

While computers can make the delimitation process more efficient, computers can also complicate it. Computer hardware and software options must be examined carefully. A decision to acquire a computer system must take into account the amount of money to be invested in hardware and software, the time and expense of acquiring and training staff to operate the system, and the time and expense of converting maps and population data to a computer readable format.

The administration of the delimitation process may be challenging. It may be difficult to find and/or train qualified staff, particularly if the process is to be computerised. There may be obstacles in obtaining accurate and up-to-date information, especially maps. And there may be problems coordinating the collection and implementation processes with election officials and local government officials. All of these challenges must be met, however, if the delimitation process is to be accurate and timely.

The price tag for redistricting varies, some countries spend relatively little to redistrict, while others face increasing costs. The redistribution in Great Britain and Northern Ireland completed in 2008 cost £UK 13.6 million. The previous redistribution cost £UK 5 million.  Similarly in Canada, the redistribution concluding in 2003 cost approximately $Can 10.1 million while the previous one in1997 cost approximately $Can 6.5 million. In the United States, enormous amounts of money are spent to redraw congressional districts every ten years. Of course, the redistricting process in the United States is quite decentralised, and the amount of money each state spends varies considerably: some states spend well over US$1 million to draw congressional districts, and millions more to defend these districts if challenged in court; other states, however, spend far less. But regardless of the actual amount spent, countries must recognize that delimitation is rarely a one-time event and must be prepared to devote the time and resources necessary to redraw electoral districts periodically.

[1] This discussion draws heavily from an article written by Michael Steed entitled “The Constituency” which appeared in Representatives of the People? Parliamentarians and Constituencies in Western Democracies, edited by Vernon Bogdanor and published by Gower Publishing in 1985.

 

Guiding Principles

Because delimitation, or redistricting, practices vary greatly around the world, there are few universal principles to guide the delimitation process. Countries disagree on fundamental issues, such as how impartial and independent the process can and should be from the legislative and political concerns. But there are several generally accepted principles: 

  • representativeness;
  • equality of voting strength;
  • independent, impartial boundary authority (or, at a minimum, reciprocity);
  • transparency;
  • non-discrimination.
  •  

Representativeness

Electoral district boundaries should be drawn such that constituents have an opportunity to elect candidates they feel truly represent them. This usually means that district boundaries should coincide with communities of interest as much as possible. Communities of interest can be defined in a variety of ways. For example, they can be geographically defined communities delineated by administrative boundaries or physical features such as mountains or islands, or they can be “communities” that share a common race, ethnic or tribal background, or the same religion or language. If districts are not composed of communities of interest, however defined, it may be difficult for representatives to serve the constituency well.

Equality of Voting Strength

Electoral district boundaries should be drawn so that districts are relatively equal in population. Equally populous districts allow voters to have an equally weighted vote in the election of representatives. If, for example, a representative is elected from a district that has twice as many voters as another district, voters in the larger district will have half the influence of voters in the smaller district. Electoral districts that vary greatly in population--a condition referred to as "malapportionment"--violate a central tenet of democracy, namely, that all voters should be able to cast a vote of equal weight.

The following are two standards developed to reflect this principle, one offered by the Organization for Security and Cooperation in Europe (OSCE) and one by the UN Committee on Human Rights (UNCHR):

  • The delineation of constituencies in which elections are conducted must preserve the equality of voting rights by providing approximately the same ratio of voters to elected representatives for each district. Existing administrative divisions or other relevant factors (including of a historical, demographic, or geographical nature) may be reflected in election districts, provided the design of the districts is consistent with the equality of voting and fair representation for different groups in society. (OSCE, “Inventory of OSCE Commitments and Other Principles for Democratic Elections”)
  • The principle of one person, one vote must apply, and within the framework of each State’s electoral system, the vote of one elector should be equal to the vote of another. The drawing of electoral boundaries and the method of allocating votes should not distort the distribution of voters or discriminate against any group and should not exclude or restrict unreasonably the right of citizens to choose their representatives freely. (UN Committee on Human Rights, General Comment 25, “The Right to Participate in Public Affairs, Voting Rights and the Right to Equal Access to Public Service”)
  •  

Independent, Impartial Boundary Authority

Ideally, the legal framework for boundary delimitation should provide that the persons or institution responsible for drawing electoral boundaries be independent and impartial. In addition, the recommendations of the boundary authority should not be subject to modification or veto by the government or by the legislature.

Failing this, the procedure for delimiting electoral districts should be clearly spelled out in legislation so that the rules regulating the process are the same, regardless of who is drawing the district boundaries – that is, the process should offer reciprocity. If political concerns are permitted to play a role in the process, then all political parties must be given access to the process. These rules must be clearly understood and must be acceptable to all major political parties and participants in the districting process.

Transparency

Because electoral systems that feature districts often produce disproportional election outcomes, it is essential that the delimitation process be considered fair if the result is to be deemed legitimate by stakeholders and voters. This means that the delimitation process should be as transparent as possible, with the methodology and guidelines clearly established and publicised in advance. Incorporating public hearings into the process to allow stakeholders to offer comments for the boundary authority to consider is also important.

Non-Discrimination

Electoral boundaries should not be drawn in a manner that discriminates against any particular minority group. For example, dividing a geographically-concentrated minority group among several electoral districts so that the group constitutes a minority of the voters in every single electoral district should be prohibited. Of course, electoral systems that rely exclusively on single-member districts cannot guarantee even some minimal percentage of seats for minority groups or minority political parties in the population. However, this fact should not open the door to active discrimination against a given minority group.

Delimiting Electoral Districts

The delimitation of electoral districts is a fairly recent phenomenon. Prior to the nineteenth century, the composition of legislatures reflected the view that distinct categories of society (i.e., towns, the clergy, and the nobility) should be represented and not individual citizens. This view of representation led to legislatures based on subdivisions that varied greatly with regard to the size of population being represented.

In the latter part of the eighteenth century, and throughout the nineteenth century, citizens began to demand a broadening of their franchise. These demands were accompanied by the belief that "fair" representation entailed an equalisation of population per representative. The response to this quest for the equalisation of representation was either to vary the number of legislators elected from a territory or to subdivide territories into electoral districts of more equal population. Single-member districts for the election of legislators were first adopted in the British colonies that later became the United States. Many European countries gradually followed suit and adopted single-member electoral districts during the nineteenth century.

The trend towards single-member districts in Europe ebbed with the advent of proportional representation in the late nineteenth century. Because multimember districts are generally used with systems of proportional representation, countries that adopted such systems no longer needed to delimit electoral districts. Instead, equality of population could be achieved by varying the numbers of legislators from a territory. The use of administrative divisions such as states, provinces, or counties as multimember districts also provided the additional benefit of community representation.

The tradition of single-member electoral districts was retained in other countries, however, in the belief that distinct benefits are derived from their use. These benefits include:

  • a strong geographic link between constituents and their representatives;
  • the ability of constituents to hold individual representatives accountable;
  • the tendency of an electoral system using single-member districts to produce strong single-party governments, and, by extension, a stable political system.

A major drawback to the use of single-member districts is that electoral district boundaries must be adjusted periodically to maintain populations of relatively equal size. Of course, in some countries, multimember districts must also be redrawn. Not only is the district delimitation process expensive and time consuming, but there is also the possibility that drawing new district boundaries could favour, intentionally or unintentionally, one particular interest or political party over others.

In some countries, multimember districts are also redrawn. How electoral districts are drawn, whether the districts are single-member or multimember, and whether they follow existing administrative boundaries or not can influence the nature of representation by elected legislators. This section considers the various types of electoral systems that delimit electoral districts (see Electoral Systems that Delimit Electoral Districts). It also considers electoral districting alternatives, such as the magnitude of the districts and whether or not the district boundaries should be aligned with pre-existing administrative boundaries (see Electoral District Alternatives).

Electoral Systems that Delimit Electoral Districts

Traditionally, three broad categories of electoral systems have been described: plurality systems, majority systems, and proportional representation systems. The most important element that differentiates these electoral systems from one another is the means by which seats in the legislature are allocated:  

  • to candidates receiving a plurality of the vote
  • to candidates obtaining a majority of the vote
  • proportionally on the basis of votes cast for political parties or candidates

A recent addition to these three broad categories of electoral systems is the mixed electoral system, which combines elements of both proportional representation and plurality or majority voting systems.

Delimiting Districts: Plurality or Majority Systems

The delimitation of electoral districts is most commonly associated with plurality or majority electoral systems. Both systems tend to rely heavily, if not exclusively, on single-member electoral districts. These districts must be redrawn periodically to reflect shifts in the population. Both systems also share one fundamental element because of their reliance on single-member districts--the number of seats that a political party receives depends not only on the proportion of the votes it received, but also on where those votes were cast. Under plurality and majority systems, minority political parties whose supporters are not geographically concentrated usually obtain fewer seats than their proportion of the vote would suggest they are entitled. The multimember districts of proportional systems can rectify this distortion in the transfer of votes to seats because the larger the magnitude of the electoral districts, the more proportional the results.

The process of electoral district delimitation in a plurality system is important because the configuration of districts can affect the partisan, and possibly even the racial, ethnic, religious or linguistic composition of the legislature. The effect may be intentional, as in a partisan "gerrymander," or it may be unintentional. Even district boundaries drawn by a neutral boundary commission may unintentionally favour one party over others.

Majority electoral systems can, and do, produce election outcomes that are just as disproportional as plurality electoral systems with regard to the ratio of seats to votes. This is not surprising, given that both systems rely almost exclusively on single-member districts. The district delimitation process in a majority system is, therefore, just as important as it is in a plurality electoral system.

For more information on plurality or majority electoral systems see the Electoral Systems: Plurality/Majority Systems.

Delimiting Districts: Proportional Representation

There are two major types of proportional representation systems: the party list system and the single transferable vote. (The mixed member proportional system also produces proportional results, but this system will be discussed under the "mixed system" category.) The party list system is the far more common of the two. Under the party list system, electoral districts rarely, if ever, require delimitation. If electoral districts are employed, they are relatively large multimember districts whose boundaries generally correspond to administrative divisions. To accommodate shifts in population, the number of seats allocated to individual multimember districts is adjusted, rather than redrawing the boundaries of the districts. The single transferable vote, used in Ireland and Malta, is the other type of proportional representation. Because voting is on the basis of candidates, not parties, these countries employ small multimember districts with only three to five members elected per district. Electoral district boundaries must therefore be redrawn periodically in these two countries.

Delimiting Districts: Mixed Electoral Systems

Mixed electoral systems are becoming increasingly popular. They employ both party list proportional representation and single-member electoral districts with plurality or majority vote requirements. The German electoral system is the prototypical mixed electoral system.

Because mixed systems incorporate single-member districts, the delimitation of electoral districts must occur periodically to adjust for shifts in the population. The importance of the delimitation process and the influence that district configurations have on the outcome of elections is dependent on whether party list seats are used to correct any distortions in the relationship between seats to votes produced by the single-member districts. In countries such as Germany, seats allocated under the party list system are used to compensate for any distortions in the seats-to-votes ratio produced at the electoral district level. Mixed systems that use party list seats in a compensatory manner are sometimes called "mixed member proportional" systems because the election results are proportional.

In countries such as Russia, party list seats are not used to compensate for any disproportionality arising from elections in single-member districts. Rather, seats allocated to the parties under the party list component of the election are simply added to the seats won at the electoral district level. The partisan seats-to-votes ratio may therefore be distorted. In this type of mixed system, sometimes called a "parallel" system, the district delimitation process is more important because it can have a more pronounced effect on the partisan composition of the legislature.

Electoral District Alternatives

 

Two important factors to be considered when contemplating electoral districting alternatives are: (1) district magnitude and (2) the alignment of electoral district boundaries with existing administrative and/or political boundaries. District magnitude refers to the number of legislative seats assigned to a district[1]. A district can be either a single-member district or a multimember district, where the number of seats may range from two to one hundred or more. Electoral districts can be aligned with administrative divisions—that is, administrative divisions can be used as electoral districts—or electoral districts can be specially drawn with little regard for administrative divisions, usually to meet equal population criteria.

These two factors form a matrix [2]. The first dimension, district magnitude, relates to the issue of single-member versus multimember districts. The second dimension refers to the alignment or nonalignment of electoral districts with administrative or political boundaries.

Most single-member districts fall into the nonalignment category. The districts tend to be artificial pieces of geography that have no meaning outside the electoral context. Some single-member districts, however, particularly those in proportional representation countries, are small, highly distinctive communities. For example, a few small cantons in Switzerland form single-member districts.

Countries with multimember districts often use existing administrative divisions as electoral districts. Each district is assigned the appropriate number of legislative seats for its population, with individual districts having as few as two representatives and most districts having far more than two representatives. These countries usually employ some form of proportional representation. The more artificially constructed multimember districts are found in countries such as Ireland and Malta, which use districts that are uniformly small in magnitude because elections are conducted using the single transferable vote.

More information on the Alignment of Districts with Administrative Boundaries can be found here.

Notes:

[1] According to Taagepera and Shugart, the term "magnitude" is preferable to "size," because size usually refers to the number of voters in an electoral district or the geographic extent of a district. See Rein Taagerpera and Matthew Soberg Shugart, Seats and Votes: The Effects and Determinants of Electoral Systems (New Haven: Yale University Press, 1989)

[2] Michael Steed, "The Constituency," in Representative of the People? Parliamentarians and Constituencies in Western Democracies, ed. Vernon Bogdanor (Grower Publishing, 1985)

 

District Magnitude

District magnitude refers to the number of legislative seats assigned to a district. Countries have adopted electoral rules that range anywhere from the exclusive use of single-member districts to a system where the entire country, in effect, functions as a single district. The United States and the United Kingdom are at one end of the spectrum, in which each and every legislator represents a single district. At the other end of the spectrum are countries such as Israel and the Netherlands, in which the district magnitude is equal to the total number of members of the legislature. Most countries are somewhere in the middle of this range; and within a country there is often a wide variation in the magnitude of districts.

Some countries set all their electoral districts at the same magnitude or within some narrow range of magnitudes. District boundaries are then usually drawn according to some voters-per-representative formula. This approach has been adopted in the United States and most other countries with plurality or majority electoral systems, where the district magnitude is set at one. This procedure is also used in Ireland and Malta, both of which employ small multimember districts and the single transferable vote. In Malta, all districts have a magnitude of five. In Ireland, the range in magnitude is from three to five.

Alternatively, some countries use existing regional, administrative or political divisions as electoral districts. Each electoral district is then assigned a specific number of seats according to its population. Most countries with electoral systems based on proportional representation use this procedure. The larger the district magnitude, the more proportional the outcome of the election--that is, the more seats per district, the closer the approximation between a political party's percentage of the vote and the number of seats that party receives in the legislature.

Single-member electoral districts must be redrawn periodically to ensure relatively equal populations. Some countries with uniformly small multimember districts must also redistrict periodically in order to comply with equal population standards. Electoral districts with large magnitudes, however, do not need to be redrawn; seats are simply reassigned from one district to another to meet equal population standards.

Single-Member and Multimember Districts

The debate about the advantages and disadvantages of single-member and multimember districts overlaps, to a large extent, with the debate over plurality or majority systems and proportional representation systems. This is because plurality and majority systems usually employ single-member districts, and proportional representation systems use multimember districts.

For the purpose of boundary delimitation, single-member districts

  • ensure geographic representation

However, single-member districts

  • must be redrawn on a regular basis to maintain populations of relatively equal size;
  • are usually artificial geographic entities whose boundaries do not delineate clearly identifiable communities, and as a consequence, the entities have no particular relevance to citizens;

Multimember districts:

  • can more easily reflect administrative divisions or communities of interest within the country because there is flexibility with regard to the numbers of representatives per district and, therefore, the size and geographic composition of the district; and
  • need not change boundaries, even if the population of a district increases or decreases, because the number of representatives elected from the district can be altered;

Advantages of Single-Member Districts

Supporters cite several advantages, namely that single-member districts:

  • provide voters with strong constituency representation because each voter has a single, easily identifiable, district representative;
  • encourage constituency service by providing voters with an easily identifiable "ombudsman";
  • maximise accountability because a single representative can be held responsible and can be re-elected or defeated in the next election;
  • ensure geographic representation.

Disadvantages of Single-Member Districts

In citing disadvantages, critics point out that single-member districts: 

  • must be redrawn on a regular basis to maintain populations of relatively equal size;
  • are usually artificial geographic entities whose boundaries do not delineate clearly identifiable communities, and as a consequence, the entities have no particular relevance to citizens;
  • because of their tendency to over-represent the majority party and under-represent other parties, cannot produce proportional representation for political parties.

Conclusion

The strengths of single-member districts rest in the close ties between representatives and constituents, the accountability of representatives to the voters, and constituency service. Because single-member districts are used in conjunction with plurality or majority voting rules, they are also said to foster strong and stable government.

Alignment of Districts with Administrative Boundaries

In many countries, especially those that employ multimember districts, the boundaries of electoral districts follow the boundaries of existing administrative divisions--usually states or provinces. Often these administrative divisions have some historical significance. Some administrative divisions, however, may be of recent vintage and of little relevance to citizens.

There are certain advantages to aligning electoral districts with administrative divisions, in that the districts will then

  • correspond to the entities that administer local governmental functions;
  • correspond to social, political, and other non-governmental organisations arranged by administrative areas;
  • be more meaningful to constituents if the districts correspond to well-recognised administrative divisions;
  • reflect geographic communities of interest when the administrative divisions themselves reflect distinct communities of interest.

Correspondence with Governmental Functions

Administrative divisions may be assigned important governmental functions. For example, local government entities may have responsibilities for levying taxes or administering justice, education or public health. Constituents of local government entities can benefit from being able to identify and relate to the elected representatives from districts that correspond to these administrative divisions, especially if constituents need help in dealing with these governmental agencies.

In addition, a country's election machinery may be organised around administrative divisions. If so, it may be easier to conduct elections if electoral district boundaries correspond to administrative boundaries.

Correspondence with Non-Governmental Organisations

Governmental agencies are not the only organisations that operate within specific administrative divisions. Many social, cultural, and political organisations are also arranged according to administrative areas. Such organisations may include political parties, trade unions, professional associations and many other occupational, social, and cultural organisations. Members of these organisations may benefit by being able to identify and relate to elected representatives from the same administrative division. Elected representatives, too, may find it easier to work with and communicate with members of these non-governmental organisations.

Recognition of Electoral Districts

Long-standing state or provincial boundaries engender a corporate identity that voters can relate to more easily than artificially created electoral districts. Voters may be able to distinguish between their district and other districts and identify their elected representative more easily if electoral districts are defined by administrative boundaries.

Reflection of Communities of Interest

Electoral districts that are composed of long-standing state or provincial territories may reflect geographically concentrated communities of interest based on a common heritage or on shared racial, ethnic, religious or language characteristics. Some of the more modern administrative entities that form the basis of electoral districts, however, may have little in the way of common roots. Consequently, these districts do not bring together constituents with common interests.

Drawbacks to Using Administrative Divisions

Administrative divisions do not have the same level of importance throughout the world. Although they reflect important regional differences in some countries, administrative divisions in other countries have been created very recently. In these countries, the boundaries are artificial and of little significance to citizens. In fact, the boundaries may divide natural communities of interest such as long-standing racial, ethnic, religious, or spoken language communities. Using administrative entities to form electoral districts in this instance may actually conflict with the creation of electoral districts that reflect strong communities of interest.

Single-Member Districts and Administrative Divisions

Aligning electoral districts with administrative boundaries is a very common practice in countries with multimember electoral districts. However, correspondence with administrative divisions is more problematic in countries that employ single-member districts exclusively. This is because there is often a conflict between drawing single-member districts that follow existing administrative lines and drawing single-member districts that are relatively equal in population.

This is not to suggest that single-member districts can never correspond to administrative divisions within a country. Many countries that use single-member districts do emphasise the need to respect administrative boundaries. But the weight given to this consideration varies, depending on the importance placed on equality of population and other redistricting criteria that may conflict with respect for administrative divisions. For example, in the United Kingdom, large numerical deviations in population were tolerated in order to accommodate local administrative areas. This changed with the passage of the Parliamentary Voting System and Constituencies Act 2011 which sets a variance of plus or minus five percent of the United Kingdom electoral quota. In the United States, congressional district boundaries never cross state lines; the strict enforcement of equality of numbers, however, is far more important than respect for local administrative boundaries.

Country Survey of Delimitation Practices

Information on the delimitation practices of 87 countries (or territories) was assembled and summarized for the Delimitation Equity Project conducted by IFES from 2004-2005.

Data was collected through a variety of means:

(1) the compilation of constitutional and election law provisions on constituency delimitation;

(2) information gathered during a series of election missions undertaken by the author on behalf of IFES and other organizations; and

(3) three surveys conducted over the last few years: 

  • A delimitation survey sent to election administrations around the world in the fall of 2004.
  • Comparative Data section of the ACE Electoral Knowledge Network, which includes a series of questions on delimitation practices.
  • A comparative redistricting project that included a conference, a survey, and a series of case studies, funded in large part by the National Science Foundation. (The Comparative Redistricting Project, which included a conference entitled "Redistricting from a Comparative Perspective" held December 7-9, 2001, was funded by grants from the National Science Foundation and the Center for the Study of Democracy at the University of California, Irvine.)

Countries that Delimit Electoral Districts

The 87 countries (or territories) for which information was collected represent a broad geographic array: 21 of the countries are located in the Americas, 34 in Europe, 15 in Africa, 2 in the Middle East, 11 in Asia and 4 in Oceania (Australia/South Pacific Islands).

Of these 87 countries, 60 (69 percent of the total number of countries) reported delimiting electoral districts. The breakdown by region of the countries that delimit electoral districts is as follows:

Region

Percent of Countries that Delimit Constituencies

Total Number of Countries

Americas

57%

21

Europe

62%

34

Africa

73%

15

Middle East

100%

2

Asia

91%

11

Oceania

100%

4

TOTAL

69%

87

The majority of countries in every region represented in the survey delimited electoral districts. Countries in the Americas were the least likely to have specifically delimited electoral districts: although every country in North America (Canada, Mexico and the United States) and most countries in the Caribbean delimit constituencies, very few in Central and South America do so. Countries in Oceania – Australia, New Zealand, and most of the South Pacific Island countries – are the most likely to have specifically delimited electoral districts.

Almost without exception, the countries that do not specifically delimit districts are countries that have List Proportional Representation (List PR) electoral systems (Georgia and Panama are the only countries surveyed that do not have a List Proportional Representation electoral system and do not periodically delimit electoral districts). Every other type of electoral system included in this study requires some delimitation of electoral districts: First Past the Post (FPTP) Systems, Two Round Systems (TRS), Alternative Vote (AV) and Block Vote (BV) Systems, and Parallel and Mixed Member Proportional (MMP) Systems.

Structure and Rules for Delimiting Electoral Districts

Countries that delimit electoral districts must establish rules and a formal structure for carrying out the process. Because different sets of districts can produce different election outcomes, even if underlying voting patterns remain the same, the choices involved in the design of the redistricting process are important. Included among those decisions are the following: 

  • Who should draw the district lines?
  • Who should decide on the final districting plan?
  • Under what grounds should the district lines be determined?
  • Should the legislature have a role in drawing the districts or deciding on the final plan?
  • Should the general public have input into the redistricting process?
  • How often do districts need to be redrawn?
  • How long will the redistricting process take to complete?
  • Should formal criteria be established for line drawers to consider? If so, what criteria should be established?
  • Should the plan be subject to challenge in the courts?
  •  

The task of drawing districts must be assigned to some boundary authority. The composition of the boundary authority and the degree of independence granted to this authority vary considerably from country to country. Traditionally, legislatures have been allowed to draw their own districts. Increasingly, however, countries are turning the process over to independent commissions. This is part of an international movement to eliminate "politics" from the redistricting process.

Reforms that have replaced legislatures with redistricting commissions have also included provisions for increasing public access to the redistricting process and identifying formal criteria for commissioners to consider when drawing districts. These redistricting criteria are usually listed in the electoral laws of the country. The criteria often include factors such as equality of population, respect for local administrative boundaries and other geographic features such as natural (physically-defined) boundaries, and recognition of communities of interest.

Countries usually have not adopted redistricting criteria pertaining to the actual outcome of the redistricting process--for example, fair representation for political parties or minority groups. This is because single-member districts, used by many of the countries that delimit electoral districts, cannot guarantee results that are proportional, or even minimal, for minority political parties or minority racial, ethnic, religious or special language groups in the population. Some countries, however, have adopted special provisions designed to modify the distorting effects of single-member districts and to ensure some degree of representation for minority groups.

This section examines the choices countries have made for rules and a structure to carry out redistricting, including: 

Designation of a Boundary Authority

 

Countries that delimit electoral districts must designate a boundary authority and establish some machinery for carrying out the task of redistricting. The task assigned to the boundary authority is the same in all countries: divide the country into districts for the election of representatives.

The composition of the boundary authority and the degree of independence from the legislature or partisan concerns granted to this authority, however, vary considerably from country to country. Some countries allow legislators to draw their own districts. Other countries, in an attempt to remove "politics" from the process, assign the task of redistricting to an independent boundary commission. In some countries, redistricting is centralised under a single redistricting authority while in other countries, states or provinces draw their own districts, with or without a uniform set of rules. In many countries, the boundary authority is granted the power to choose the final districting plan. But in some countries with non-legislative boundary authorities, the legislature or the government must approve the final districting plan before it can be implemented.

The types of boundary authorities countries have established and the degree of independence countries have accorded these authorities cover a broad spectrum. At one end of the spectrum is the United States, where the redistricting process is very political and decentralised. The responsibility for drawing districts for the United States Congress rests individually with the fifty states. There are few limitations on the states, and the boundary authorities are almost always political entities, i.e., state legislatures.

At the other end of the spectrum are many of the Commonwealth countries, where politicians have opted out of the redistricting process and granted the authority to redistrict to neutral or independent commissions. A central agency may draw districts for the entire country. If the central agency does not actually draw the districts, it establishes guidelines for regional commissions and oversees the redistribution process. The final decision as to which district boundaries should be implemented often rests with the commission and not with the legislature.

This section will examine alternative approaches to the designation of a boundary authority. It will discuss the composition of the boundary authority, whether the authority should be partisan or non-partisan, and whether a central authority or regional authorities should perform the task of redistricting. Who has the authority to make the final decision as to which set of district boundaries are to be implemented will also be considered.

 

Composition of the Boundary Authority

A substantial majority of the countries that delimit electoral districts employ a specially designated boundary commission or an election management body to draw these boundaries. The legislature serves as the boundary authority in several countries. And in a few countries, government agencies are charged with the task of redistricting.

Britain was an early pioneer of establishing an independent boundary commission to define electoral districts[1].  Many established democracies once governed by the United Kingdom followed this example and adopted boundary (or delimitation) commissions, including Australia and Canada, Caribbean countries such as the Bahamas, Barbados, St. Lucia and St. Vincent and the Grenadines and several Anglophone African countries (i.e., Botswana, Namibia and Zimbabwe) also adopted boundary commissions for delimiting constituencies.

Composition of Boundary Commissions

Boundary commissions tend to be relatively small in size, ranging from three to seven or nine members. Canada, for example, has three-member commissions, the United Kingdom has four-member commissions, and a number of Caribbean countries have five-member commissions (e.g., Bahamas, Barbados). Germany and New Zealand each have seven-member commissions (New Zealand actually has 8 members but one, the Chair of the Local Government Commission, does not have a vote); Albania has a nine-member commission.

The commissions often include non-partisan (non-political) public officials with backgrounds in election administration, geography, and statistics. In Australia, New Zealand, and the United Kingdom, for example, the commissions incorporate electoral officers or registrar-generals, as well as the Director of Ordnance Survey (United Kingdom) and the Surveyor-General (Australia and New Zealand). Statisticians have an important role on Australian commissions because population projections are used to draw electoral district boundaries. In Canada, academics knowledgeable about elections and/or geography may be asked to serve on boundary commissions.

Members of the judiciary are also well represented on districting commissions in many countries. They often chair the commissions, as in Canada and New Zealand. In the United Kingdom, senior judges serve as Deputy Chairs of the four Boundary Commissions in England, Scotland, Wales, and Northern Ireland. In India, two of the three members of the Delimitation Commission are required to be judges.

Many countries with boundary commissions exclude anyone with political connections from serving on the commission. On the other hand, some countries specifically include representatives of the major political parties on the commission. For example, in New Zealand, two “political” appointees, one representing the governing party and one the opposition parties, serve on the seven-member Representation Commission. The theory behind their presence on the commission is that it helps ensure that any political bias in a proposed delimitation plan is recognized and rectified. However, because the two political appointees constitute a minority of the commission, they cannot outvote the non-political commissioners. Other countries that incorporate political party representatives on the boundary commission include Albania, Bahamas, Barbados, Fiji, Papua New Guinea, and St. Vincent.

Botswana is one of the countries that specifically excludes any person with political connections from serving on the boundary commission. Other examples include Australia, Canada, India, and Mauritius.

Election Management Bodies

Another, equally common, approach to delimiting constituencies is the use of the election commission. In some countries, the election commission is quite independent of the executive and the legislature (Lithuania, Mexico, and Poland, for example), but in other countries this is less true (e.g., Kenya, Nigeria, Tanzania).

Legislature

Although many countries have delegated the task of delimitation to an authority other than the obviously self-interested legislature, in some countries the legislature has retained this responsibility.

However, a number of countries in which the legislature is responsible for delimitation are countries with List Proportional Representation (List PR) electoral systems. The legislatures in these countries (Belgium, Bulgaria, Croatia, Finland, Iceland, and Sweden) originally defined a set of electoral district boundaries (often multimember districts) in the constitution or electoral law, and these constituencies have remained in place for subsequent elections – although the number of seats assigned to the multimember constituencies vary over time depending on the population size.

A second set of countries in which the legislature plays a role in the delimitation process are countries with mixed electoral systems like Italy, Korea, Kyrgyzstan, and Panama. The boundaries of the constituencies in these countries are of less political consequence than in those with a First Past the Post electoral system because a separate set of legislative seats are filled via proportional representation. Elections in mixed systems usually produce outcomes that are far more proportional than FPTP systems.

The United States is one of the very few democracies that allows the legislature a dominant role in the delimitation process given that the election of legislators is based solely on single-member constituencies. The consequence of this approach, at least in the United States, is that partisan politics plays a very large role – and often quite explicit role – in the redistricting process. For example, on several occasions when a redistricting plan was challenged in court on the grounds that the plan constituted a racial gerrymandering, defendants claimed that politics, and not race, was the motivating factor behind the plan hence the plan was neither illegal nor unconstitutional.

 

Notes:

[1] New Zealand established an independent boundary commission in 1887 that included government-appointed members; however, the government-appointed members never exceeded the number of politically neutral public servants included on the commission.

Partisan Considerations in Choosing a Boundary Authority

Since different electoral district configurations produce different election outcomes, even if voting patterns remain constant, the designation of a boundary authority and the powers granted to that authority are very important.

In the nineteenth century, districts almost everywhere were drawn by the legislatures. Legislators from the majority political party were often tempted to draw districting plans that favoured their own candidates at the expense of the candidates from other parties. The redistricting, process came under increasing attack for political bias because of this practice.

Over the past fifty years, a growing number of countries have moved to neutral or non-partisan commissions to avoid politically biased redistricting. In 1964, when Canada adopted independent electoral boundary commissions for redistricting, the United States was left as one of the few long-standing liberal democracies where the redistricting process remains in the hands of the politicians. The United States has also been one of the only countries to accept partisanship as an inevitable part of the redistricting process.

Disagreeing with the proposition that politics cannot be removed from the redistricting process, many countries have established neutral commissions and enacted rules for drawing district boundaries. These commissions have no formal links to political parties, and commissioners are not permitted to consider political data when drawing district boundaries. Instead, commissioners are obliged to consider criteria such as equal population and respect for administrative boundaries in deciding upon district configurations. Since the adoption of these reforms, few have questioned the neutrality of the redistricting process in these countries.

The only drawback to the non-partisan approach to redistricting is that it does not necessarily produce a politically unbiased redistricting plan: Ignoring politics does not mean that a districting plan has no political effect. The non-partisan approach guarantees only that any political bias in a districting plan is unintentional.

Solutions for Political Bias in Redistricting

Some countries have attempted to devise solutions to the problem of political bias in redistricting outcomes. One solution, for example, is to allow representatives of all the major political parties to serve on the body that makes redistricting decisions. Another solution is to permit representatives of major political parties to analyse the potential partisan impact of a districting plan before the plan is enacted and comment on the plans. In the United States, one or both of these approaches have been adopted by most of the states.

In the United States, a legislative committee or a bipartisan commission may draw a districting plan. Political data to ascertain the potential partisan effect of a plan will almost certainly be employed. Because a districting plan is almost always enacted by the state legislature, the likely result, however, is a plan that knowingly favours one political party or the incumbent legislators of both parties rather than a politically unbiased plan.

New Zealand has adopted another approach to the problem of politically biased outcomes: Two of the seven members of the New Zealand Representation Commission are political appointees. One of these two partisan members represents the governing party, the other one represents the opposition parties. Their presence on the commission helps to ensure that any egregious political bias is recognised and rectified. Because the two political appointees constitute a minority on the commission, they cannot outvote the non-political commissioners. The neutrality of the commission is, therefore, unquestioned.

Degrees of Boundary Authority Centralisation

Countries vary in the degree to which the delimitation process is centralised. At one end of the spectrum, the delimitation, or redistricting, process is very decentralised, with regional entities such as states or provinces responsible for drawing their own federal electoral districts. Little, if any, federal guidance is provided to these regional entities. At the other end of the spectrum are those countries in which a single central agency is charged with drawing districts for the entire country. In the middle of the spectrum are countries that have established central agencies, but these agencies do not actually draw federal electoral districts. Instead, they may establish guidelines for regional commissions to follow when drawing district boundaries, and they may oversee the federal redistricting process.

The United States is at one end of the spectrum. There the redistricting process is completely decentralised. Once the U.S. Congress apportions congressional seats among the states, each of the fifty states is responsible for drawing the allotted number of congressional districts within its own borders. Each state adopts its own redistricting procedures and determines its own redistricting criteria. There is some guidance from the federal government and the courts, however, but this guidance is limited for the most part to the areas of population equality and minority voting rights.

Australia and Canada, despite employing federal systems like the United States, have adopted redistricting procedures that are more centralised. In Australia, separate commissions for the redistribution of federal electoral districts were established in each state at the turn of the century. Canada borrowed this practice in 1964 when it determined that federal redistribution should be conducted by independent commissions established in each province. However, both Canada and Australia provide the state or provincial commissions charged with creating federal electoral districts with a uniform set of criteria for redistribution. Both countries also provide some degree of central co-ordination for federal redistribution.

Elections Canada, a permanent federal agency, coordinates the process of federal redistribution in Canada by bringing commission members (including the chairpersons) together for discussions before the process begins. Elections Canada also provides each provincial commission with a database for federal redistricting and trained support staff.

In Australia, the federal electoral commissioner--the administrative head and one of the three members of the Australian Electoral Commission--has a seat on each of the state redistribution commissions charged with federal redistricting. The other two members of the Australian Electoral Commission are added to form augmented redistribution commissions for federal redistributions within each state.

In most other countries, redistricting is centralised in a single federal agency that draws districts for the entire country. In Germany, a permanent seven-member constituency committee determines the district boundaries for the entire country (although each state has a voice in the process). In New Zealand, the seven-member Representation Commission undertakes redistribution for the entire country. In France, the Ministry of Interior drew district lines for the entire country when single-member districts were restored in 1986. And in India, a delimitation commission whose decisions were implemented in 2008, conducted the process administratively for the entire country.

Conclusion

A major advantage of a centralised redistricting process is that the redistricting criteria can be interpreted or applied without regional variation and, as a result, districts may be more uniform in construction. A major advantage of a decentralised process is that district boundaries are drawn by individuals more familiar with regional geography, communities of interest, and other local circumstances.

Authority for Choosing the Final Districting Plan

In the nineteenth century, in nearly every country that delimited districts, legislative approval was required before a redistricting plan could be implemented. Recent reforms designed to remove politics from the redistricting process have reduced the power of legislatures to approve redistricting plans. In many countries today, the legislature plays only a limited role or no role at all in the redistricting process. However, some countries do require executive approval, rather than legislative approval, before a redistricting plan can be implemented. While this removes the decision from legislators--those who directly benefit from the districting plan--it still leaves the redistricting process open to charges of political influence.

In the majority of countries that assign election management bodies the task of delimiting constituencies, the election commission serves as the final authority; the approval of the legislature or executive is not required to implement the delimitation plan. This is less true of boundary commissions – more often than not, a constituency plan proposed by a boundary commission must be enacted by the legislature (or signed by the executive) before it can be implemented. However, in New Zealand, for example, the final plan of the Representation Committee, once published, cannot be changed or appealed. Since 1983, Australia’s augmented Electoral Commission has had the same power. The constituency boundaries created by the Delimitation Commission in India are also final.

In other countries, the legislature can debate and, possibly, even delay the enactment of a commission's plan, but it cannot modify the plan. In Canada, for instance, the 1964 Electoral Boundaries Readjustment Act removed the responsibility for redistricting from the Parliament and gave it to independent electoral commissions in each province. Parliament is permitted to consider plans produced by the commissions but has no vote on their implementation. Parliament used this provision to delay the implementation of plans, however, prompting a change in the law--there is now a sixty day limit on Parliamentary debate and consideration.

In the United Kingdom, the final proposals of the four Boundary Commissions take effect only after an affirmative vote by Parliament. But Parliament's power to accept or reject a plan is a formality. It has almost always affirmed commission proposals; to do otherwise would be viewed as "political." The only two exceptions were in 1948, when Parliament proposed the addition of seventeen seats for under-represented urban areas, and in 1969, when Parliament delayed the implementation of a redistribution plan on the grounds that impending changes to local government boundaries would render the plan obsolete. Conservatives viewed both of these actions by the Labour government as political.

Some countries have provisions requiring the legislature to either accept or reject the proposed delimitation plan, but specifically do not grant parliament the authority to modify the plan. Examples of this approach include Malaysia, Mauritius, and Papua New Guinea.

The United States is the anomaly with regard to legislatures and the adoption of redistricting plans. Most states assign the task of federal redistricting to the state legislature. The few that assign the task to an agency or a commission still require an affirmative vote of the state legislature to enact a redistricting plan.

Conclusion

Allowing the legislature to accept or reject a plan--let alone create it--opens the redistricting process up to charges of political bias. In fact, many plans adopted by legislatures do favour one political party over others. Still, plans drawn by neutral commissions can also produce politically biased election outcomes, however unintentional.

Frequency of Electoral District Delimitation

The majority of countries that delimit electoral districts have established some mandatory time interval within which delimitation must occur. Although there is no standard time period, the range of intervals for delimitation is not particularly large. The Seychelles requires the delimitation of new constituency boundaries as often as every three years if necessary. On the other hand, France requires the delimitation of electoral districts only every 12 to 14 years.

The most popular choice for periodic delimitation appears to be ten years: Botswana, Canada, India (although amendments to the Constitution cancelled the delimitations following the 1981 and 1991 censuses and delayed the exercise following the 2011 census), Japan, Kenya, Lesotho, Malaysia, Mauritius, Mexico, Nepal, Nigeria, Pakistan, Papua New Guinea, Tanzania, the United States, and Yemen all have electoral laws or constitutional provisions requiring delimitation at least every 10 years (in the case of Botswana, the requirement is every five to ten years; in Kenya, the law dictates that delimitation occur every eight to ten years).

Albania, Bahamas, Fiji, New Zealand, Turkey, and Zimbabwe redraw their electoral districts every five years. Australia delimits at least every seven years. Ireland is required to delimit multimember constituencies for their Single Transferable Voting System every 12 years; the United Kingdom also permits up to twelve years to lapse before undertaking another delimitation exercise.

Of course, the establishment of a mandatory time interval does not necessarily mean that redistricting will occur. After delimiting constituencies in 1973, India placed a moratorium on delimitation until after the year 2000, despite a legal provision requiring redistricting after every decennial census.

No specific time interval has been established in about one third of the countries that delimit electoral districts. Common triggers for delimitation other than a specified time period include: following a national census, a change in the number of seats apportioned to an area, changes in administrative boundaries, and reaching a prescribed level of malapportionment. For example, in Macedonia, the degree of malapportionment cannot exceed three percent; if it does, delimitation must occur. In the Czech Republic, the prescribed level of malapportionment prompting a delimitation exercise is 15 percent; in Germany, the trigger is 25 percent.

The disadvantage of infrequent redistricting is the wide discrepancies in district populations that often result over time. Districts that are drawn with very small population deviations at the beginning of a decade, for instance, may vary considerably in size by the end of the decade.

Australia has adopted a unique solution to this problem. Rather than use current enrolment figures, federal electoral districts are redrawn using projections, such that the number of electors enrolled in each district will be equal three and one-half years into the future, the midpoint of Australia's seven-year redistricting cycle.

Conclusion

The disadvantages of frequent redistricting are the costs in money and time to undertake the process. In addition, frequent changes to district boundaries disrupt ties between representatives and their constituencies. To determine the optimal interval, countries must balance the costs of redistricting with the benefits of maintaining districts with some measure of population equality.

Length of Time Permitted for the Delimitation Process

While there are no time constraints on the delimitation process in some countries, other countries impose quite rigid constraints on the length of time permitted for the redistricting process. The existence or absence of a legal deadline, however, does not necessarily relate to the time actually taken to redraw districts.

Neither the United States nor the United Kingdom imposes a mandatory time limit for completion of the redistricting process. Although there is no federal law and few state constitutions that set time limits, candidate-filing deadlines for upcoming congressional elections serve as a practical deadline for federal redistricting in the United States. If a state legislature does not complete redistricting by the candidate-filing deadline, the courts will intervene and either draw a plan of their own or implement a districting plan proposed by one of the parties to the court case. Hence, the practical, if not legally imposed, time period for redistricting in the United States is two years, beginning with the date that numbers from the decennial census are released and ending with the candidate-filing deadline for the first congressional election of the decade.

In the United Kingdom, the redistribution process can take more than twice as long as it does in the United States. The last English Boundary Commission Report, completed in 2006, took six years to prepare.

By contrast, redistricting in Australia and New Zealand takes less than one year. In 1984, for example, Australia undertook an extensive redistribution. Districts for the 125 seats in the Lower House were redrawn to create 148 new districts. The entire process--drawing up proposals, holding public inquiries, and modifying proposals to produce a final plan--took only six months to complete. In New Zealand a representation commission is required file its report no later than eight months after formal deliberations begin.

Two factors that appear to affect the length of time needed for redistricting are the design of the public inquiry process and whether the legislature is permitted to debate or modify a redistricting plan. The elaborate public inquiry process utilised in the United Kingdom adds considerably to the time needed for redistribution. (See the case study of the United Kingdom, The United Kingdom Redistribution Process, for a description of the public inquiry process in that country.) In Canada, the major source of delay has traditionally been Parliament. A sixty-day restriction was eventually placed on Parliament for the review of proposed federal redistribution plans.

Given the political nature of redistricting in the United States, it is remarkable how quickly the process is completed. Legislatures in most states are responsible for drawing the lines and enacting the final congressional districting plan. Part of the reason for the timeliness in most states is the lack of any structured public input into the redistricting process. While some states did hold public hearings during redistricting in 1991, it is doubtful that these hearings led to significant modifications in any redistricting plans.

Conclusion

The major disadvantage of a lengthy redistricting process is the same as that associated with a lengthy time interval between redistributions, i.e., large population deviations may result. In England, for example, there were wide discrepancies in district populations following the 1983 redistribution because the Boundary Commission redrew constituencies with voter registration counts from 1976, the year the redistribution process began.

Public Access to the Delimitation Process

One distinction between countries with redistricting commissions and countries where legislatures or government agencies conduct redistricting is public access to the process. Many countries that have adopted neutral redistricting commissions have incorporated public access provisions as part of the reforms to limit the influence of legislators and political parties in the redistricting process.

One of the aims of Canada's Electoral Boundaries Redistribution Act of 1964 was to increase the public's awareness of and involvement in the redistribution process. The Act, modelled on Australia's redistribution process, borrowed Australia's practice of granting the public an opportunity to present suggestions or objections to commission proposals. In Canada today, once an independent electoral commission has completed its proposal and published the map in the local newspapers, the general public is invited to present written briefs or oral representations at public hearings held by the commission. Commissions have received thousands of comments from a wide variety of sources. Local jurisdictions, political parties, members of Parliament (MPs), candidates for Parliament, political activists and other interested citizens have all offered comments on proposed federal redistribution plans. Redistribution plans have often been revised after these hearings.

Although Australia's redistribution process has been modified since Canada borrowed major portions of it, public involvement is still an important part of the process. In 1983, for example, the process was changed to give the public two opportunities to offer comments on proposed redistricting plans. Australia's four-member Redistribution Committee receives suggestions from the public, political parties, candidates, and MPs before beginning to draft a plan. Once the Redistribution Committee has completed a plan, the augmented Electoral Commission hears public objections to the proposed plan, if there are any, and produces a final map. The only avenue for legislators in Australia to comment on a federal redistribution plan is through the public hearing process. See the case study on Australia, Federal Redistribution in Australia, for more details on the public inquiry process in this country.

In the United Kingdom, the process of public consultation is similar, although a public inquiry is held only if local authorities or at least one hundred electors object to the proposed map. Despite this caveat, the public consultation process takes much longer to complete in the United Kingdom than in other countries. One reason is that second inquiries may be held if a proposed plan is modified and new objections are raised. The public consultation process is described in detail in the case study of the United Kingdom, The United Kingdom Redistribution Process. In countries without standard procedures for public access, litigation may be the only avenue for the public to challenge a redistricting plan. In the United States, civil rights organisations, public interest groups, and interested citizens frequently file lawsuits if they deem a redistricting plan unfair. One consequence is that the number of redistricting lawsuits filed in the United States is enormous, far greater than in any other country. Increased public access to the redistricting process may or may not reduce the number of court challenges to redistricting plans in the United States since there are other reasons as well for the proliferation of lawsuits. But greater public access to the process would certainly make the process appear more open and democratic.

Conclusion

The primary advantage to granting public access to the redistricting process is that the outcome is more likely to be viewed as fair if the process is perceived as open and accessible. In countries with plurality or majority electoral systems, granting the public access to the process may be especially important because of the tendency of single-member districts to distort the relationship between the percentage of votes a political party receives and the number of seats the party wins.

Establishment of Criteria for Delimiting Districts

Countries often institute a set of formal rules, or criteria, for their boundary authorities to consider when drawing electoral districts. Although this is especially true in countries that employ boundary commissions or election commissions to draw districts, many countries that allow the legislature or a government agency to redistrict have enacted criteria as well. These rules are usually listed in the electoral law, but they can sometimes be found in the country's constitution. Examples of redistricting rules enacted by selected countries can be found in Election laws or constitutional provisions listing redistricting criteria for selected countries.

The rules often specify that districts should be as equal in population as possible. Administrative and/or natural boundaries and other geographic features such as sparsely populated or isolated territory are factors also commonly listed. Respect for communities of interest is another factor many countries specify. In some countries, especially developing countries, those in charge of redistricting are asked to consider the means of transportation and/or communication as well.

Almost all countries that have formal redistricting criteria require that districts be as equal in population as possible. Many countries specify tolerance limits or allowable deviations from the population or electoral quota. The population or electoral quota is simply the population of the territory to be redistricted divided by the number of legislative seats (or districts) to be allocated to that territory.

Although the representation of voters has taken precedence over the representation of communities in the twentieth century--as evidenced by the number of countries that require districts to be as equal in population as possible--respect for administrative areas, physically-defined natural communities and geographically concentrated communities of interest continue to play prominent roles in redistricting. In fact, a primary function of single-member districts is to provide representation for geographically defined communities.

Criteria Relating to Election Outcomes

Equal population, geographic considerations and communities of interest are criteria that relate directly to the process of creating districts. Other criteria relate to the outcome of the redistricting process--for example, requiring that district plans be drawn so that political parties are fairly represented or that racial, ethnic, religious or linguistic minorities have an equitable chance of representation. But countries that delimit districts usually do not adopt criteria relating to the fairness of the outcome. This is because countries with single-member districts can rarely meet these standards, if fairness of outcome is defined as proportional or near proportional representation for political parties and minority groups.

Conclusion

Redistricting criteria can conflict with one another. For instance, although almost all countries list population equality as a redistricting criterion, few countries actually have districts that are very close to equal in population. The reason is that other criteria have been deemed more important than strict adherence to population equality. Respect for administrative boundaries and natural communities, for example, often require districts to be smaller or larger than the electoral quota. Countries that establish redistricting criteria must either prioritise the criteria or accept certain inconsistencies in district configuration. Many countries choose to specify the most important criteria and/or place limits (such as tolerance limits) on the boundary authority, but still permit redistricters some discretion in balancing these criteria.

Equal Population in Redistricting

The most widely accepted rule for redistricting is that districts should be relatively equal in population. This is because representation by population is a central tenet of democracy, and, in countries that employ single-member districts, this rule translates into the principle of equal populations across districts. Equally populous districts are necessary if voters are to have an equally weighted voice in the election of representatives. If, for example, a representative is elected from a district that has twice as many voters as another district, voters in the larger district will have half as much influence as voters in the smaller district.

The degree to which countries require population “equality” and the population figure (for example, total population, citizen population, registered voters) that is used to determine equality differs across countries. Approximately half of the countries that delimit districts use “total population” as the population base for determining equality across electoral districts. Another third of the countries employ registered voters as the population base. Several European countries use citizen population as the relevant base for determining population equality. Lesotho uses the voting age population as the base; Belarus uses the number of voters in the previous election.

The degree to which countries demand population equality also varies. Many countries have no established tolerance limit regarding the extent to which constituencies are permitted to deviate from the population quota. Among those that have established limits there is a range from “virtually no deviation allowed” (the United States) to as high as a 30 percent tolerance limit (Singapore).

Deviations from Population Quotas

Minimal

The United States is unique in its adherence to the doctrine of equal population. No other country requires deviations as minimal as the “one person, one vote” standard that has been imposed by U.S. courts since the early 1960s. In the 1983 court case Karcher v. Daggett, the U.S. Supreme Court held that there is no point at which population deviations in a congressional redistricting plan can be considered inconsequential: “there are no de minimus variations which could practically be avoided but which nonetheless meet the standard of Article I, Section 2 [of the U.S. Constitution] without justification.” The Court went on to reject a New Jersey congressional redistricting plan that had a total population deviation of only .7 percent. Following this decision, most states interpreted Karcher as requiring the adoption of congressional redistricting plans with exact mathematical population equality or, at minimum, with the lowest possible population deviation. Although the courts later upheld the legality of some redistricting plans that had less than the absolute de minimus population variation possible, none of the plans upheld contained total deviations of even one percent.

Medium

Macedonia, with a Regional List PR electoral system and six electoral districts, comes closest to this strict standard, with allowable deviations of no more than plus or minus three percent from the population quota. New Zealand, Albania, and Yemen allow deviations of up to five percent from the population quota. Australia, Belarus, Italy, and the Ukraine specify 10 percent as the maximum allowable deviation.

The population requirement in Australia, however, is actually more complicated than a 10 percent tolerance limit: Australian election law also requires that electoral districts deviate by no more than 3.5 percent, three years and six months after the expected completion of the redistribution. This criterion was devised to produce equality of population halfway through the seven-year Australian districting cycle and to avoid wide discrepancies at the end of the delimitation cycle. To meet this requirement, the Australian delimitation commission (referred to as the Redistribution Commission) must use population projections as well as current population data. Australia's close attention to population equality is relatively recent. Thirty years ago, the practice of heavy rural loading--creating rural districts that were much smaller in population than urban districts--was quite common. (For more information on Australian redistricting practices, see the case study on Australia, Federal Redistribution in Australia.)

Large

Armenia, Germany, and the Czech Republic allow population deviations of no more than 15 percent. (In Germany, proposed electoral districts cannot deviate by more than 15 percent, and districts that deviate by more than 25 percent must be redrawn.) Zimbabwe and Papua New Guinea have set the maximum allowable deviation at 20 percent. In Canada, the independent commissions charged with creating federal electoral districts are allowed to deviate by up to 25 percent from the provincial quotas. But since 1986, commissions have been permitted to exceed the 25 percent limit under "extraordinary circumstances." This provision was used to create five of the 295 seats in the Canadian House of Commons in 1987, two of 301 seats in 1996 and two of 308 seats in 2003. In 2003, one Ontario district was 43.7 percent below the provincial average and one Newfoundland district was created with a population 61.9 percent below the provincial average (For more information on Canadian redistribution, see the case study on Canada, Representation in the Canadian Parliament.) The United Kingdom allows even larger deviations in district populations. The original standard was set at 25 percent in 1944. But the standard was repealed only two years later. The current rule however requires that constituencies be "within plus or minus five percent of the United Kingdom quota", but this rule must be balanced against other factors such as local boundaries and special geographical considerations. Allowances for natural communities prompted English boundary commissioners in 1983 to leave the Isle of Wight with 95,000 electors as a single constituency, while respect for local London boundaries left suburban Surbiton with only 48,000 electors. The last redistribution saw the Isle of Wight apportioned an additional district, the two districts having 56,000 and 55,000 electors respectively. Likewise, recognising the difficulties of island travel, the commissioners in Scotland granted the Western Isles (2011 population 22,000) and Orkney and Shetland (2011 population 34,000) their own representatives.

Conclusion

The degree to which a country adheres to strict equality of population is related to the significance attached to individual political equality. The United States is strongly committed to individual rights, so perhaps it is not surprising that it developed the strictest population deviation standards of any country using single-member districts. Other countries, while recognising the importance of population equality, have chosen to balance this factor against other redistricting criteria perceived as equally valid. In the United Kingdom, respect for local administrative boundaries is given precedence over exact equality of number. In many African countries, the need to keep individual tribes intact in a single electoral district may take precedence over population equality. Each country must determine how much variation from the ideal of exact population equality will be tolerated to accommodate other redistricting goals.

Geographic Criteria for Delimiting Electoral Districts

In many countries, the electoral laws specify that geography, or certain geographic factors, be taken into account when delimiting electoral district lines. Geographic criteria can be divided into two categories--criteria relating to geographic boundaries and criteria relating to geographic size and/or shape. A boundary authority may be asked to consider factors from either or both criteria.

Criteria Related to Geographic Boundaries

Respect for clearly established boundary lines is often specified as a criterion for those redistricting to consider when drawing electoral district lines. These boundaries can include administrative boundaries such as county and municipality lines and/or natural boundaries created by dominant topographical features such as mountain ranges, rivers or islands.

Perhaps the most commonly mentioned geographic factor listed by countries is consideration for local administrative boundaries. Dozens of countries list this as a criterion to consider including: Albania, Bangladesh, Barbados, Bulgaria, Cameroon, Canada, Croatia, Czech Republic, Fiji, France, Germany, India, Indonesia, Italy, Japan, Kenya, Lithuania, Malaysia, Mexico, Pakistan, Panama, Tanzania, Uganda, United Kingdom, and Yemen. Botswana’s Constitution specifies consideration of not only administrative district boundaries, but the boundaries of tribal territories.

Another geographic feature commonly listed is population density or sparseness of population. Several Caribbean countries as well as Kenya, Mauritius, Nepal and Papua New Guinea identify this as a factor to take into account when redistricting. In Malaysia, the Election Commission is required to weight sparsely populated rural constituencies in a manner to guarantee their over-representation in the legislature.

Geographic redistricting criteria such as respect for administrative boundaries and physically defined natural communities are a higher priority in some countries than in others. In the United Kingdom, for example, respect for local administrative boundaries and natural communities is the most important concept guiding boundary commissioners. Large population disparities are tolerated as a result.

Criteria Related to Geographic Size and Shape

Two other factors that are sometimes listed as redistricting criteria relate specifically to the geometric shape of a district: contiguity and compactness. Advocates of these criteria hold that districts should not be oddly shaped and that all pieces of a district should be inter-connected. The election commission in Mexico, for example, is required to create electoral districts in which the perimeters are regular in shape. Other countries that specify that constituencies be compact include Albania, Armenia, Bangladesh, Barbados, Belarus, Dominican Republic, India, Italy, Pakistan, and the United States.

In the United States, district compactness has not been required by federal law since 1929, but when a number of states created some bizarrely-shaped districts in the1990s round of redistricting, the U.S. Supreme ordered the redrawing of a number of these districts. Although the shape of these districts was not actually the basis for the Court's decision, the fact that the districts were not compact was considered evidence of an impermissible motive in creating the district boundaries. (For additional discussion of these court cases see Role of the Courts in Electoral District Delimitation.)

Special Provisions for Minority Groups When Delimiting Electoral Districts

Criteria specifying fairness for minority groups within a country focus on the electoral outcome, rather than the process, of redistricting. Electoral systems that rely on single-member districts, however, cannot guarantee proportional representation or even some minimal percentage of seats for ethnic, racial, religious or other minority groups in the population. This is particularly true of electoral systems that rely solely on single-member districts for the election of representatives (i.e., FPTP and AV systems). On the other hand, List PR and Mixed systems – such as Parallel and MMP systems – can accommodate requirements for minority representation within the context of the party lists if so desired.

In districted systems, voters of a specific minority group will find it very difficult to elect members of their group to legislative office if voting is polarized along majority-minority lines. Only if separate seats are reserved for this minority group, or if special electoral districts are drawn for the group, will minority voters succeed in electing minority representatives. A few countries have made such special provisions to ensure that racial, ethnic, or religious minorities are represented in the legislature. Examples of these countries include Croatia, Fiji, India, Mauritius, New Zealand, Pakistan, the Palestinian Territories, Papua New Guinea, Singapore, and the United States.

Croatia, which has a List PR electoral system with electoral districts that are not typically redrawn, reserves specific districts for members of the (1) Hungarian, (2) Czech and Slovak, and (3) Ruthenian and Ukrainian and German and Austrian minorities. In addition, three seats are specifically reserved for the Serbian minority within the Republic of Croatia.

In the Block Vote (or Party Block Vote) systems of Mauritius, Singapore, and the Palestinian Territories, a number of seats are reserved for minorities:

  • Singapore – Most members of parliament are elected through a “Party Block Vote” in multimember Group Representative Constituencies (GRCs). Parties contesting a GRC must propose a slate that includes at least one member of an official minority (listed as Indian, Malay, Eurasian, or Other). Within the GRCs, voters select from among closed party lists, with the party receiving a plurality of votes winning all seats in the district.
  • Mauritius – In addition to the 62 representatives elected from 21 multimember constituencies, there are a maximum of eight additional seats allocated to the “best losers.” These “best loser” seats are apportioned among four constitutionally recognized ethnic or religions communities (Hindus, Muslims, Chinese, and “Creole”) to ensure some representation for each of these minority groups.
  • Palestinian Territories – The West Bank and Gaza Strip are divided into 16 multimember electoral districts. In the 1996 elections, the political party obtaining the greatest number of votes in each district took all the seats allocated to the district. Six seats across four districts (Jerusalem, Bethlehem, Ramallah, and Gaza) were reserved for the Christian population; one seat (in the Nablus district) was set aside for the Samaritans.

India and Pakistan, both with FPTP electoral systems, have specifically reserved single-member districts to ensure the representation of certain minorities:

  • Pakistan – There are three categories of seats in the National Assembly: (1) 272 general seats; (2) 60 seats reserved for women; and (3) 10 seats reserved for non-Muslims (Hindus, Christians, and others). Representatives of the general seats are elected by simple majority on the basis of 272 single-member constituencies. The seats reserved for women are filled on the basis of a proportional representation system based on the number of general seats won by each political party by province. The seats reserved for non-Muslims are filled under the same proportional representation system, except that the entire country constitutes a single constituency. Both women and non-Muslim candidates are chosen from closed lists filed by the political parties.
  • India – A certain number of parliamentary constituencies in each state are reserved for members of Scheduled Castes and Scheduled Tribes based on their proportion of the total state population. In reserved constituencies, only candidates from these communities can stand for election. These reserved constituencies shift from one election to the next. In total, there are 79 parliamentary seats reserved for scheduled Castes, and 41 seats for Scheduled Tribes.

Fiji and Papua New Guinea, each with Alternative Vote systems, have separate sets of communal seats to guarantee representation of the major ethnic groups. In Fiji, for example, the 71 legislative constituencies are comprised of 46 “communal” constituencies and 25 “open” constituencies (where all eligible voters, regardless of race/ethnicity, caste votes), with the “communal” members elected as follows:

  • 23 elected from a roll of voters registered as indigenous Fijians, 19 elected from a roll of voters registered as Indians, one elected from a roll of voters registered as Rotumans, and * three elected from a roll of voters not registered as Fijians, Indians, or Rotumans (this is the “general voters” roll).

Minority Representation in the United States

The United States, because of its sizeable racial and ethnic minority population and its history of discrimination against certain minority groups, has had to address the issue of fairness to minorities in promulgating redistricting plans. The Voting Rights Act of 1965 and its amendments in 1982 have established that a redistricting plan that dilutes the voting strength of minority voters by dividing the minority community among different districts may be invalid. Protected minority groups (blacks, Hispanics, Asians, and Native Americans) must meet three conditions to qualify for this protection:

  • the group must be sufficiently large and geographically compact to form a majority in a single-member district; the group must be politically cohesive (they must share common political interests); the group must be able to demonstrate that the majority population votes as a bloc against the minority community's preferred candidates and that the minority-preferred candidates usually lose.

If a minority group is able to satisfy all three of these conditions, a redistricting plan must be fashioned such that minority voters constitute a majority of voters in one or more districts. The minority community must demonstrate that these conditions are satisfied in a court proceeding. In fact, in a series of recent court decisions, the U.S. Supreme Court held that several jurisdictions that created "majority minority" districts voluntarily--that is, without being required by a court to do so--must redraw these "majority minority" districts without taking race or ethnicity into account.

The Voting Rights Act guarantees racial and ethnic fairness in some minimal sense in the United States. It is minimal because only minority communities that are able to satisfy all three of the conditions are given an opportunity to form the majority of a district and elect a candidate of choice. Blacks, Hispanics, Asians, and Native Americans are far from proportionally represented in the United States Congress. The minority community in New Zealand is better represented in the legislature because of a more effective provision.

Minority Representation in New Zealand

A unique feature of New Zealand's electoral system is a provision for representation of the descendants of New Zealand's aboriginal Maori population. In addition to sixty general legislative districts, the Representation Commission creates several Maori districts (five Maori districts were created in 1993, six in 1998, and seven in 2001, for example). These Maori districts are geographically defined and overlay the general electoral districts. To vote in a Maori district, rather than a general election district, a Maori voter must register on the Maori roll. Registration on this roll is optional; Maoris can choose to register on the general roll instead. Because of this electoral feature, Maoris have been represented in the legislature roughly in proportion to their percentage of the population for more than a decade. (See the case study on New Zealand, Electoral Redistribution in New Zealand, for a more detailed description of this provision.)

Conclusion

Countries that delimit single-member districts cannot guarantee proportional representation to minority political parties or to minority groups within their borders, at least not without special provisions or additional seats elected by a party list vote. Instead, redistricting criteria may be adopted to ensure a fair and impartial redistricting process. Although this will not necessarily produce proportional, or even minimal, representation for minority parties or groups, it does guarantee that any bias is unintentional.

Countries with deep racial, ethnic or religious divisions usually opt for some form of proportional representation rather than relying on single-member districts to elect representatives. Unless the minority group is geographically concentrated or special provisions for minority representation are adopted, the election outcome produced by single-member districts will benefit some groups at the expense of others. In a deeply divided country, this fact may well lead to instability rather than foster strong and stable governments.

Communities of Interest: Delimiting Boundaries

Because of requirements that single-member districts be relatively equal in population, single-member districts often do not reflect distinct geographic communities as signified by municipal, county or other administrative boundary lines. This does not mean, however, that political representation has been divorced from the notion of "community" in countries that delimit single-member districts.

Many countries that delimit single-member districts continue to emphasise the importance of creating districts that correspond as closely as possible to pre-existing communities, defined as administrative divisions and/or "communities of interest." The rationale for recognising communities in redistricting is that electoral districts should be more than conglomerations of arbitrary, random groups of individuals. Districts should, as much as possible, be cohesive units with common interests related to representation. This makes a representative's job of articulating the interests of his or her constituency much easier.

Defining Communities of Interest

A "community of interest" is rarely defined by statute but it is generally thought of as a group of individuals united by shared interests or values. These shared interests may be the result of a common history or culture, a common ethnic or tribal background, or a variety of other ties that create a community of voters with distinct interests.

Although the perimeter of a community of interest may correspond to the boundaries of an administrative division, this is not necessarily the case. For example, a river may form a boundary between two administrative divisions, but the entire river valley may comprise a unified community of interest. In this instance, an electoral district that follows the administrative boundary would divide a community of interest.

In general, criteria related to communities of interest can be divided into three categories: (1) criteria related to administrative or geographic boundaries; (2) criteria related to common interests or common characteristics; and (3) criteria related to patterns of interaction. Criteria related to administrative or geographic boundaries are discussed under Geographic Criteria for Delimiting Electoral Districts.

Some of the criteria related to common interests or characteristics are:

  • Shared racial or ethnic background
  • Common history and/or culture
  • Common religion or language
  • Shared socio-economic status
  •  

Some of the criteria related to patterns of interaction are: 

  • Transportation patterns
  • Economic ties
  • Communication networks (media markets)
  •  

Most countries’ electoral laws do not elaborate on what specific communities of interest are relevant to delimitation; the boundary authority is simply instructed to take into account “communities of interest.” German electoral law states that constituencies should form a “coherent” area. Nepal, Pakistan, and Papua New Guinea electoral law instruct the boundary authority to consider “community and diversity of interest” or “homogeneity and heterogeneity of the community.” Australian electoral law offers more guidance, stating that the Redistribution Committee shall give due consideration to “community of interests within the proposed Electoral Division, including economic, social and regional interests.”

A handful of countries offer more explicit instructions as to what communities of interest are particularly pertinent when delimiting constituencies. In Hungary, for example, the boundary authority is to take account of ethnic, religious, historical, and other local characteristics when creating electoral districts. Panama and the Ukraine also require consideration of minority populations: in the Ukraine, the “density of national minority populations” is to be taken into account; in Panama, “concentrations of indigenous populations” must be considered. Minus electoral law provisions specifically designed to promote minority representation, however, criteria requiring “due consideration” of the minority population is likely to have little impact on integrating the halls of government with minority representatives.

Conclusion

Redistricting criteria inevitably conflict with one another. One possible means of resolving a conflict between criteria is to determine the most salient or most important "community of interest" in a given instance. Public hearings are essential to this process. For example, a redistricting plan that follows ethnic community boundaries rather than administrative boundaries may prevail if members of the public assert that the ethnic community boundaries are more relevant to them than administrative boundaries.

Role of the Courts in Electoral District Delimitation

 

It appears that the courts have no role at all in the delimitation process in the majority of countries that delimit electoral districts. In fact, in some countries, such as Pakistan and Tanzania, there is a specific bar against court involvement in the delimitation process. Other countries grant the court some function in the delimitation process, although in some instances, only in a very limited capacity. Examples of countries in which the court has a role in the delimitation process include Australia, Canada, Czech Republic, Fiji, France, Indonesia, Ireland, Japan, Lithuania, Mexico, New Zealand, Nigeria, Uganda, United Kingdom, and the United States.

Delimitation plans can be challenged, and have been to a limited degree, in the courts in Nigeria, Uganda and other Anglophone African countries. In Fiji, judicial review is permitted, but no one has challenged a delimitation plan to date. The only court challenge to a delimitation plan filed to date in the United Kingdom was unsuccessful, and this appears to have discouraged subsequent litigation on the issue of fairness of a delimitation plan or the delimitation process in this country. The Canadian courts have only recently ventured into consideration of delimitation acts; the first challenge to a federal electoral district plan was filed in Canada in 1987. In May 2004, the Federal Court of Canada made its decision in Raîche v. Canada (Attorney General), concerning a portion of the electoral boundary between the ridings of Miramichi and Acadie–Bathurst. The Court held that, in transferring certain parts of parishes from the riding of Acadie–Bathurst to Miramichi, the Federal Electoral Boundaries Commission for New Brunswick erred in its application of the rules governing the preparation of its recommendations, that is, the rules governing minority language protection under the Official Languages Act.

The major exception to limited judicial involvement is the United States, where the courts have decided hundreds of cases brought against congressional and state legislative districting plans.

Britain

In 1982, the Labour Party brought suit against the English Boundary Commission, challenging the Commission's newly completed redistribution plan. The Labour Party argued that the commission had given too much weight to "natural communities" and county boundaries in the plan and too little weight to ensuring equal electorates. There were, in fact, large disparities in population across constituencies. Both the Isle of Wight and the London suburb of Surbiton, for example, were designated as single seats, but the Isle of Wight had an electorate of 95,000 and Surbiton had only 48,000 electors.

The court, however, in its decision in R. v. Boundary Commission for England ex parte Foot[1], found no evidence that the commission had failed to undertake its statutory obligation to ensure equality of numbers. The court found that the boundary commission necessarily enjoyed a considerable degree of flexibility in interpreting redistribution rules. Furthermore, the court indicated a reluctance to interfere in a sphere that was clearly within Parliament's jurisdiction. To date, the court in Britain has not been asked to consider the fairness of another redistribution plan.

United States

American courts entered the political thicket of redistricting in 1962 when the United States Supreme Court ruled in Baker v. Carr that voters could challenge redistricting plans. Prior to this decision, the courts had refused to become involved in the line drawing process. The courts considered redistricting to be a political question, best resolved by the state legislatures.

Since the Baker decision, the courts have become active participants in the redistricting process to an extent unparalleled in any other country. Courts have established many of the rules that govern the redistricting process in the United States. These include rules on equal population, minority voting rights, political and racial gerrymandering, and numerous provisions of individual state redistricting laws. In addition, the courts are frequently called upon to draw district boundaries when the legislatures are unable to agree on redistricting plans that satisfy legal or constitutional requirements.

The United States Supreme Court's initial involvement with the redistricting process concerned the issue of equal population between districts. During the early part of the twentieth century, the American population was largely transformed from a rural to an urban majority. Politicians from rural areas, afraid of losing power and representation in the legislatures, either refused to redraw districts or redrew districts that clearly favoured the rural minority. As a result, legislative districts were often severely malapportioned. Although originally hesitant to address the issue, the Supreme Court ultimately determined, beginning with Baker, that large inequalities in district populations violated the Fourteenth Amendment of the U.S. Constitution.

In the case of Wesberry v. Sanders (1964), the Supreme Court ruled that the U.S. Constitution required congressional districts to have populations that were as "nearly equal as practicable." This standard was further refined in Karcher v. Daggett (1983), in which the Court rejected a congressional redistricting plan for the State of New Jersey that contained a population deviation between districts of less that one percent. Under Karcher, unless a congressional redistricting plan contains the least possible population deviation between districts, a state may be required to prove that the deviation was necessary to achieve a legitimate goal. The practical effect of this ruling has been to require states to draw congressional districts with nearly the exact same populations.

Equality of population is not the only redistricting criterion American courts have addressed. The United States Supreme Court has also recognised the right of voters to challenge redistricting plans as dilutive of minority voting rights under the Voting Rights Act or as unconstitutional partisan or racial gerrymanders under the Fourteenth Amendment.

In Davis v. Bandemer (1986), the United States Supreme Court ruled that a redistricting plan that discriminates against an identifiable political group or party may violate the U.S. Constitution. The Court recognised, however, the highly partisan nature of the American redistricting process and imposed a very difficult burden on voters who make such a claim. To succeed, voters must prove that they have been denied any influence in the electoral system and that they are substantially shut out of the political process. Despite a number of challenges, no congressional or state legislative redistricting plan has been invalidated by the courts on the grounds that the plan constitutes a partisan gerrymander.

Voters seeking redress for redistricting plans that dilute minority voting strength have been far more successful, until recently. The Voting Rights Act of 1965 was designed to prevent the dilution or abridgement of the voting rights of minority voters. The Act was amended in 1982 to make it clear that redistricting plans that diluted minority voting strength were illegal. In Thornburg v. Gingles (1986), the Supreme Court was asked to consider the 1982 amendments to the Act. The Court ruled in Gingles that to succeed on a voting rights act claim, minority voters must establish three factors:

  • the minority group must be sufficiently large and geographically compact to constitute a majority in a single-member district;
  • the minority group must be politically cohesive;
  • the white majority must vote sufficiently as a bloc to defeat the minority group's preferred candidates.
  •  

The Gingles decision established a clear and objective standard for minority claimants to satisfy, which encouraged minority groups to bring suits against redistricting plans that they felt were discriminatory. If minority groups were able to satisfy the three Gingles factors, the courts required the redrawing of the district boundaries. This led to a substantial increase in the number of "majority minority" districts and to an increase in the number of minority representatives elected to office. These gains in minority representation, however, have recently been threatened.

In a string of court cases beginning with Shaw v. Reno in 1993, the United States Supreme Court ruled that voters could challenge a redistricting plan that contained "majority minority" districts drawn on the basis of race. If voters can prove that race was the predominant motivating factor in the drawing of particular districts, a state must show that the challenged districts were "narrowly tailored to further a compelling state interest." This standard has proved virtually impossible for states to satisfy. The effect of Shaw and successive decisions has been to place in jeopardy numerous majority black and Hispanic districts that were drawn following the 1990 census and to make it more difficult to create such districts in the future. (For a more detailed discussion of the role of the U.S. courts in the creation of "majority minority" districts, see US: Ethnic Minorities and Single-Member Districts.)

In a single generation, American courts have moved from declining to exercise jurisdiction in redistricting disputes to participating in the redistricting process as one of the key players. Today in the United States, it is not at all unusual for a redistricting plan not only to be challenged in court but even to be drawn by a court.

And even though the plan may be drawn by a court, that, in itself, does not mean that the court-drawn plan will go unchallenged. The Supreme Court of the United States set aside election maps drawn by a federal court in Texas in 2012. The Supreme Court said that “a district court should take guidance from the state’s recently enacted plan in drafting an interim plan” paying attention to any parts that might violate the Constitution and the Voting Rights Act.

Canada

The role of Canadian courts in the redistricting process has been minor compared to the very active role played by courts in the United States [2]. In fact, it was only recently that Canadian voters could even request that the courts consider the fairness of an electoral boundaries plan. Prior to the passage of the Canadian Charter of Rights and Freedoms in 1982, opponents of a redistricting plan had no recourse in the courts. The charter provided the first constitutional mechanism for challenging electoral boundaries and the legislation under which electoral commissions carry out their mandates.

The first case to accept the justiciability of the issue of fairness of an electoral boundaries map was Dixon v. Attorney General of British Columbia. The Dixon case, decided in 1989, involved a challenge to British Columbia's provincial electoral map. The B.C. Supreme Court found that the province's electoral districts, varying in population from 5,511 to 68,347, violated the right to vote guaranteed by Section 3 of the Charter of Rights and Freedoms. The province had used a complex quota system rather than a tolerance limit, such as the 25 percent rule that guided the drawing of federal electoral maps. The British Columbia Supreme Court decreed that "equality of voting power is the single most important factor to be considered in determining electoral boundaries." It ruled that a new set of districts with more equitable populations must be created.

A second challenge to a provincial map was filed two years later in Saskatchewan. The Saskatchewan Court of Appeal found the electoral boundaries to be unconstitutional on the grounds that the right to vote includes the requirement that "one person, one vote" must be the ideal in evaluating electoral distribution schemes. It objected to constituencies that ranged in population from 6,309 to 12,567 electors and to the legislatively prescribed distribution of constituencies among urban and rural seats. The decision was appealed to the Supreme Court of Canada, which reversed the lower court decision in Reference Re Provincial Electoral Boundaries, Saskatchewan (1991), commonly referred to as the Carter decision.

In Carter, the Supreme Court of Canada reinstated the Saskatchewan district plan and held that "the purpose of the right to vote enshrined in Section 3 of the Charter is not equality of voting power per se, but the right to 'effective representation.'" In rejecting a strict population equality requirement, the Court indicated that effective representation could be achieved by "relative parity of voting power," modified where necessary to take into account other valid factors. In the Court's view, geography, community history, community interest, and minority representation should also be considered when redrawing district boundaries "to ensure that our legislative assemblies effectively represent the diversity of our social mosaic."

The courts have been called upon to render opinions on the constitutionality of provincial electoral maps or the legislation under which the commissions carry out their mandates in other provinces as well. For example, in 1991, the Alberta Court of Appeal upheld electoral boundary legislation which included a variance rule of 25 percent, with some exceptions for sparsely populated areas. In a subsequent case decided in 1994, the Alberta Court of Appeal upheld an electoral map produced by a committee of the legislature that had undertaken its work independently of the prescribed readjustment rules.

The Prince Edward Island provincial electoral map that replaced an earlier, unconstitutional map was upheld in 1996. The original provincial map, which had remained largely unchanged for a hundred years, and contained very large differences in population across districts, was struck down in Mackinnon v. Prince Edward Island in 1993. Its successor map, while containing no population deviations in excess of 25 percent, was challenged on the grounds that it over-represented rural areas and failed to conform to the municipal boundaries of cities such as Charlottetown. The Prince Edward Island Supreme Court, however, in City of Charlottetown et al. v. Prince Edward Island et al. (1996) upheld the map.

In addition to the  issue of voter equality, in 2004, the court ruled  on the application of the rules governing the preparation of a commission’s recommendations related to the rules governing minority language protection (Raîche v. Canada (Attorney General)). The Canadian courts have clearly chosen not to follow the path of American courts and their very strict adherence to population equality. Whether Canadian courts can avoid the profusion of redistricting lawsuits on other redistricting issues that has been experienced in the United States, however, remains to be seen. An explanation of the relative paucity of court challenges in Canada could simply be the relatively short history of the Canadian Charter of Rights and Freedoms.

Alternatively, the explanation for the relatively few number of court challenges may lie in the fact that the Supreme Court of Canada has accepted both the 25 percent population deviation limits and the "extraordinary circumstances" clause permitting deviations in excess of 25 percent. This acceptance may provide independent electoral commissions with sufficient leeway to save their maps from legal challenge [3].

Conclusion

In most countries, court challenges to electoral district boundaries are few in number or non-existent – either there is no right to appeal a redistricting plan to the courts or the grounds for such an appeal are very limited. It is only in the United States that the courts play a major role in the redistricting process. But in the United States redistricting tends to be very partisan, and public access to the process is very limited. Often, the only recourse voters have to challenge a plan is through the courts. The courts, at least in theory, serve as the safeguard against egregiously unfair redistricting plans in the United States.

Notes:

[1] The discussion of R. v. Boundary Commission for England ex parte Foot is based on Robert Waller, "The 1983 Boundary Commission: Policies and Effects," Electoral Studies 2, no. 3 (1983): 195-206. [2] The discussion of the role of the courts in Canadian redistribution draws heavily from a paper by Jennifer Smith entitled "Drawing Electoral Boundaries in Canada: Current Representation Dilemmas." This paper was presented at a conference hosted by the National Center for Geographic Information and Analysis in Buffalo, New York, October 24-26, 1997. [3] This observation was offered by John Courtney, a Canadian academic with considerable experience with and insight into the redistribution process in Canada.

 

Tasks Involved in Drawing Electoral District Boundaries

The process of drawing electoral district boundaries is time consuming and labour-intensive. The delimitation process usually begins with an allocation of seats to sub-regions of a country, such as states or provinces. Seats are almost always allocated to regions on the basis of population. But under special provisions, certain regions may receive more or fewer seats than population alone would dictate.

Once seats have been allocated, the process of drawing district lines within a region commences. A redistricting database is created using population data and, in some countries, political data as well. Maps are collected. After all of the necessary information has been gathered and synthesised, the process of assigning geographic units to electoral districts can begin. Each geographic unit--whether it is a county, city, town or village, or some smaller geographic census unit or voting area--is assigned to a specific district. After all geographic units in the region have been assigned to one and only one electoral district; the plan is complete and ready to be evaluated.

The process for evaluating a redistricting plan depends in large part on the redistricting criteria that have been adopted. A statistical summary of population or other demographic data by district is straightforward as long as the requested data has been included in the redistricting database. The plan may require a more sophisticated or subjective assessment, as well, depending on the criteria adopted. For instance, are communities of interest intact? Do minority voters have an opportunity to elect candidates of their choice?

Computers can be used to make the drawing of district boundaries more accurate and efficient. Consultants can be brought in to assist in any aspect of the redistricting process. A decision on whether computers or consultants should be employed depends on the need for them and what they will cost.

This section discusses the information needed to conduct redistricting: supporting the redistricting process, population data, maps and political data (see Information Required to Draw Electoral Districts). It outlines steps in the line drawing process, including the allocation of seats, the preparation of a database, the drawing of district boundaries, and the summary description of the plan for evaluation (see Steps in the Electoral District Delimitation Process). And it considers the possible use of computer technology and/or consultants (see Use of Computers and Software for Delimitation and Use of Consultants).

Information Required to Draw Electoral Districts

Delimitation, or redistricting, requires the collection of several different types of information. The two essential pieces of information are population data and maps. The population data, which may be in the form of census enumeration data or voter registration data, provide the only means of creating districts that are relatively equal in population. The population data must be associated with a specific geographic area and must be as accurate and up-to-date as possible. Maps are needed to ensure that only contiguous geographic population units are assigned to districts.

A third piece of information that may or may not be utilised for redistricting is political data. Political data usually refers to election results --tabulations of votes for candidates and ballot measures from previous elections by voting area. Including political data in the redistricting database allows line drawers to produce a political profile of proposed districts and to predict, to some degree, the partisan implications of a redistricting plan.

Election results can easily be entered into the redistricting database if they are reported for the same geographic unit as the population data. This will likely be the case when the population units for redistricting are based on voter registration data. If, however, the population units are based on a census enumeration, the geographic units for population and political data may not be the same. In that case, census geography and election geography may have to be matched in some manner to create geographic units that can be associated with both population and political data.

In the United States, for example, electoral districts are usually created using census geography (census blocks or tracts), but election results are reported at the voting area (election precinct) level. These two units of geography--census blocks and election precincts--are not equivalent. States that wish to use political data in conjunction with population data must develop some method of matching political data with the corresponding units of census geography (see also GIS - Geographical Information Systems).

Use of Population Data in the Delimitation Process

 

Population data is essential to the redistricting process. Redistricting operates by combining discrete geographically-based units of population in a series of assignments designed to produce districts of relatively equal population. The population data can be either total population counts based on a census enumeration or voter registration data; the geographic population units will therefore reflect either census geography or election geography (these units of geography are often, in fact, the same).

The Use of Census Enumeration Data

In the United States, the reapportionment of the House of Representatives of the Congress and the redrawing of congressional districts have always been based on a census count of the total population. The number of seats to which each state is entitled in the U.S. Congress is dependent on the relative size of the state's population according to the census enumeration. The U.S. Constitution requires the Census Bureau to conduct an enumeration of the entire population every ten years.

Therefore, a national census is conducted as of April 1 in the first year of each decade (for example, 1980, 1990, and 2000). The Bureau must report the results of the census to the president by December 31 of the census year. Since 1975, the Census Bureau has also been required to send population data specifically for redistricting purposes to each state governor within a year of conducting the census.

The redistricting data produced by the Census Bureau is referred to as PL 94-171 data (for Public Law 94-171). It includes population counts for several different levels of census geography, i.e., units as large as whole counties and units as small as census blocks (usually the equivalent of one city block) are reported in PL 94-171.

Although the Census Bureau collects data on a large number of population characteristics, only a few items are included in the data files sent to the states for redistricting. These items are total population, voting age population, and subtotals of the population for persons of Hispanic origin and for five major racial groups--white; black; Asian and Pacific Islander; American Indian, Eskimo, and Aleut; and "other" races. Population data by race and Hispanic ethnicity are needed to ensure that communities in which these minority groups predominate are not divided by electoral district boundaries. (An example of PL94-171 population data reported at the city level can be found in Use of Population Data in the Delimitation Process.)

Because children and non-citizens are counted in the U.S. census, the population base for redistricting in the United States includes many persons not eligible to vote. In the United Kingdom and many other countries, especially those with compulsory voter registration, the redistricting base is the number of qualified voters rather than the total population. The use of qualified voters as the redistricting base automatically eliminates children and non-citizens from the calculation. Whether this practice makes sense from the perspective of democratic representative theory is arguable.

The use of census population data as the redistricting base presents some problems, such as:

 

 

  • census data may be inaccurate;
  • census data can become obsolete;
  • population equality does not necessarily produce an equal electorate.
  •  

     

Inaccurate Census Data

The U.S. Census Bureau estimates that in 1990 it undercounted the population by 1.6 percent. If certain racial or ethnic groups are undercounted at a higher rate than others groups, both the allocation of seats in a legislative body and the redrawing of legislative districts can be affected. Blacks in the United States, for example, were undercounted in 1990 at a much higher rate (approximately 5 percent) than the population in general. If a state is disproportionately undercounted because it has a large black minority population, the state could conceivably lose a congressional seat when the Congress is reapportioned.

In 2011, Canada adopted the use of Census population estimates for the initial assignment of seats to a province. The Minister responsible for introducing the legislation stated : “In determining the actual populations of the provinces, we’re using the population estimates.  These are the same numbers, the same population estimates, that are used for the federal-provincial equalization program, the same numbers that are used for the Canada health transfer, the same numbers used for the Canada social transfer.  So this is the best data available for the population figures for the provinces themselves.” The estimates correct for net undercoverage in the census and provide a more accurate representation of total provincial population.


Obsolete Census Data

Even if it is accurate at the time of enumeration, census data can rapidly become obsolete because of growth rates. The use of census data for redistricting may discriminate against localities experiencing high population growth, such as, for example, suburban areas in the United States. Over time, people in these suburban districts will become underrepresented compared to those in rural or urban districts.

Census Data May Produce Unequal Electorates

Even if all districts are drawn such that they have the exact same population, the number of electors per district would not necessarily be equal because the proportion of nonvoters—that is, noncitizens, those too young to vote, and even registered voters who choose not to cast a ballot – is not uniform across the population. In the United States, for example, a congressional district with a large Hispanic population is likely to have more children and more non-citizens than a district with few Hispanics. For this reason, the number of voters on Election Day is likely to be lower in the heavily Hispanic district than in other districts.

The Use of Voter Registration Data

These same problems affect the use of voter registration data as the redistricting base, although not to the same degree. Unless the registration system is very accurate, the registration count could be incorrect. For example, the registration rolls could be inflated due to failure to remove from the list deceased persons or persons who had moved from the district; alternatively, the rolls could be missing some potential voters if the lists are not updated often enough. And unless voting is compulsory, some registered voters may neglect to cast ballots on Election Day, making the number of voters unequal across districts. After some period of time, population shifts will render districts of unequal size, regardless of whether population or registration counts are used.

Australia has adopted a unique solution to the problem of population shifts over time. Federal electoral districts are redrawn in Australia using registration, or enrolment, figures--but not current enrolment figures. Instead, districts are redrawn based on projections of what the enrolment of each locality is likely to be in three and a half years, the midpoint of the redistribution period. (See the case study, "Federal Redistribution in Australia", for more information on the use of population projections.)

Conclusion

The choice of whether to use census data or voter registration data may be guided by either practical or theoretical concerns. For instance, reliable registration data may simply not be available or may not be adequate for redistricting purposes. In the United States, for example, racial data is needed to ensure that minority communities are not divided between districts, and racial data are usually not available from the registration rolls. On the other hand, registration rolls may provide the better measure of the current population when a general enumeration of the population is unavailable, outdated, or inaccurate.

Redistricting based on registration data is likely to produce districts that are more equal with regard to the number of voters contained in them. But are voters the only persons deserving of representation? From the point of view of democratic theory, an argument could be made that all persons, and not simply voters, should be represented. If a broader definition of "representation" is adopted--one that views representatives as acting on behalf of all his or her constituents, non-voters as well as voters--then the justification for the use of total population is that it produces equal constituent representation in a more expansive sense[1].

Notes:

[1] This discussion draws heavily from an article written by Carlton Henry entitled "The Impact of New Technology and New Census Data on Redistricting in the 1990s," in Redistricting in the 1990s: A Guide for Minority Groups, ed. William O'Hare (Population Reference Bureau, Inc., 1989)

 

Use of Political Data in the Delimitation Process

Because population data does not provide information about the political composition of a proposed district, political data is sometimes added to a redistricting database. Political data may include statistics on the partisan affiliation of electors (if available) and tabulations of votes for candidates and ballot measures in prior elections. When political information is included in a redistricting database, it may be possible to predict how support for a particular candidate or political party might be affected by a change in the geographic composition of a district.

The major consumers of political information in the redistricting process are political parties and incumbent legislators. Political parties are interested in determining how changes in district boundaries are likely to affect the election of their candidates and, ultimately, the partisan composition of the legislature. Incumbent legislators, of course, are interested in ensuring that they will continue to be re-elected by the voters in their new districts.

In the United States, the courts are also interested in political information, but usually for very different reasons. Under the Voting Rights Act of 1965, certain minority groups are protected from redistricting plans that might dilute the effect of their vote. The courts use election results from prior contests in minority vote dilution claims to determine if there is (1) minority group cohesion in support of certain candidates and (2) bloc voting on the part of white voters against the minority group's preferred candidates. If these two conditions are met, and if the minority group is sufficiently large and geographically compact, then boundaries must be redrawn to create a district in which the minority community has the opportunity to elect a representative of its choice. (For more information on minority vote dilution cases in the United States, see Role of the Courts in Electoral District Delimitation.)

Problems with Using Political Data for Redistricting

Adding political data to a redistricting database may be problematic since election geography may not correspond with census geography. In the United States, for example, where census geography is usually used to redraw districts, the Census Bureau requires that census geographic units--census blocks and tracts--follow easily recognised features, such as roads and rivers. The boundaries of election precincts, however, have historically been based on property lines and therefore do not necessarily coincide with the boundaries of census geographic units.

If political data is added to a redistricting database in the United States, the data must somehow be matched to its corresponding census geography. This process can be difficult and time-consuming, as well as fraught with inaccuracies. Countries that redistrict on the basis of voter registration figures do not face this problem since registration figures and political data are usually reported at the same geographic level.

Advantages and Disadvantages of Political Data

The primary disadvantage associated with the use of political data is that it enables redistricters to create a districting plan that may favour one political party at the expense of others. In countries that have neutral boundary commissions, political information is often purposely excluded from redistricting databases to prevent political gerrymandering. Of course, political parties with sufficient resources may collect their own political data to learn in advance about the potential partisan implications of a redistricting plan.

The advantage of including political information in a redistricting database is that the partisan implications of a new redistricting plan may be determined prior to its implementation. And the ability to make such a determination is not limited solely to political parties or special interest groups with sufficient funds to carry out the necessary analysis. Since any changes in boundary lines are likely to have partisan consequences, whether intentional or not, it may be better to know in advance what these effects will be.

Of course, the partisan impact of a redistricting plan is not necessarily predictable, especially if voters are not strongly partisan. It is an inexact science in any case, and some rather infamous gerrymandering has backfired on their creators.

Use of Maps in the Delimitation Process

Maps are essential in the redistricting process, but obtaining adequate maps can be one of the most challenging obstacles facing officials in charge of redistricting. While standard maps, such as road maps, identify geographic features and the boundaries of administrative areas such as counties, cities and towns, they may not necessarily provide sufficient detail for drawing electoral districts.

If redistricting is based on a census count of the population, then the maps must show the boundaries of the census geographic units for which population statistics are available. If redistricting is based on the number of registered voters, then the maps must show the boundaries of the voting areas for which voter registration statistics are available. The boundaries of the existing electoral districts are usually needed as well.

Considerable time and effort may be needed to assemble the necessary maps and to determine the precise boundaries of the new district lines on these maps. Sufficient detail is needed to enable election administrators to assign each and every voter to an electoral district. This can be a very difficult and time-consuming process, even for jurisdictions with technologically sophisticated computer mapping systems.

In the United States, maps provided by the Census Bureau were once drawn by hand and, as a result, were often difficult to read and frequently filled with errors. In preparation for the 1990 census, however, the Census Bureau created a computerised database of the entire country called TIGER (Topographically Integrated Geographic Encoding and Referencing). The TIGER database depicts:

  • visible geographic features such as roads, rivers, and railroads;
  • a whole hierarchy of census geographic units (for example, census blocks, block groups, and census tracts) for the collection and reporting of population data;
  • the boundaries of administrative subdivisions such as counties, cities and towns; and
  • political geography such as congressional districts and, in some cases, voting areas, referred to as election precincts in the United States
  •   

Maps produced from the TIGER database for use with GIS (geographic information systems) software provide redistricters in the United States with uniform, digitised maps on a county-by-county basis for the entire country. The availability of these maps has made the process of district building in the United States more accurate and less time consuming.

Very few countries, however, have computerised maps available for redistricting purposes – most redistricters have access only to paper maps. Depending on the country, there can be serious problems with accuracy, legibility and differing scales with these paper maps. Digitised computer maps of the country can be created, but only at an enormous cost. Of course, these detailed computer maps can be used for many purposes other than redistricting and, therefore, may be worth the time, effort and expense. 

Steps in the Electoral District Delimitation Process

There are usually two phases in the electoral district delimitation process. The first phase is the allocation, or apportionment, of seats in the legislature to regional entities such as states or provinces. This is usually a very mechanical process, with the number of seats assigned to each state or province usually dependent on the relative population of that state or province. In countries that do not delimit single-member districts, the reallocation of seats, or reapportionment, is the only step taken to equalize population across electoral districts.

In countries that do redistrict, the second phase of the process is usually the adjustment of the boundaries of current districts and/or the creation of new districts within the states or provinces themselves. In countries that do not allocate seats regionally, this is the only phase in the process.

Drawing electoral district boundaries is much less mechanical and much more time consuming than allocating seats. It proceeds in three stages:

  • preparation of a redistricting database
  • assignment of geographical units to districts until all units have been assigned and the boundaries of all districts established
  • production of a summary description and maps for evaluating and implementing the redistricting plan

Allocating Seats

Many countries, especially those very large in size and/or having a federal system rather than a unitary system, delegate the responsibility for drawing national legislative districts to regional entities such as states or provinces. For example, in the United States, seats in Congress are allocated to the states on the basis of each state's relative population. Each state is then responsible for drawing the prescribed number of congressional districts within its borders. Australia and Canada use a similar system of allocating seats in the national parliament to the states or provinces and then having independent electoral commissions in each state or province draw district boundaries.

In the United States, the decennial process of allocating congressional seats among the states is called "reapportionment." In Canada, the decennial reallocation of Parliamentary seats among the provinces is known as "redistribution." But in many countries, no special terminology is applied to this allocation process. This is especially true of countries that do not have federal systems or countries where regional entities have little power over redistricting decisions. Population (either census enumeration data or voter registration figures) is almost always used as the basis for apportioning seats to national legislatures. In fact, the stated purpose of decennial census in the United States is to determine the proper allocation of congressional seats among the states. The U.S. Constitution, in Article I, Section 2, provides that:

“Representatives shall be apportioned among the several states...according to their numbers....The actual Enumeration shall be made within three years after the first meeting of the Congress and within every subsequent term of ten years in such manner as they shall by law direct.”

Before seats are allocated, however, the total number of seats in the legislative body is often determined. But this is not always the approach adopted. In its first 120 years, for example, the U.S. House of Representatives simply grew in size to accommodate new states and a rapidly growing population. From 1790 to 1900, seats were allocated to states on the basis of a population quota that, despite its gradual increase over time, resulted in the House of Representatives expanding in size from 106 to 391 during this time period. Since 1910, however, a permanent ceiling on membership in the House of Representatives has been set at 435 legislators. Seats were allocated by first giving each state one seat and then awarding the remaining seats in succession to the states with the largest remaining quota. In 1950, the statute defining the formula was modified slightly. Each state was given one seat, and the remaining 385 seats were allocated in succession under a priority numbers formula. This formula is referred to as the "method of equal proportions.”

Although the process is a mechanical one, the formula for the allocation of seats has not escaped controversy. Since 1790 states in the United States have argued over how congressional seats are apportioned, and states that have lost representation have been especially sensitive about the process.

Some countries have adopted solutions designed to alleviate this problem. Canada, for example, has solved this problem by never allowing the number of seats allocated to a province in the House of Commons to decrease. Canada, like the United States, requires that the distribution of seats in the House of Commons be governed by the "proportionate Representation of the Provinces," based on a decennial census. However, a "grandfather clause" was adopted to protect provinces with declining populations relative to the other provinces. This clause, first enacted in 1976 and then re-enacted in different legislative form nine years later, ensures that no province can ever have fewer electoral districts than it had in 1976. The size of the House of Commons has gradually increased as a result. (For more information on the redistribution process in Canada, see the case study Representation in the Canadian Parliament.)

There are two problems with allowing states or provinces to retain more seats than population alone would dictate: 

  • the size of the legislature could become unwieldy;
  • the population or electoral quota (the population of the state divided by the number of seats allocated to the state) can vary considerably across states or provinces.
  •   

The latter occurs when a state or province is awarded a greater number of seats than it would normally be entitled, based on its relative proportion of the population. In Canada, these are referred to as "add-on" seats.

But a broad range in population quotas is possible even without add-on seats. In the most recent round of congressional redistricting in the United States in 1991, the population quota for congressional districts ranged from 455,975 in Wyoming to 803,655 in Montana. This broad range occurred because each state is awarded at least one seat, even if its population is below the "national quota" for a seat. No national electoral quota is actually applied in the United States; quotas are calculated only on a state by state basis. The population of the state of Wyoming was below the "national quota" for a congressional district, but was awarded a seat; whereas the population of Montana, more than ample for one congressional seat, was not quite large enough to be allocated two seats.

Drawing District Boundaries

Drawing electoral district boundaries is the second phase of the redistricting process. The drawing of federal electoral districts can either be done at the federal level or at the state or provincial level, once a specified number of districts have been assigned to the state or province. Drawing district lines proceeds in three stages:

  • preparation of a redistricting database
  • assignment of geographic units to districts, forming district boundaries
  • production of a summary description and maps of the redistricting plan

Preparation of a Database

Before line drawing can commence, a redistricting database must be created. The database is composed of all of the geographic units that are to be used as the building blocks, and any population data, as well as any additional demographic and political data associated with these geographic units and deemed useful for redistricting. The geographic units may be as large as counties, cities or towns, or as small as voting areas or city blocks.

Formation of District Boundaries

After the database has been created, district boundaries can be formed. This involves adjusting the boundaries of existing districts and/or creating new districts. The process entails assigning, or reassigning, geographic units to districts until all units have been assigned to the requisite number of districts.

Production of a Summary Description and Maps

After the districts have been formed, a summary description of the redistricting plan must be produced. This description usually consists of a list of the geographic components of each district, a map of the district plan, and a statistical summary of the districts using selected population and possibly political data. This information may be used to evaluate the redistricting plan on a variety of levels, such as equal population, and is needed to implement the plan for election purposes.

Preparing a Districting Database

The first step in the line drawing process is preparing a database. In a redistricting database, each data item reflects a single geographic unit. The geographic units, to be used as the basic building blocks for creating the districts, can reflect administrative areas such as counties or cities and towns, census geography such as census blocks, or election geography such as voting areas. The database should include as much of the following information as possible for each geographic unit:

  • a unique name or identification number for each unit
  • additional identifiers, such as town or city, county, and state or province for each unit
  • the total population or the number of registered voters for each unit
  •   

Additional demographic data, such as total population and voting age population counts by race or ethnicity for each unit, may also be included if useful for redistricting.

The Use of Digitised Geographic Data

If geographic information systems (GIS) software is to be used to develop district plans, the redistricting database must also include information related to the location of the geographic unit (spatial co-ordinates). This information allows the computer to automatically generate computer screen maps that reflect each new assignment of a geographic unit to a district.

The Use of Political Data

If the redistricting database is to include political information, then voter registration counts by political party, if available, and votes for candidates from previous elections should be entered for each geographic unit. The number of previous elections to be included in the database depends on the availability of previous election results, the amount of time required to enter the results into the database, and the financial resources available for creating the database.

If a redistricting database is to include both political and census demographic data, the election geography and the population geography must correspond with one another. If they do not, they will have to be matched.

Matching Census and Election Geography

Matching census geography and election geography can be quite challenging. The easiest method for matching these units is to merge the smaller geographic units until they approximate the boundaries of the larger unit. If, for example, census data is available for units as small as city blocks, then these blocks can simply be aggregated together until they match an election precinct (voting area). On the other hand, if the units of election geography are smaller, then they can be merged to produce political data for the larger pieces of census geography (e.g., villages and towns). It is always easier to aggregate data up to the higher level than it is to try and break it down to smaller levels of geography.

In the United States, however, the courts often require states to draw districts with census geography rather than election precincts (voting areas). This requirement can be handled in one of two ways:

  • election precincts and their corresponding political data can be merged until they approximate census tracts;
  • alternatively, precinct level election results can be disaggregated down to the census block level.
  •   

If redistricters want the flexibility of being able to draw at the smaller census block level (and many states do, in fact, prefer this approach), then political data reported at the precinct level have to be disaggregated and distributed among the census blocks that fall within that election precinct. This process is difficult, involves a great deal of estimating, and is subject to a great deal of error. Because it is a very tedious and labour intensive task, it will absorb a substantial amount of the database construction time.

Drawing Electoral District Boundaries

Once a database has been prepared, the next step in the redistricting process is the formation of districts. This is the step in which the line drawers create a redistricting plan either by assigning geographic units to districts for the first time or by moving geographic units in an existing plan from one district to another. A redistricting plan is complete when all geographic units in a given territory are assigned to a district and all districts in the plan meet the predetermined redistricting criteria. The geographic units to be assigned may be as large as counties or cities and towns, or they may be smaller units of geography such as census blocks or voting areas.

If the line drawers are to create a new redistricting plan by modifying districts in an already existing plan, their tasks may be relatively straightforward. This is especially true if a decision has been made that districts should be modified only as much as necessary to meet equal population standards. Many countries, for instance, list consideration of existing district boundaries as a redistricting criterion. If there is no existing redistricting plan to modify or if the number of districts in an existing plan is to change substantially, the line drawers' tasks become more challenging.

Modifying Districts in an Existing Plan

One strategy that could be adopted if districts already exist and simply need to be modified is to (1) determine the current population of each of the existing districts, (2) calculate the minimum amount of change required to meet equal population standards, and (3) focus only on the districts that need to add or subtract population. Of course, it is never as simple as adding or subtracting population only from the malapportioned districts. Since the process also affects adjoining districts, a ripple effect usually occurs across the districts, necessitating at least some changes to other districts. But an effort could be made to minimise these changes.

This incremental approach to redistricting is often adopted by line drawers who are required to consider existing district boundaries when redistricting. It may also be a good strategy if the legislative body for which new districts are being drawn must approve the redistricting plan before it can be implemented. The chances of obtaining legislative approval for a redistricting plan are obviously affected by the extent to which existing districts are disrupted. In the United States, for example, the wisest strategy is often to move only the minimum amount of geography necessary to achieve population equality – this ensures bipartisan legislative support for the redistricting plan.

Creating a New Plan or Substantially Modifying a Plan

Drawing districts where districts have not previously existed or where district boundaries must be changed substantially is a more complex process. The task of drawing districts for the first time can be simplified by dividing a large territory into manageable sub-regions and drawing districts in two or three stages. If a country is divided into states or provinces, for example, allocating districts first to the states or provinces and then drawing districts within these states or provinces will make the process more manageable. A second level of district allocation may also be desirable if the states or provinces are large. If further allocations are made, however, the subdivisions should possess some sense of geographic unity.

Choosing the Basic "Building Block"

One of the first decisions to make in creating districts is what unit of geography and, possibly, what level of geography to use as the basic building block. This decision is required whether the line drawers are to modify an existing districting plan or create an entirely new plan. Whether composed of census or election geography, the redistricting database has, presumably, been built upon a unit of geography small enough to give the line drawers some flexibility in creating districts.

The line drawers, however, need not use the smallest unit of geography available in the redistricting database as the only building block when creating or modifying a plan. For example, redistricters may be able to assign whole counties or cities in many areas and resort only to individual census blocks or voting areas in certain, usually urban, areas. And, in fact, it is easier and faster to create the broad outlines of districts with large geographic units and then use the smaller units to make the refinements necessary to comply with equal population standards or other redistricting criteria.

Choosing Where to Begin Drawing Districts

Another decision that should be made early in the process of creating districts is where to begin drawing district lines. Redistricters often begin at the corners of a territory and work towards the centre. This is done to avoid any distortion which may result at a corner or edge if the process starts at one side of the territory and moves to a final corner or edge.

There may be reasons to begin the line drawing elsewhere. If, for example, the redistricters wish to create a district for a geographically concentrated minority community or some other community of interest, they may begin by drawing the boundaries of that particular district and then move the line drawing process outward from that district.

Drawing District Boundaries Interactively

Once decisions have been made about what units of geography to use as building blocks and where to begin the line drawing process, the actual line drawing can commence. The line drawing process is an interactive one. The redistricters assign a piece of geography and then determine how the inclusion of that particular geographic unit affects the size and composition of the district.

Although aggregating the population totals for the districts after each new assignment is a straightforward process, it can be a tedious one as well. And as geographic units continue to be shifted, this process is repeated over and over again. A computer can automate this process, but it can also be done manually with a hand-held calculator, if necessary.

Describing and Evaluating an Electoral District Plan

Once the boundary authority has successfully completed a redistricting plan by assigning all geographic units in the territory to a district, summary information for the plan should be produced. This information may be used to evaluate the plan. Also, it is often needed to prepare legislation in order to enact a plan. Certainly, it will be needed to implement a plan once the final plan has been approved. A summary description of a redistricting plan should include: 

  • a description of the plan listing the geographic components of each district
  • map(s) of the plan showing the district boundaries
  • a report summarising the most relevant statistical information for each district in the plan

Description of the Plan

A description of the redistricting plan should include a detailed list of all of the geographic units contained in each district. A narrative description of the plan could also be created. In the United States, for example, a detailed street-by-street description of district boundaries, called "metes and bounds," is often required, especially in urban areas.

Advances in Geographic Information Systems (GIS) now allow for the production of highly detailed maps.  The British Columbia (Canada) Electoral Boundaries Commission in its redistricting of provincial electoral boundaries ”chose to forego metes and bounds and submitted their proposed boundaries digitally. A DVD packaged with the commission’s final report contained their proposed electoral districts saved as shape files, a data format common to GIS software. The DVD included software allowing users to view the electoral district boundaries in the context of geographic features and local administrative boundaries. Together with the printed maps available in their reports, the commission felt satisfied that “the traditional metes and bounds descriptions of proposed electoral boundaries [had become] unnecessary.” In establishing the new electoral districts in the Electoral Districts Act, the Legislative Assembly referenced the digital files rather than including metes and bounds descriptions in the Schedule to the Act. By doing so, the Legislative Assembly established the digital files as the legal source of the new boundaries.” (Report of the Chief Electoral Officer on the 2008 Electoral Boundary Redistribution)

Maps of the Plan

Maps of the redistricting plan may need to be created at different scales. A large-scale map will be needed to show the entire redistricting plan while a series of smaller-scale maps may be needed to show district boundaries in more detail in areas with higher population densities. These maps may be used to inform legislators and voters of the new district boundaries. In addition, these maps will be needed by election administrators to implement the new plan.

Summary Report of the Plan

Summary reports of the redistricting plan can be produced displaying district population totals and any other statistics for data items that have been included in the redistricting database. These reports may include detailed demographic data as well as political data, if desired. The design of the report should reflect its intended purpose. Of course, more than one report can be compiled for each plan. The only constraint on reporting capabilities is that the information must have been included in the database. For this reason, it is important to construct a database that includes as much of the necessary information as possible.

Summary reports allow redistricters to evaluate a redistricting plan according to established criteria. For example, the redistricting criteria may specify that districts be as equal in population as possible, and that some districts afford minority voters the opportunity to elect candidates of choice. In this instance, the summary report of the redistricting plan should list the total population, the population deviation from the electoral quota, and the racial and ethnic composition of each district.

Other interested groups may request summary reports for different sets of data items. Political parties and legislators, for example, may be interested in a detailed political profile of the proposed districts. A political report could include such information as the electoral history of the district. This can be calculated by summing all the previous election results for each of the geographic units included in the new district.

It is a relatively straightforward procedure to construct reports that calculate summary statistics for data items such as total population, population by race, and votes for candidates in previous political contests, provided that this information has been included in the redistricting database. But there may, or may not, be constitutional, legal, or court-specified standards that specify, for example, what constitutes an acceptable level of population deviation. If no set standards exist, then it is still a matter of discretion as to what the summary statistics tell us.

Delimitation Technology

Delimitation, or redistricting, can be accomplished using manual techniques such as colour markers, road maps, and calculators, or by using sophisticated computers and geographic information systems (GIS) software. The technology employed may change how the redistricting process is conducted, but it does not change the task of the redistricters in drawing district boundaries. Computers and computer software can, however, add a great deal of speed, accuracy and efficiency to the redistricting process.

Computers can reduce the time it takes to draw a redistricting plan, provide accurate statistical reports to help evaluate a redistricting plan, and produce detailed colour maps of the new boundaries. With computers, report and map production can be as simple as issuing a single computer command. This increase in speed, efficiency, and accuracy has changed the redistricting process to the extent that it allows line drawers to consider a wider range of district plan options. It also permits interested parties, both inside and outside of the process, to evaluate the redistricting plans more easily and more thoroughly.

Computers and computer software can be very expensive, however, and it may not be worth the investment. This is particularly true in countries with small populations and/or few districts to draw. Computer-assisted redistricting also may not be worth the investment in countries where maps have not and cannot be digitised.

In most situations, inexpensive personal computers and simple computer programs, or even adding machines or hand-held calculators, can be used with paper maps to assign geographic units to districts and to keep track of demographic and political data for the new districts. Until very recently, this was the method employed by all line drawers and is still the method used by the vast majority of redistricters outside of Europe, North America and Australia.

Use of Computers and Software for Delimitation

 

Computers can be extremely useful in the delimitation process[1]. They can reduce the time required to draw redistricting plans and also aid in the evaluation of plans. They can produce detailed maps and statistics based on the demographic and political information in the redistricting database. And because computers allow redistricters to create plans more quickly, they enable redistricters to generate a larger number of redistricting plan alternatives for comparison and evaluation.

In the United States, about half the states used computers when congressional districts were redrawn after the 1980 census. In most of these states, computers were used as "super calculators" to summarise demographic and political data and to produce printouts of demographic information for the proposed districts.Today, all states use GIS systems.

Two important developments in the late 1980s contributed to a far greater use of computers for upcoming redistricting. These were (1) the development and the extensive use of powerful, affordable desktop computers and (2) the development of geographic information systems (GIS) software.

By 1990, every state in the United States used computers for redistricting, and almost every state used GIS software. The move towards computer-assisted redistricting was also aided by the availability of detailed demographic and geographic data in computer readable format from the U.S. Census Bureau.

 

Computer Software

Redistricters have several different options from which to choose with regard to software. A simple spreadsheet or database program can calculate the statistics that are commonly used in redistricting, such as total population, racial composition or overall population deviation from the electoral quota.

The resources necessary to employ computer technology for redistricting include:

  • a computer (preferably desktop)
  • spreadsheet or database software, or GIS software
  • population and, if desired, political data available in electronic form (if only paper copies of the information are available, the data will have to be keypunched into the computer program)
  • maps (either paper maps or digitised maps if GIS software is to be used)

 

A more sophisticated and more expensive option is GIS software.

 

Notes:

[1] This discussion draws heavily from an article written by Carlton Henry entitled "The Impact of New Technology and New Census Data on Redistricting in the 1990s." This article appeared in Redistricting in the 1990s: A Guide for Minority Groups, edited by William O'Hare and published by the Population Reference Bureau, Inc. in 1989.

 

Low-Technology Delimitation

In large countries, delimitation, or redistricting, can be a major technical challenge. Manipulating the enormous amount of geography and population data required to create new districts is a very complex procedure. However, redistricting in countries with small populations and/or few districts presents fewer technological challenges.

Computers and computer software may not be worth the investment in these countries.

Although computers and geographic information systems (GIS) software can speed up the redistricting process and make the process more accurate and efficient, redistricting with computers can be very expensive. This is particularly true if the maps to be used for computer-assisted redistricting have not already been digitised.

Regardless of the technology to be employed, it is necessary to obtain accurate summaries of demographic data for current districts and to re-tabulate that data as geographic units are reassigned among the districts. Simple computer programs, or even adding machines or hand-held calculators, can be used to summarise demographic data at the district level after each new geographic assignment.

Maps must be utilised during the redistricting process to ensure that the geographic units being assigned to a district are contiguous with the district. If redistricting is to be done without GIS software, district boundaries must be drawn by hand on paper maps. Drawing the boundaries of proposed districts onto transparent acetate overlays will preserve the original paper map and can facilitate the process of manually reassigning population units among districts. Once a final redistricting plan has been decided on, maps of the new district boundaries must be produced by a separate process.

Recommended resources for low-technology redistricting include:

  • a computer, adding machine or hand-held calculator
  • census and/or political data on paper
  • paper maps

GIS (Geographical Information Systems)

Geographical Information Systems (GIS) store spatial and attribution data

For redistricting purposes, the spatial data is location-based information; the attribute data is data that describes a particular location. An example of spatial data would be the location of a house (spatially known as a point) or occupied dwelling while examples of attribute data could be the address of the house, the street name on which it is located, the electoral district in which it is found and the number of people living in the house to mention only a few.  The ability to relate spatial information to attribute data is what makes GIS a powerful tool for redistricting purposes.

While there are a number of software packages specifically marketed as redistricting software, whichever package is chosen should contain the following functions to be considered optimal:

  • Create/modify/delete new boundaries
  • Assign census polygons (counting areas) to new boundaries using a wide variety of methods
  • Display running population counts and demographics as census blocks (the smallest population count aggregation area) are assigned to new boundaries
  • Display and print maps and reports
  • Query and select data
  • Validate new boundaries
  • Overlay with satellite imagery
  • Allow block (counting areas) splitting
  • Multi-lingual (tools and reports/maps for countries with more than one ”official” or ”national” language)
  • Access to other Geographic data via the Web

 

Redistricters using GIS software have a variety of sophisticated tools at their disposal for creating district plans. GIS software can be used to draw new districts by assigning geographic units to districts or by moving particular geographic units out of one district and into another. The effects of the reassignment on the population, or any other demographic or political characteristics of the districts, can be immediately reflected on the computer screen as shown in the following screen capture:

In the above, each boundary line can be dragged to a new location. Each polygon (enclosed area) can be added to the one touching it.  The corresponding demographic variables, shown in the table under the map, update automatically.
Once a new plan has been created, the GIS software can produce a map and report of the new district plan or any individual district. GIS software essentially consolidates all redistricting operations into one efficient system.

Limitations of GIS

 

While GIS is a great improvement over manual redistricting, GIS software does not provide a push button (”push a button and it’s done”) solution to boundary delimitation. The basic activity of redistricting, aggregating census blocks into a new or existing electoral district to approach population equality in the redistricting plan (and mapping the result) is only one requirement in redistricting, albeit the most important.  Others factors that may need to be considered include legislation such as the Voting Rights Act in the United States and the Official Languages Act in Canada, minority representation, linguistic representation, racial representation, geographical size (geographically large and/or compact districts), transportation routes, administrative boundaries, community of interest (or identity), historical pattern of a district and contiguity. It would be very difficult to weight the preceding factors to the effect that they could be used efficiently within a GIS system, hence, a great deal of human interaction with the redistricting scenarios generated is required.

Data preparation is also important. Differences in census geography and electoral geography need to be resolved.  In most countries, the two geographies are different, that is, census areas do not exactly match electoral areas.  A method for reconciling the differences must be established to use the software effectively.  Canada has advanced considerably in this area.  In the mid 1990s, Elections Canada (EC), the electoral agency, and Statistics Canada (STC), responsible for conducting the census, entertained the idea of building a common database, the National Geographic Database (NGD) that would enable both organizations to conduct their operational business.  A diagrammatic representation of that geographical partnership appears below.

 

Built on a common database of roads and addresses and other topographical features such as hydrology, railroads, power lines (all used in the descriptions of electoral districts), each organization overlays their business layers onto the base (electoral districts, polling districts for Elections Canada, census blocks, enumeration and other collection units for Statistics Canada).  Because the basic geography is the same, each organization can use the other’s business layer. (Elections Canada uses the census information for redistricting; Statistics Canada uses the electoral district for reporting the results of the census.)

Complexity of the software must be considered.  The most optimal software packages also require a knowledgeable user. To get a boundary delimitation body operating in as little time as possible, it may be necessary to hire consultants or GIS technicians to assist.  Alternatively, the electoral management body, if it is required to assist the boundary authority, may have GIS staff available for that purpose.

Summary

GIS shortens the time required to draw boundary lines and aggregate/display the demographic data related to the map.  While GIS can be costly, it may be a cost worth bearing depending on the frequency of redistricting and on the size of the area to be redistricted.  The instantaneous feedback and ease of creating multiple scenarios for consideration improves the redistricting process.

An emerging benefit in the use of GIS in redistricting is the capacity to engage the public in the redistricting process. Digitized redistricting data is now accessible though the use of most common web browsers; no special software is required. The public can see the proposed new districts, the current districts, chose to overlay the new district on the current one and see where the differences lie and provide feedback to the boundary authorities. GIS systems can allow the public to ”redline” or mark up an online map and transmit that map, along with comments, to the boundary authority.

Elections Canada noted in an environmental scan (2009) of redistricting software that ”[t]here is, nonetheless, a widespread belief that public participation in redistribution could be increased and improved. All of those involved in the process, from commissioners, to members of Parliament, to the public, have expressed the wish to see the current process for public consultation expanded and updated to reflect the improved communication technology of today.”

In increasing numbers, the public is becoming more spatially aware, nowhere more evident than in the use of Google maps. Where once a novelty, web users today include location mapping and map annotation in their online presence. This provides a natural communication bridge from the public to the boundary authority and back again, all made possible through the use of GIS.

Use of Consultants

Redistricting consultants offer a wide range of services to countries contemplating or undertaking the process of drawing district boundaries. They can serve in a broad range of capacities, from providing limited advice on a particular facet of redistricting to managing the entire redistricting process.

Some redistricting consultants may be of particular use in computer-assisted redistricting. They can provide custom software, for example, or advise on the selection of software. They can assist with the acquisition of hardware and in installing redistricting computer systems. And they can provide training in system operations.

Some of the more general services a redistricting consultant may provide include:

  • help in designing a formal redistricting process, including suggesting electoral law relating to the delimitation process;
  • help in the construction of a redistricting database;
  • strategic advice on creating districts;
  • help in evaluating a redistricting plan;
  • litigation support in a court challenge to a redistricting plan.

Depending on what a consultant is being asked to do, the consultant should be conversant with comparative methods of redistricting and the laws that govern redistricting, including defining and measuring redistricting criteria, and with administering and managing the redistricting process. If the consultant is being asked to help with the line drawing process itself, the consultant should have experience with:

  • the use of maps and population data;
  • the construction of databases (including the merging of political and census data, if applicable);
  • the creation of districts;
  • the evaluation of redistricting plans.

If a number of broad tasks are required, a consulting firm or a group of consultants may be of greater service than an individual consultant.

Redistricting consultants are often trained as geographers or political scientists. Whatever their training, they should be knowledgeable about the election process and electoral law, geography, political methodology and statistics as well as the redistricting process in particular. Computer-assisted redistricting may require consultants with computer-programming and other technical skills.

The costs of hiring a consultant are as wide ranging as the services offered by consultants. Using a consultant in an advisory capacity is clearly much less expensive than bringing in a consultant to oversee or actually create a database or draw districts.

Supporting the Redistricting Process

In many countries, redistricting is conducted by special bodies that are set-up particularly for a specific redistricting exercise, and, at the end of their mandate, are disbanded; some countries might have permanent redistricting bodies.  In either case, the permanent electoral management body might be asked to support and lend assistance to the appropriate redistricting body or even to conduct the redistricting exercise itself. Planning for the redistricting exercise should include three broad phases:

  1. Planning and preparation before the redistricting exercise begins
  2. Support to the redistricting process once it is underway, and
  3. Returning to election readiness

In Canada, 10 three-member commissions, one for each province, are established to conduct the redistricting exercise (popularly known as redistribution). Because these are temporary bodies, federally constituted at the beginning of each redistribution exercise once every ten years following the decennial census, the members have little or no experience working within a government structure. To help them manage the administration side of the exercise, the Office of the Chief Electoral Officer of Canada (Elections Canada) provides legal, technical, financial and administrative support to the independent commissions responsible for the periodic process of readjusting federal electoral boundaries, to ensure that representation conforms to the Electoral Boundaries Readjustment Act.

The following tables lists the major activities undertaken by Elections Canada before, during and after the redistribution process

PHASE 1: Planning and Preparation

Up to two years before the independent commissions are established to conduct the redistribution, Elections Canada sectors/divisions begin the planning of Redistribution activities and complete the development of tools and the implementation of infrastructure to support Phase 2 (the independent commissions) of the program. Major activities by sector/division lead follow:

 

Division/Sector

Activities

Redistribution Directorate (this directorate is re-established every ten years to plan for the 3 broad phases of the redistribution exercise.

Staffing of key positions (within Elections Canada)

 

Electoral Geography

Development and implementation of Redistribution tools:

Commission Redistricting Tool (CRT – a GIS application developed by Elections Canada for boundary delimitation)

Data and Mapping Tools (DMT - applications developed to manipulate data, print reports and maps)

Public Web Tools (PWT - offers the public web mapping capabilities to respond to the commission’s proposals in preparation for the hearings.

 

Prepare for the following:

Provide geography specialists to support Commissions in the field

Providing geography specialists to assist committee of the House of Commons in its review of objections from Members of Parliament

Production of all geography products requested by legislation (gazetted proposals, report, newspaper and Web map)

Information Technology (IT)

Support for GIS Technology

Support for Redistribution website development

Support for Field application deployment

Policy, Planning and Public Affairs (PPPA)

Development of a global communication strategy including Internet strategy

Development of communication plans for each Commission

Design and construction of Redistribution website

Provision of translation, editing, and graphic design services

Human Resources

Developing job descriptions and staffing key Redistribution positions

Legal Services

Review of past recommendations for amending redistricting legislation

Impact evaluation of potential modifications of legislation and Representation formula on the Redistribution process

 

PHASE 2: Support of the redistricting process

Elections Canada supports the redistricting commissions in every stage of their operations.  Major activities and deliverables by sector/division lead follow:

Division/Sector

Activities

Redistribution Directorate

Apply representation formula

Organize public sittings

Tabling of reports with the Speaker

Completion of reports following public sittings. Reports referred back to commissions for reconsideration (as necessary)

Preparation of draft representation order

Overseeing Proclamation of draft representation order

Electoral Geography

Provide geography specialists to support Commissions in the field

Providing geography specialists to assist committee of the House of Commons

Production of all geography products requested by EBRA (gazetted proposals, report, newspaper and Web map)

Information Technology (IT)

On-going support for Redistribution activities (in the field and at Headquarters)

Planning (business requirements, design) for adapting corporate systems in managing two sets of electoral districts (in the transition period before the new boundaries become effective).

Policy, Planning and Public Affairs (PPPA)

Publication of Representation formula results in Canada Gazette

Gazette proclamation establishing commissions, naming each member

Publication of commissions’ proposals

Publication of notices for public sittings through public advertisements

Gazetting representation order and proclamation

 

Note: this phase of redistribution places the heaviest demand on the services provided by Electoral Geography and by Policy, Planning and Public Affairs.  Most of the milestones noted above are governed by legislated (EBRA) timeframes and must be completed on schedule.  The services need to be provided to each of the ten commissions, at times concurrently, at times separately but experience has shown that providing the services to the commissions sequentially is not feasible.

PHASE 3: Return to Readiness

Following proclamation of the draft representation order, there is a minimum 7-month period prescribed by the Electoral Boundaries Readjustment Act before the representation order comes into force. This period enables Elections Canada to return to a state of readiness under the new electoral boundaries. (It also enables political parties to adjust their internal operations and organizational structure to reflect the new boundaries/districts.)  Any general election called before the 7 month period has passed is conducted under the current boundaries. Any general election called after the 7 month period has passed is conducted under the new boundaries.  Any by-election called before the new boundaries come into effect is conducted under the current boundaries.

Two concurrent activities dominate return to readiness across all EC sectors:

  1. Maintain a state of readiness under the current electoral districts; and
  2. Establish and maintain a state of readiness under the new electoral districts.

Major activities and deliverables by sector/division lead follow:

Division/Sector

Activities

Electoral Geography

Maintain current Electoral District (ED) and Polling District (PD) boundaries

Revise polling divisions to respect new ED boundaries

Prepare new maps at national, provincial, municipal and ED levels and PD maps

Transpose votes of previous general election to new boundaries

National Register of Electors (NROE)

Adjust Corporate Address Register

Prepare two releases (current and new boundaries)

Field Readiness and Management (FREM)

Hiring of New Returning Officers

Training of Returning Officers, Assistant Returning Officers, Automation Co-ordinators,
 and Support Network staff

Revise Electoral Event Systems (all systems must be revised to the new boundaries)

Distribution Centre to prepare material/supplies for new districts

Note: two full sets of Returning Officers  (current and new) and all other field election staff may be required

Information Technology (IT)

Adjust Corporate Data Base (CDB) to maintain current structure and new required structure

Policy, Planning and Public Affairs (PPPA)

Review of Enquiries Unit (Manual and tools) and WEB tools and content (structure, tools, and all information)

Review of all Backgrounders, Information Kits,  Exploring Canada’s Electoral System, Advertising program and Media buying strategy

Finance

Establish accounts for new electoral districts

 

As can be seen, the actual setting of new boundaries is just a small part of the redistricting exercise. Planning for, carrying out, and implementing the new boundaries requires a major effort on behalf of the Elections Management Body (Elections Canada) and requires close coordination both within the various sections of Elections Canada and with the external, independent boundary delimitation commissions.

Special Considerations for Delimiting Voting Areas

Voting areas are administrative units that are used only for conducting elections. They are contiguous geographic areas where all voters within the circumscribed territory are assigned to the same polling place. Voting areas are known by a variety of different labels, depending on the country. In Commonwealth countries, for example, they may be referred to as polling areas, voting or election districts, or election precincts.

Voting areas are necessary for the technical implementation of an election. A given territory must be subdivided in such a way as to enable voters to travel as conveniently as possible to a polling site and cast their ballots. In addition, assigning electors to voting areas allows election administrators to keep track of who is voting. This ensures that no one casts more than one ballot.

Most countries, regardless of the type of electoral system employed, delimit voting areas. Unlike electoral districts, where the type of electoral system determines how crucial delimitation is to the outcome of an election, the delimitation of voting areas has a minimal effect on election outcomes. Voting areas are used merely to collect votes; they are not used to translate votes into seats in a legislative or parliamentary body.

Because voting areas are used for election administration only, the delimitation of voting areas is not controversial and is normally left to the discretion of election administrators. Electoral laws or regulations, however, may specify certain criteria for delimiting voting areas.

Authority for Delimiting Voting Areas

The delimitation of voting areas is often performed by local election officials, but the delimitation may be carried out by federal election administrators. For example, the election commission in Ghana is responsible for drawing all political boundaries, from constituency (electoral district) boundaries to voting area boundaries.

In some countries, the same voting areas are used for all elections. In other countries, different voting areas are created for different elections. For example, in the United States, county election officials delimit voting areas for all elections: federal, state, and local. In Canada, a federal agency, Elections Canada, draws voting areas for federal elections, while provincial election administration authorities draw voting areas for provincial and local elections.

Criteria for Delimiting Voting Areas

Election administrators usually consider the following criteria when delimiting voting areas, even if no criteria are specified by law:

  • population size
  • pre-existing administrative and electoral district boundaries
  • convenience and accessibility for voters

Although voting areas differ dramatically in size of population--within a country as well as in different countries--there is a minimum and a maximum number of voters that can be efficiently and effectively served by a single polling site. It may not be feasible to establish a polling site for only a handful of voters. On the other hand, assigning too many voters to a single polling site can result in long lines of frustrated voters waiting to cast their ballots at an election. The optimal minimum and maximum numbers vary, depending on local conditions and available resources and technology.

Administrative and electoral district boundaries should be taken into account when creating voting areas because these boundaries determine who votes for a particular set of offices and candidates at an election. If the boundaries of voting areas cross administrative or electoral district boundaries, election administration will become more complex. Different ballot styles listing different offices and candidates will be needed for voters within a single voting area. Producing and disseminating several different ballot styles within a single voting area can be complicated and expensive.

Convenience and accessibility are also important factors to consider when drawing voting areas. The boundaries of a voting area should be drawn around a polling site that is centrally located, easy to travel to, and accessible to all eligible voters assigned the polling site. Factors such as the time needed to travel to the polling site and accessibility to public transportation should also be taken into account. Some countries, for example, specify by law the maximum distance that voters can be expected to travel to cast their ballots.

The Need to Redraw Voting Areas Periodically

Voting areas may need to be redrawn because of population changes or changes to administrative or electoral boundaries. For example, it may be necessary to redraw a voting area if the area's population has grown too large for a single polling site or, alternatively, the voting area has lost population and it is no longer cost effective to keep the polling site operational. After redistricting, a voting area may need to be redrawn to realign its boundaries with the boundaries of the new electoral districts. If not redrawn, a voting area may be divided between two or more districts, complicating the administration of the election.

Many countries redraw voting areas on a regular basis, for example, after the redistricting of electoral districts or the completion of a voter registration campaign. Some countries redraw voting areas on an ad hoc basis, for example, whenever the voting areas become too large or too small.

Tasks of Delimiting Voting Areas

Two essential pieces of information for delimiting voting areas are:

  • population data
  • detailed local maps

Population data for delimiting voting areas usually consists of voter registration data. A reliable count of the number of eligible voters in the territory to be delimited is needed as well as information on the residential location of each voter. Accurate and up-to-date maps are also needed to delimit voting areas. The maps should clearly delineate local features and indicate the boundaries of administrative and electoral districts.

The first step in the process of delimiting voting areas is to obtain maps and mark relevant administrative and electoral boundaries. The next step is to generate a list of registered voters by location--by a street address, if possible. The number of voters on each side of the street, or at each location, is then counted and recorded on the map. After the voter counts have been recorded, election officials can begin to create or adjust voting area boundaries, tallying and re-tallying the counts with the assignment of each new piece of territory. Defining new voting area boundaries requires some experimentation to determine where boundaries must be moved to most closely match criteria such as population size and distance from a polling site. The process is similar to redistricting electoral boundaries, except that the territories involved are usually much smaller.

Conclusion

Voting areas are required to implement elections efficiently and effectively. Although the delimitation of voting areas is rarely a controversial process, it is an important one for the administration of elections.

Annexes

Glossary

Boundary Delimitation: (or districting) the process of drawing electoral districts or voting area boundaries.

District Magnitude: the number of legislative seats assigned to a district.

Redistricting: the periodic delimitation of electoral districts or voting areas for representation purposes.

Electoral District: geographic unit that defines legislative representation (conversion of votes) according to type of electoral system being used.

Voting Areas: (also called polling areas, districts or election precincts) administrative units where all voters within the circumscribed geographic territory are assigned to the same polling place.

Gerrymandering: manipulating the electoral boundaries so as to give undue influence to some party or class; districting or redistricting in a discriminatory manner.

Malapportionment: electoral districts that vary greatly in population

Geographic Information Systems (GIS): technology that is used to view and analyze data from a geographic perspective. This data may exist as maps, 3D virtual models, tables, and/or lists.

Community of Interest: generally thought of as a group of individuals united by shared interests or values.

Boundary Authority: specially designated commission or group that divides a country or region into districts for the election of representatives.

Reapportionment: periodic redrawing of geographic boundaries of electoral districts from which legislative representatives are elected; a primary purpose of early census enumeration

Bibliography

 

Published Resources

Barnes, G.P. "The General Review of Parliamentary Constituencies in England (1976 - 83)." Electoral Studies 4, no. 2 (1985): 179-81.

Barnes, G.P. "The Use of Computers in Redistributing Constituencies." Electoral Studies 6, no. 2 (1987): 133-38.

Blais, Andre and Louis Massicotte. "Electoral Systems." In Comparing Democracies: Elections and Voting in Global Perspective, edited by Lawrence LeDuc, Richard Niemi and Pippa Norris. Thousand Oaks, California: Sage Publications, 1996.

Bogdanor, Vernon and David Butler, eds. Democracy and Elections: Electoral Systems and Their Political Consequences. Cambridge: Cambridge University Press, 1983.

Butler, David and Bruce Cain. Congressional Redistricting: Comparative and Theories Perspectives. New York: Macmillan, 1992.

Butler, David, and Bruce Cain. "Reapportionment: A Study in Comparative Government." Electoral Studies 4, no.3 (1985): 197-213.

Cain, Bruce. The Reapportionment Puzzle. Berkeley: University of California Press, 1984.

Carstairs, Andrew McLaren. A Short History of Electoral Systems in Western Europe. London: Allen and Unwin, 1980.

Courtney, John C. *Commissioned Ridings: Designing Canada’s Electoral Districts*, Montreal: McGill-Queens’ University Press, 2001.

Courtney, John C. "Discrimination in Canada's Electoral Law." In Discrimination in the Law and the Administration of Justice, edited by Walter Tarnopolsky, Joyce Whitman and Monique Ouellette. Montreal: Les Editions Themis, 1993.

Courtney, John C. "Drawing Electoral Boundaries." In Canadian Parties in Transition, edited by A. Brian Tanguay and Alain G. Gagnon. Scarborough, Ontario: Nelson Canada, 1996.

Courtney, John C. "Electoral Boundary Redistributions: Contrasting Approaches to Parliamentary Representation." In Comparative Political Studies Australia and Canada, edited by Malcolm Alexander and Brian Galligan. Melbourne: Pitman, 1992.

Courtney, John C. "Parliament and Representation: The Unfinished Agenda of Electoral Redistributions." Canadian Journal of Political Science 21, no.4 (December 1988): 675-90.

Courtney, John C. "Theories Masquerading as Principles: Canadian Electoral Boundary Commissions and the Australian Model." In The Canadian House of Commons: Essays in Honour of Norman Ward, edited by John Courtney. Calgary: University of Calgary Press, 1985.

Courtney, John C., Peter MacKinnon and David Smith, eds. Drawing Boundaries: Legislatures, Courts and Electoral Value. Saskatoon: Fifth House Publishers, 1992.

Dixon, Robert. Democratic Representation: Reapportionment in Law and Politics. New York: Oxford University Press, 1968.

Farrell, David. Comparing Electoral Systems. London: Prentice Hall, 1997.

Grofman, Bernard and Arend Lijphart, eds. Electoral Laws and Their Political Consequences. New York: Agathon Press, 1986.

Grofman, Bernard, Arend Lijphart, Robert McKay and Howard Scarrow, eds. Representation and Redistricting Issues. Lexington, Massachusetts: Lexington Books, 1982.

Handley, Lisa and Bernard Grofman, eds. Redistricting In Comparative Perspective. Oxford ; New York : Oxford University Press, 2008

Henry, Carlton. "The Impact of New Technology and New Census Data on Redistricting in the 1990s." In Redistricting in the 1990s: A Guide for Minority Groups, edited by William O'Hare. Washington D.C.: Population Reference Bureau, 1989.

Hughes, Colin. "The Case of the Arrested Pendulum." In The Australian National Elections of 1977, edited by Howard Penniman. Washington D.C.: American Enterprise Institute, 1977.

Jaersch, Dean. Election! How and Why Australia Votes. St. Leonards, Australia: Allen and Unwin, 1995.

Johnston, R.J. "Constituency Redistribution in Britain: Recent Issues." In Electoral Laws and Their Political Consequences, edited by Bernard Grofman and Arend Lijphart. New York: Agathon Press, 1986.

Johnston, R.J. Political, Electoral and Spatial Systems. London: Oxford University Press, 1979.

Johnston, R.J. "Redistricting by Independent Commissions: A Perspective from Britain." Annals of the Association of American Geographers 72, no. 4 (1982): 457-70.

Lijphart, Arend. "Proportionality by Non-PR Methods: Ethnic Representation in Belgium, Cyprus, Lebanon, New Zealand, West Germany and Zimbabwe." In Electoral Laws and Their Political Consequences, edited by Bernard Grofman and Arend Lijphart. New York: Agathon Press, 1986.

Mair, Peter. "Districting Choices under the Single-Transferable Vote." In Electoral Laws and Their Political Consequences, edited by Bernard Grofman and Arend Lijphart. New York: Agathon Press, 1986.

McLean, Iain and Roger Mortimore. "Apportionment and the Boundary Commission for England." Electoral Studies 11, no. 4 (1992): 293-309.

McRobie, Alan. "The Electoral System." In Essays on the Constitution, edited by Philip Joseph. Wellington, New Zealand: Brookers' Ltd., 1995.

McRobie, Alan. "The Electoral System and the 1978 Election." In New Zealand at the Polls: The General Election of 1978, edited by Howard Penniman. Washington D.C.: American Enterprise Institute, 1980.

Milen, Robert, ed. Aboriginal Peoples and Electoral Reform in Canada. Volume 9 in a series commissioned by the Royal Commission on Electoral Reform and Party Financing. Toronto: Dundurn Press, 1991.

Morrill, Richard. Political Redistricting and Geographic Theory. Washington D.C.: Association of American Geographers, Resource Publication in Geography, 1981.

O'Hare, William P. and Taynia Mann. Redistricting: A Guide to Technical Issues for Minority Groups. Washington D.C.: Population Reference Bureau, Inc., 1991.

Reynolds, Andrew and Ben Reilly. The International IDEA Handbook of Electoral System Design. Stockholm: International Institute for Democracy and Electoral Assistance, 1997.

Rossiter, D.J., R.J. Johnston, C.J. Pattie, *The Boundary Commissions: Redrawing the UK’s map of Parliamentary constituencies*, Manchester: Manchester University Press, 1999.

Rush, Mark. Does Redistricting Make a Difference? Baltimore: The Johns Hopkins University Press, 1993.

Steed, Michael. "The Constituency." In Representatives of the People? Parliamentarians and Constituencies in Western Democracies, edited by Vernon Bogdanor. Hants, England: Gower Publishing Company, 1985.

Small, David, ed. Drawing the Map: Equality and Efficacy of the Vote in Canadian Electoral Boundary Reform. Volume 11 in a series commissioned by the Royal Commission on Electoral Reform and Party Financing. Toronto: Dundurn Press, 1991.

Smith, Jennifer. "Drawing Electoral Boundaries in Canada: Current Representation Dilemmas." Paper presented at the at a conference hosted by the National Center for Geographic Information and Analysis in Buffalo, New York, October 24-26, 1997.

Stewart, Alan. "Community of Interest in Redistricting." In Drawing the Map: Equality and Efficacy of the Vote in Canadian Electoral Boundary Reform, edited by David Small. Volume 11 in a series commissioned by the Royal Commission on Electoral Reform and Party Financing. Toronto: Dundurn Press, 1991.

Taagepera, Rein and Matthew Sobert Shugart. Seats and Votes: The Effect and Determinants of Electoral Systems. New Haven, Connecticut: Yale University Press, 1989.

Taylor, Peter J. and Graham Gudgin. "A Fresh Look at the Parliamentary Boundary Commissions." Parliamentary Affairs 28, no. 4 (1975): 405-415. 1955-70."

Taylor, Peter J. and Graham Gudgin. "The Myth of Non-Partisan Cartography: A Study of Electoral Biases in the English Boundary Commission's Redistribution for 1955-70." Urban Studies 13 (1976): 13-25.

Taylor, Peter J. and R.J. Johnston. Geography of Elections. New York: Holmes and Meier Publishers, 1979.

Waller, Robert J. "The 1983 Boundary Commission: Policies and Effects." Electoral Studies 2, no. 3 (1983): 195-206.

 

Contributors

 

The Boundary Delimitation topic area was written by Dr. Lisa Handley and further updated by Dr. Handley under the supervision of IFES from 2005-2006 thanks to the generous financial support of the United States Agency for International Development (USAID).

Lisa has more than fifteen years of experience in the area of electoral boundary delimitation (redistricting), both as a practitioner and academic. She is a partner at Frontier International Electoral Consulting, LLC, a company with expertise in boundary delimitation, voting analyses and political technology. She has advised numerous jurisdictions and other clients on redistricting and voting rights issues and has served as an expert in dozens of redistricting and voting rights court cases.

Additional updates in 2012 were undertaken by Herschell Sax who was Elections Canada’s subject matter expert for boundary delimitation  from 1993 to his retirement in 2011. During this period, he met with all foreign delegations to Elections Canada and represented Canada abroad on the subject of redistribution.

 

Case Studies

 
 

Afghanistan: Delimiting Districts for Transitional Elections in a Post-Conflict Society

Note: This case study is from before the 2005 elections  

Organizing elections in Afghanistan that will be viewed by all major stakeholders as free and fair presents a major technical and logistical challenge to the United Nations (UN) [1] – a challenge that is even further complicated by the tight timeframe and the ongoing security issues. Two of the many problems facing the UN in planning for the parliamentary and local elections in Afghanistan are the delimitation of provinces and districts and the allocation of parliamentary seats to provinces on the basis of population.

Background

Afghanistan has been at war for most of the last 25 years. The Soviet Union invaded in 1979 (following a communist coup within the country), but was finally forced to withdraw 10 years later by anti-Communist mujahidin forces supplied and trained by the US, Pakistan, and others. Fighting subsequently continued among the various mujahidin factions, giving rise to a state of “warlordism” that eventually spawned the Taliban.

Backed by foreign sponsors, the Taliban developed as a political force and eventually seized power over most of the country, aside from Northern Alliance strongholds primarily in the northeast. Following the September 11th 2001 terrorist attacks, however, a U.S.-lead invasion forced the Taliban from power.

Shortly after the fall of the Taliban, representatives of various Taliban opposition groups met under the auspices of the United Nations in Bonn, Germany, and agreed on a plan for the formulation of a new government. Hamid Karzai was inaugurated Chairman of the Afghan Interim Authority (AIA) on December 22, 2001 and in June 2002 he was elected President by a national Loya Jirga (General Assembly).

The Transitional Authority was given an 18-month mandate within which to convene a Loya Jirga and adopt a constitution, and a 24-month mandate to hold nationwide elections. A constitution was ultimately adopted on January 4, 2004 (several months after the deadline); elections were originally scheduled for June 2004, but these were postponed until October 2004. Although both the Bonn Agreement and the Constitution specify that, if at all possible, the first elections should include both presidential and parliamentary elections, only presidential election will be held in 2004. Parliamentary elections are currently scheduled for April 2005.

Delimiting Boundaries for the Parliamentary Elections

The Constitution dictates that the National Assembly in Afghanistan be composed of two chambers: the Wolesi Jirga and the Meshrano Jirga. Only when both houses are in session can the National Assembly fulfill its legislative functions (Chapter 5, Article 87 & 94).

Both parliamentary and local elections must be conducted in order to convene the National Assembly: Parliamentary elections are required to select Wolesi Jirga representatives and local elections (provincial and district) must be held in order to indirectly elect members to the Meshrano Jirga.

Before either set of elections can occur, however, consensus must be reached on the configuration of provinces and districts within Afghanistan because both parliamentary and local elections depend on these administrative units to serve as electoral constituencies.

Constituencies for Parliamentary and Local Elections

The boundaries of electoral constituencies must be established before parliamentary and local elections can proceed.

Using Provinces as Electoral Constituencies for the Wolesi Jirga Because a regional rather than a single-constituency List Proportional Representation (PR) electoral system is being introduced for election to the Wolesi Jirga, some delimitation of constituencies will be necessary. This process is likely to be limited to the adoption of provinces as electoral constituencies and the apportionment of parliamentary seats to these provincial constituencies on the basis of population.

In Afghanistan, this process will be complicated by several factors, one of which is a possible change in the number, and therefore the boundaries, of the provinces[2]. In early March 2004 (and for many years prior to this), there was general agreement that 32 provinces exist; however, pending before the Ministry of the Interior were plans for at least three additional provinces. At the end of March, one of these proposed provinces was declared a new province by President Karzai. More decrees establishing new provinces may follow even closer to the election.

Electing Representatives to the Meshrano Jirga According to the Constitution, elected members of the Meshrano Jirga are to originate from two sources[3]:

  • Each provincial council is to elect one member of the council to serve as a member of the Meshrano Jirga
  • From among the district councils of each province, the respective councils are to elect one person to serve as a member of the Meshrano Jirga

One important consequence of this constitutional provision is that provincial and district (local) elections must occur before the Meshrano Jirga can be convened. Local elections will require clearly defined provinces and districts. However, as mentioned above, provincial lines may be in flux. District boundaries are even more problematic.

Delimiting District Boundaries

There is no consensus on the number of districts, or on what the boundaries of these districts might be in Afghanistan. The Appendix provides a comparison of some of the more relevant lists of districts in Afghanistan: the United Nations Assistance Mission in Afghanistan (UNAMA) list of districts by province used by the Electoral Component of UNAMA for voter registration purposes; an updated list of provinces and districts provided by the Afghan Central Statistics Office (CSO) in March 2004; and the list of districts used for the Emergency Loya Jirga (ELJ) elections in 2002 and the Constitutional Loya Jirga (CLJ) elections in 2004.

A comparison of district lists from these sources provides only an indication of the problem – even if all of the lists concur on the number of districts to be found in a given province (which they do not), this does not mean that there is agreement on what the boundaries of these districts are, and hence what villages are assigned to each district. On the other hand, if the total number of districts per province does not correspond across lists, then clearly there is disagreement as to village-to-district-to-province assignments.

According to the village-to-district-to-province assignment list provided by the CSO to UNAMA in November 2003, there are 32 provinces and 387 districts. Although UNAMA regarded this information as static, the CSO is in fact updating this information as the pre-census enumeration process continues and more up-to-date information is obtained.

The CSO began the pre-census process in January 2003 with what was known to be an outdated list of provinces, districts and villages – the list was compiled in 1979 for the only attempted census in Afghanistan, which was never completed. The CSO census team has been updating this list on a rolling basis[4]; as of March 2004, district lists for 22 of the currently existing 32 provinces have been updated.

UNAMA designed the voter registration process on the basis of the village-to-district-to-province list provided by the CSO in November 2003; the UNAMA list has not updated since this date, despite periodic updates in the village-to-district-to-province list issued by CSO.

The UNAMA staff has recognized the problematic nature of their village-to-district assignment list[5]. Internal checks on the voter registration data collected to date, for example, have determined that:

  • 12% of the villages identified by registered voters are in “unlisted” villages (that is, do not appear on the UNAMA list)
  • 3% of those registered thus far have registered to vote in an “unlisted” village and this percentage is likely to increase substantially as the registration process moves from the urban to more rural areas[6].

Even more problematic, however, than the discrepancies in district lists among various agencies is the fact that even the most recent list of districts (as represented by the updated CSO list) may not reflect reality. Even though there are ostensibly rules and regulations for the establishment of new provinces and districts, these rules are not strictly enforced, especially with regard to the formation of new districts[7]. As became evident during preparations for the Emergency Loya Jirga (ELJ) elections in 2002, quite a number of districts have “emerged” that the CSO seems unaware of.

In many cases, the new districts are the result of provincial governors rewarding supporters with administrative positions, or local commanders declaring the existence of new districts to award themselves with administrative positions. These districts, although not necessarily “recognized” by the CSO or the Ministry of the Interior, are usually acknowledged by the local elders and accepted by local voters. In order to conduct ELJ and CLJ elections, “negotiations” often had to occur between the ELJ/CLJ Commission and the local communities, usually resulting in the de facto recognition of these “new” districts for the ELJ and CLJ elections[8]. Ultimately, 465 districts were included in the ELJ/CLJ elections.

Data for Apportioning Wolesi Jirga Seats to Provincial Constituencies

Wolesi Jirga seats are to be apportioned to electoral constituencies on the basis of “population” according to the Constitution. The Constitution does not specify the type of population data to be used, but the apportionment process is almost always based on either census enumeration data or voter registration data[9].

Population data may also be required to determine the number of members to sit on the provincial and district councils. Although it is not strictly necessary, presumably the same data that is used to apportion parliamentary seats will be used to gauge the size of local councils.

At this point in time (March 2004), there is a complete absence of reliable population data on which to apportion parliamentary seats. No census, with the exception of an aborted pre-census in the late 1970s[10], has ever been conducted in Afghanistan. Nor is there a pre-existing civil registry or voters’ list available. By Election Day, however, there will hopefully be several sources of population data available for apportionment purposes. These include:

  • Pre-census enumeration data
  • Voter registration data
  • Voter turnout data
  •  

Census Data A pre-census household enumeration is currently being conducted by the Central Statistics Office (CSO), with the assistance of the United Nations Population Fund (UNFPA). This phase of the census process, referred to as the household-listing phase, entails visiting every household in Afghanistan and counting the total number of persons in each household. The enumeration includes not only the total number of persons, but number of males and females and the number of persons under the age of 18 and over the age of 18 per household.

The enumeration process commenced in the province of Kunduz in January of 2003 and 22 provinces have since been completed. The anticipated completion date for this project, assuming no major obstacles are forthcoming (i.e., security issues), is June 2004. Depending on whether this schedule is maintained, and when the parliamentary elections are scheduled, this enumeration data may be available for apportioning parliamentary seats to provincial constituencies.

Voter Registration Data The Electoral Component of the United Nations Assistance Mission in Afghanistan (UNAMA) is currently compiling a voters list which will, of course, have to be completed by Election Day – how many days, weeks or months before Election Day, however, remains a question. This voter registration data could be used to apportion seats to parliamentary constituencies.

The advantage of using voter registration data, as compared to census data, is the fact that this data is likely to be less outdated than census data for seat apportionment purposes, at least in the long run. (For this first election, both the enumeration data and the voters’ list data should be equally up-to-date.) The disadvantage of utilizing this data is that the voters’ list may not be reflective of the total population, and may be more biased in some regions than others. This is especially true in areas where potential voters have decided to boycott the registration process for social or political reasons.

Voter Turnout Data Apportioning seats on the basis of posteriori votes is another, albeit far less common, possibility[11]. Theoretically, this approach offers an increased incentive for voters to turn out and cast a ballot. However, this supposition has never been tested and proven. What we do know is the following disadvantages of using posteriori votes: 

  • The ambitious election timetable currently in place in Afghanistan may not be able to accommodate the added pressure on the civic education program to ensure that community leaders and voters understand all of the ramifications of NOT voting, and the voter registration program to make certain that ALL eligible voters can cast a ballot.
  • Any inclination to boycott the voter registration process or vote (or any bias in the registration process) will have not only immediate partisan ramifications (number of parliamentary seats won), but ramifications for geographic representation that will remain in place until the next set of elections. 
  • Given the currently scheduled release date for the CSO pre-census enumeration data, election officials (and any other stakeholders interested in carrying out the exercise) will be able to calculate the "effectiveness" of the registration exercise and to ascertain any bias in seat apportionment due to "ineffective" registration even before Election Day. If such a bias exists, this cannot be corrected (and may, in fact, be exacerbated) by apportioning seats on the basis of voter turnout.

Problematic Draft Electoral Law

Although the draft electoral law (dated 29 February 2004) on “Electoral Boundaries” provides for a streamlined process and a dispute resolution mechanism, there are some problems with the law as drafted. The major problem is the lack of a clearly delineated (or insufficient) timeframe for the identification of electoral boundaries and the resolution of boundary disputes. The lack of specification in portions of the “Electoral Boundaries” draft law may also cause difficulties. What follows are some comments on each of the provisions encompassed in Chapter III.

Article 11. Use of Existing Administrative Boundaries For the purposes of electing members of the Wolesi Jirga, provincial councils, and district councils, the provinces and districts entitled to elect representatives will be those designated by the President in a decree no later than 90 days prior to the election.

Article 11 does not make clear whether the President’s decree merely lists the names of the provinces and districts entitled to elect representatives or specifies the boundaries of these provinces and districts (and records the villages assigned to each district, and the districts assigned to each province). If the boundary delimitation process is to proceed in a timely manner, it is important that the decree include all information necessary to assess the electoral constituencies (provincial and district boundaries) and assign eligible voters to the correct district and province.

The time period established for the release of the Presidential decree – 90 days prior to the election – does not provide sufficient time to resolve possible boundary disputes and still organize elections effectively. (If a boundary dispute arises, this must be resolved before election administrators can assign voters to the correct districts and provinces.) In fact, the entire process, including the resolution of boundary disputes, should almost certainly be completed approximately 90 days prior to the election. Since the boundary dispute process is liable to take a minimum of 60 days to resolve (from the initiation of the evaluation to the resolution of any disputes), this suggests that the 90 day deadline for issuing the decree must be moved back at least an additional 60 days to 150 days prior to Election Day.

Article 12. Electoral Boundary Disputes (1) Disputes relating to the precise boundary of any province or district shall be heard and resolved for electoral purposes only by the JEMB (Joint Election Management Body).

Article 12 does not indicate who is allowed to initiate a dispute. If no limitations are placed on potential complainants, then the dispute process may be subject to manipulation. On the other hand, the process should be as open and democratic as possible, so limiting possible complainants excessively would not be wise. At a minimum, only citizens that are residents of the district or province being disputed should be permitted to lodge a complaint.

Article 12 also does not specify how a dispute is to be recorded. For example, is a written complaint submitted to the JEMB? Or should the JEMB (or perhaps some local entity like the Regional Electoral Coordinator) be required to hold a hearing, or series of hearings, allowing citizens to object to provincial/district boundaries? A public hearing process would be time-consuming, but would be more inclusive.

There is no time period stipulated for submitting a dispute to the JEMB. But if disputes are to be resolved in a sufficiently timely manner (approximately 90 days prior to the election), then they must be initiated well before 90 days prior to the election. (Allowing disputes closer to the election could result in the disruption of an election.)

(2) The Regional Electoral Coordinator shall gather information regarding the dispute from the disputing parties and other local sources and forward that information to the JEMB.

A time limit should be placed on the Regional Electoral Coordinator for gathering and forwarding information so that the dispute resolution process is not delayed. Furthermore, if the JEMB is to decide boundary disputes on the basis of “technical criteria” (Article 12.3), then a requirement that the Regional Electoral Coordinator gather information related to these “technical criteria” should be included.

(3) The JEMB, following consultation with the Minister of Interior, shall decide the dispute on the basis of technical criteria, taking into account the information made available to it by the Regional Electoral Coordinator, and shall notify the Electoral Secretariat, the disputing parties, and local government offices of its decision.

No indication of what “technical criteria” might be is provided in the text. Perhaps the purpose of including this stipulation is to avoid a decision motivated by political considerations, in which case “objective criteria” might be a better choice of words. On the other hand, perhaps “technical criteria” is meant to refer to such “traditional”, or “internationally accepted”, districting criteria as the size of the population, geographic contiguity and compactness, and remoteness of territory. Some consideration should be given to rephrasing the “technical criteria” requirement.

If the dispute resolution leads to a change in the boundaries as provided by the Presidential decree, some thought should be given to implications of this. Despite the admonishment that the boundary resolution is “for electoral purposes only”, the election of a district council to a district not included in the original Presidential decree, for example, has ramifications far beyond Election Day. For instance, district and provincial councils presumably have administrative functions associated with them. Are the President and Ministry of the Interior obliged therefore to accept the JEMB resolution? If so, perhaps the resolution of boundary disputes should not reside solely with the JEMB (in consultation with the Minister of the Interior). It may be that a separate Delimitation Commission, with representatives from both the JEMB and the Ministry of the Interior, should convene to resolve boundary disputes.

Finally, no deadline for resolving the dispute and notifying the Electoral Secretariat, the disputing parties and the local government offices, is included in the provision. But any boundary dispute must be resolved in sufficient time to allow election administrators and candidates to prepare for the election. Therefore the inclusion of a timetable for the dispute resolution process, and an adjustment to the deadline for the Presidential decree, is quite important.

 

Task

Time Frame

Presidential decree identifying provinces and districts, and the boundaries of the provinces and districts, entitled to representation

150 days prior to election

Submission of written complaints or completion of public hearing process

120 days prior to election

Resolution of boundary disputes

90 days prior to election

Conclusion

Before parliamentary and local elections can take place, a decision will have to be reached as to the number of districts and provinces and the boundaries of these districts and provinces (or, if not the precise boundaries, then at least the village-to-district-to-province assignments). If new provinces (and districts) are to be created, acknowledgment of this must be forthcoming immediately. Furthermore, a moratorium on the formation of additional provinces and districts must be declared well before elections are to be held.

Although it is premature at this point to decide what data to use to apportion Wolesi Jirga seats, an analysis of potential data sources, and the reliability of these sources, should be conducted closer to Election Day. Assuming that the pre-census process has been completed and the enumeration data has been deemed reliable, this data may be the best option for apportioning parliamentary seats and determining the size of provincial and district councils. This is because the pre-census program is likely to reach more people than the voter registration process[12].

The electoral law, as currently written, should be revised. The major problem with the current draft is the inadequate timeframe for identifying electoral boundaries and resolving boundary disputes. A lack of specification in portions of the “Electoral Boundaries” draft law may also cause difficulties.

SEE TABLE in APPENDIX

Notes:

[1] The United Nations was responsible for organizing and conducting the upcoming elections in Afghanistan.

[2] According to the Cartography Department in Afghanistan, three potential provinces have been mapped, not just one – but these maps are not to be released until/unless the Interior Minister approves the establishment of the new provinces.

[3] The President is to appoint one-third of the Meshrano Jirga members; the other two-thirds are elected from the provinces, one-third from among the members of the provincial councils and the other one-third from among the district councils within each province.

[4] The list for each province was updated in consultation with the provincial governor, district leaders and village elders. Any disagreement in the list was to be resolved by the provincial governor.

[5] The information reported here was supplied by Matthew Blakely, Data Manager, Electoral Component of UNAMA.

[6] In one relatively rural province in which voter registration is currently underway (Bamyan), 9% of those that have registered have identified “unlisted” villages as their place of residence.

[7] The rules and regulations for creating new provinces and districts are the responsibility of the Local Areas Management Body, located within the Ministry of the Interior.

[8] The negotiation process at times was quite “delicate” and even led to violent clashes on occasion, according to informed sources.

[9] The choice of whether to use census data or voter registration data may be guided by either practical or theoretical concerns. For instance, census data may not be the best option if a general enumeration of the population is unavailable, outdated or inaccurate. On the other hand, registration data may not adequate for districting purposes if it fails to include information (i.e., demographic data) that is essential given the specific country context. From a theoretical perspective, delimitation based on registration data is likely to produce constituencies that are more equal with respect to the number of voters contained within them, but a counter-argument could be made that representatives serve all persons, not simply voters, and therefore districts should contain equal numbers of persons, not voters.

[10] The census was aborted in 1979 when the Soviets invaded Afghanistan.

[11] According to research in ACE's Comparative Data, only one country (Belarus) surveyed uses voter turnout data for apportioning seats to the legislature.

[12] The first phase of the census is relying on door-to-door contact rather than centrally located registration sites; the enumeration program is also less likely to be boycotted for social or political reasons.

Australia: Using Projections to Equalize Electoral District Populations

In Australia the overriding principle in delimiting electoral boundaries, or redistribution as the process is referred to in Australia, is population equality. Redistributions in Australia endeavor to ensure that each State and Territory is granted representation in the House of Representatives in proportion to the State or Territory’s population, and that there are as nearly as practicable the same number of electors in each electoral division (district) for a given State or Territory [1].

The electoral law is very specific about how much variation from population equality will be tolerated: At the time of the redistribution the number of people enrolled in each electoral division may not vary from the population quota (average) by more than 10 percent. More unique is the requirement that at the point three and a half years after the expected completion of the redistribution, the division populations should not vary from the average by more or less than 3.5 percent. It is these very precise, objective numerical criteria that drive the redistribution process in Australia.

Timing of Redistributions

The Commonwealth Electoral Act 1918 provides the machinery and principles for revising the electorate boundaries in each State for the House of Representatives. Under Section 59 of the Electoral Act, there are only three situations in which a redistribution of a State or Territory represented in the Parliament can be initiated: 

  • When changes in the distribution of population require a change in the number of members of the House of Representatives to be allocated to a State or Territory;
  • When more than one third of the divisions within a State deviate from the average divisional enrolment for the State by more than 10 percent, and have done so for more than two consecutive months; or
  • When seven years have elapsed since the State or Territory was last redistributed

Of the three possible “triggers” for redistributions, change in representation entitlements has accounted for the majority of redistributions initiated since 1984; redistributions triggered by malapportionment (deviating from the divisional enrolment by more than 10 percent) are rare. (Appendix A provides a list and the dates of all redistributions commenced since 1901.)

Calculating Apportionment

The method used to determine the number of House of Representative seats each State is entitled to is specified both in section 24 of The Constitution and section 48 of the Commonwealth Electoral Act 1918. During the thirteenth month after the first sitting of the House of Representatives, the Electoral Commissioner is required to ascertain the population of the Commonwealth (excluding the Territories) according to the latest official statistics available from the Australian Statistician. These figures are then used to determine how many Members of the House of Representatives (and therefore divisions) each State is entitled to. A similar exercise is used to calculate the entitlements of the Territories [2].

Section 48 of the Electoral Act prescribes the calculation as follows:

(a) A quota shall be ascertained by dividing the number of people of the Commonwealth, as ascertained in accordance with section 46, by twice the number of Senators for the States; (b) The number of members to be chosen in each State shall be determined by dividing the number of people of the State, as ascertained in accordance with section 46, by the quota, and had if on such a division there is a remainder greater than one-half of the quota, one more member shall be chosen for the State.

For example, the determination of representation entitlements on 19 February 2003 was made as follows:

Step 1:   Quota   

 

total population of the six states                  or         19205190

2  x  number of Senators for the States                    72 x 2

Quota= 19205190/(72 x 2)= 133369.375

  

Step 2:  Number of Members

total population of individual State or Territory/Quota

 

Based on these calculations, the following apportionment of Members (and electoral divisions) was announced:

 

 

Population

Entitlement

Number of Members

Change

New South Wales

6657478

49.9176

50

 

Victoria

4888243

36.6519

37

 

Queensland

3729123

27.9609

28

+1

Western Australia

1934508

14.5049

15

 

South Australia

1522467

11.4154

11

-1

Tasmania

473371

3.5493

5*

 

Australian Capital Territory (ACT)

322871

2.4209

2

 

Northern Territory

199760

1.4978

1

-1

Total

 

 

149

 

 

 

  • The Constitution states that at least five Members shall be chosen from each of the original six States and therefore Tasmania is guaranteed a minimum of five Members.

Redistribution must ensue if the number of Members allocated to a State or Territory changes relative to the previous number. The 2003 apportionment triggered redistributions in Queensland (which gained a Member relative to the previous allocation) and South Australia (which lost a Member relative to the previous allocation).

Redistribution Authority

Once it is determined that a State or Territory must be redistributed, a Redistribution Committee for that State or Territory is established. The Commonwealth Electoral Act 1918 requires that each committee consist of the following people: 

  • The Electoral Commissioner (EC) of the Australian Electoral Commission – this statutory officer is the administrative head and one of three members of the national commission
  • The Australian Electoral Officer (AEO) for the particular State or Territory – this statutory officer heads the Australian Electoral Commission’s administrative structures in the State
  • The Surveyor-General (SG) for the State (or, if the State has no Surveyor-General, the person holding an equivalent office)
  • The Auditor-General (AG) for the State (or, if unavailable, a senior Australian public servant)

The Redistribution Committee has the task of producing a set of boundaries and names for electoral divisions for the House of Representatives in that particular State or Territory.

After the proposed redistribution has been published, an augmented Electoral Commission is established to consider these objections and make a final determination of the boundaries. This body consists of the Redistribution Committee for the State, plus the two members of the three-member Australian Electoral Commission who were not members of the Redistribution Committee: the Chairperson of the Commission and the so-called non-judicial Commissioner (the Australian Statistician).

There appears to be a consensus among Australia’s main political parties that the membership of the Redistribution Committees and augmented Electoral Commissions is appropriate and that these bodies operate in an independent and politically neutral manner.

Steps in the Redistribution Process

Once the Electoral Commission directs the commencement of the redistribution, the enrolment quota for the particular State or Territory is determined by the Electoral Commissioner and the Redistribution Committee for that State or Territory is appointed by the Electoral Commission.

The Electoral Commissioner invites written suggested from interested person or organizations which must be provided to the Redistribution Committee within 30 days. The suggestions are made available for public inspection and comments may be lodged within 14 days.

The Redistribution Committee then considers the public suggestions and comments and develops a set of proposed boundaries. The plan is published and maps showing the suggested boundaries and names are publicly exhibited. The public has 28 days to lodge objections to the proposals. Then there is a period of 14 days in which interested people and organizations may make comments on the objections.

The augmented Electoral Commission is established to consider these objections and make a final determination of the boundaries. The augmented Electoral Commission has 60 days to consider objections (including initial and any further objections). After considering any further objections, the augmented Electoral Commission makes a final determination of the boundaries of the electoral divisions. The Commission is required to state in writing its reasons for its final determination, and provision exists for any member who dissents to indicate his or her reasons for doing so [3]. (Appendix B provides the timetable for redistribution.)

The determination made by the augmented Electoral Commission is final and cannot be vetoed at the political level [4]. It is also not subject to judicial review, except on very limited constitutional grounds.

Provision for Public Input

Extensive provision is made for public input into the redistribution process. Redistribution Committees are required by law to call for public suggestions at the outset of the redistribution process, and any suggestions received are made publicly available, and can be the subject of further public comment. All suggestions and comments are required to be considered.

After a proposed redistribution has been published, objections to it may be lodged. The augmented Electoral Commission is required to conduct a public inquiry into an objection unless the objection covers matters which were already substantially raised in earlier suggestions or comments, or is frivolous or vexatious. Submissions regarding the objection may be made at the public inquiry by the person or organization objecting, and any person who or organization that made suggestions or comments. The augmented Electoral Commission may also invite other witnesses to appear.

The augmented Electoral Commission is not bound by the legal rules of evidence, and in general has considerable flexibility in determining how inquiries will be conducted. Inquiries are typically held in the capital city of the State, but on some occasions in the past an augmented Electoral Commission has chosen to hold an inquiry in a provincial city, particularly where such an approach facilitated the examination of a controversial proposed boundary.

Having held such public inquiries as are necessary, the augmented Electoral Commission is required to make a further proposed redistribution. As part of that process, the augmented Electoral Commission must determine whether in its view the proposed redistribution is “significantly different” from that put forward earlier by the Redistribution Committee. If the proposed redistribution is not regarded as significantly different from that of the Redistribution Committee, there is no provision for further objections, and the proposed redistribution is in practice reflected in the final determination made by the augmented Electoral Commission. If the proposed redistribution is regarded as significantly different, further objections can be lodged, and another public inquiry, or series or public inquiries, must be held.

Participants at inquiries into objections have typically been given extensive opportunities to elaborate on their submissions such that the augmented Electoral Commission gains the maximum benefit from their views and so that the participants can be satisfied that they received a proper hearing. In addition, the Redistribution Committees and augmented Electoral Commissions have tended to provide substantial explanations of their proposals, addressing in some detail the suggestions, comments and objections have been raised, in their statements of reason. As a result, those who have made suggestions, comments or objections can better appreciate why their own preferred positions may not have been accepted.

The net effect of these provisions for public input has been to make the redistribution process open and accountable and to ensure that arguments of the substance of a particular redistribution has tended to be worked out within the framework of the process, rather than forming the basis for ongoing questioning of the legitimacy of a particular redistribution.

Criteria Governing the Drawing of Electoral Boundaries

The criteria in accordance with which boundaries are to be devised are fully spelt out in the Commonwealth Electoral Act 1918. These criteria are as follows:

  • The State or Territory must be distributed into a number of electoral divisions equal to its representation entitlement in the House of Representatives as most recently determined.
  • A quota is determined by dividing the State or Territory’s total electoral enrolment as at the end of the period for the lodging of comments by its representation entitlement in the House of Representatives. An electoral division, which deviates in enrolment from the quota by more than 10 percent, cannot be proposed.
  • Subject to the rules above, the Redistribution Committee or augmented Electoral Commission must, as far as practicable, endeavour to ensure that the boundaries of the electoral divisions are such that the number of electors enrolled in each division will not, at the point three-and-a-half years after the redistribution has come into effect, be less than 98 percent nor more than 102 percent of the average divisional enrolment for the State or Territory at that time.
  • Subject to the three rules above, due consideration must be given, in relation to each proposed division, to:

    1) Community of interests within the proposed division, including economic, social and regional interests;

    2) Means of communication and travel within the proposed division;

    3) The physical features and area of the proposed division; and

    4) The boundaries of existing divisions. 

The criteria listed in the final paragraph are clearly subordinate to the numerical criteria outlined in preceding points and, as such, tend to come into play “at the margins”, when the issue is not the major structure of the redistribution, but rather the precise positioning of a few of the boundaries.

The relative importance of the criteria is clearly defined in the legislation, and is not a matter on which the bodies conducting redistributions have any significant discretion. Furthermore, the Redistribution Committee and the augmented Electoral Commission are not required to take any account of the possible partisan implications of redistributions, and in fact have not done so.

Population Equality

In determining electoral boundaries within a State the overriding consideration in Australia is population equality.

While the total population is used in allocating Members to the States and Territories, actual voter numbers are used in determining the delineation of electoral districts. Since voter registration is compulsory in Australia, these figures are readily available and are for all practical purposes accurate at the time of calculation.

In applying the principle of “one person, one vote”, the starting point is to determine the ideal population for each electoral district. This is calculated by dividing the total number of electors in a particular State or Territory by the number of House of Representative seats allocated to the State or Territory. The result is rounded to the nearest integer. The population of electoral districts within a given State or Territory can deviate from this ideal population by as much as 10 percent at the time of the redistribution.

In addition to the requirement that an electoral district population not deviate by more than 10 percent from the ideal population at the time of redistribution, another metric is used to prevent malapportionment occurring during the life of the redistribution: Each electoral district can deviate by no more than 3.5 percent in 3 years and 6 months from the expected completion of the redistribution (or the midpoint in the possible 7 year period since the last redistribution). This requirement necessitates the production of population projections [5].

Because of this latter requirement, two enrolment quotas must be calculated during the redistribution process:

  • The current quota or average district enrolment; and
  • The projected quota or average district enrolment three and a half years after the expected completion of the redistribution

In each case, the quota is calculated by dividing the number of people enrolled by the number of members to which the State or Territory is entitled.

The current quota uses the number of electors on the roll when the redistribution commences and is permitted to vary up to 10 percent in each district. The projected quota, on the other hand, is based on a projected enrolment figure three and a half years after the expected completion of the redistribution and may vary by no more than 3.5 percent in each division. In aiming for this quota, the Redistribution Committee is able to allow for population growth or contraction in particular areas.

For example, a determination of enrolment quotes for the State of Victoria at the beginning of the redistribution process in 2003 was as follows: Number of electors enrolled in VIC (3259454) = 88093 Number of districts in which VIC is to be distributed (37)

Therefore at the time of the redistribution the number of electors in the electoral districts could vary up to 10% from 88093 (with the acceptable population range therefore being 79284 to 96902).

The projected quota for the State of Victoria was calculated as follows:

Projected enrolment in VIC at 31.03.2006 (33473637) = 93882 Number of members VIC is entitled to (37)

The projected average enrolment at March 2006 (in three and a half years time), on which the Redistribution Committee based its proposal was 93882 electors, with the acceptable population range therefore being from 97168 to 90596 electors. (See Appendix C for the complete summary statistical report outlining the current and projected population totals and population deviations for the federal redistribution of Victoria in 2002.)

Computerization of Redistribution Process

The basic building blocks used to draw electoral districts are referred to as Census Collection Districts (CCDs). These small geographical units are used by the Australian Bureau of Statistics in the collection of the five yearly census. Each unit contains detailed demographic and electoral data such as population, elector population and projected population.

The geographic boundaries for the CCDs, as well as the associated demographic and electoral data, are computerized and available for the Redistribution Committee to use for redistribution. The manual manipulation and amalgamation of these units within the context of GIS has been named “passive redistributing”. Using GIS has shortened the timeframe for redistributions considerably throughout the years.

Conclusion

The assumption underpinning the current Australian redistribution process, spelled out in the Commonwealth Electoral Act 1918 and substantially implemented in 1984, is that the legitimacy of the process can be guaranteed by:

  • Timing of redistributions: Provisions are made for redistributions to be conducted with sufficient frequency to limit the development of malapportionment. In addition, the timing is specified by law and cannot be manipulate for political advantage.
  • Independent redistribution authority: Redistributions are undertaken by politically neutral and independent bodies.
  • Public input: The redistribution process is very public and the views of interested individuals and organizations must be taken into account
  • Established criteria for drawing boundaries: Redistribution Committees are required to work in accordance with a specific set of well-defined criteria which are broadly supported across the political spectrum
  • Automatic implementation of redistributions: The final redistribution proposal is not subject to veto at the political level, or by Parliament.

According to Maley, Morling and Bell, “the emphasis in the Australian scheme is very much on the legitimacy of [the redistribution] processes rather than specific outcomes, it being generally believed that if the mechanisms leading to a specific redistribution are acceptable, then the redistribution itself will be accepted, even by participants in the electoral process who see themselves as being disadvantaged by it.” [6] In fact, federal redistributions in Australia have largely ceased to be a matter for partisan debate.

Appendix A: Dates of Redistribution since 1901

 

 

 

NSW

VIC

QLD

WA

SA

TAS

ACT

NT

1900

11 Dec

26 Sep

04 Dec

05 Dec

 

 

 

 

 

1903

 

 

 

 

02 Oct

02 Oct

 

 

 

1906

13 Jul

13 Jul

13 Jul

13 Jul

 

 

 

 

 

1913

27 Feb

01 Feb

01 Feb

01 Feb

 

 

 

 

 

1922

13 Sep

04 Oct

13 Sep

13 Sep

13 Sep

13 Sep

 

 

 

1934

01 Aug

 

01 Aug

 

01 Aug

 

 

 

 

1937

 

21 Jul

 

28 Jul

 

 

 

 

 

1949

11 May

11 May

11 May

11 May

11 May

11 May

 

 

 

1955

30 Aug

10 Aug

10 Aug

10 Aug

19 Oct

30 Aug

 

 

 

1968

21 Nov

21 Nov

 

21 Nov

21 Nov

21 Nov

 

 

 

1969

 

 

27 Feb

 

 

 

 

 

 

1974

 

 

 

19 Apr

 

 

19 Apr

 

 

1977

31 Oct

31 Oct

31 Oct

07 Nov

31 Oct

31 Oct

 

 

 

1980

 

 

 

28 Feb

 

 

 

 

 

1984

11 Oct

14 Sep

13 Sep

31 Aug

03 Sep

12 Sep

23 Aug

 

 

1989

 

05 Jun

 

31 Mar

 

 

 

 

 

1992

31 Jan

 

28 Jan

 

17 Jan

01 Apr

23 Mar

 

 

1994

 

20 Dec

01 Dec

 

 

 

30 Sep

 

 

1997

 

 

10 Dec

06 Mar

 

 

10 Dec

 

 

1999

 

 

 

 

13 Aug

 

 

 

 

2000

11 Feb

 

 

20 Nov

 

11 Feb

 

 

21 Dec

2003

 

29 Jan

 

 

 

 

 

 

19 Feb*

NT did not undergo a redistribution, but reverted to a single division as a result of the determination of entitlement made on 19 February 2003.

Appendix B: Redistribution Timetable 

Australian Electoral Commission directs the commencement of the Redistribution and quota of electors is determined;

Electoral Commissioner invites written suggestions from the public;

Appointment of Redistribution Committee

(As soon as practicable)

↓ 

Public suggestions

(Closing date for suggestions – 30 days)

↓ 

Suggestions available for public comments

(Closing date for written comments – 14 days)

↓ 

Redistribution Committee considers suggestions and comments and develops a set of boundary proposals

(No time specified)

↓ 

Redistribution Committee publishes and exhibits maps showing proposed boundaries and names of divisions

(No time specified)

↓ 

Public objections to proposals

(Closing date for written objections – 28 days)

↓ 

Objections available for public comments

(Closing date for written comments – 14 days)

↓ 

Augmented Electoral Commission considers objections and makes final proposal

(open hearings)

(60 days)

↓ 

Final determination

(As soon as practicable)

 

 

img_au3

img_au4

img_au5

Notes:

[1] The Australian parliament is bicameral, with the Senate elected using a proportional representation system and the House of Representatives elected from single member districts. To ensure equal representation, the boundaries of these divisions must be redrawn periodically. Australia is not redistributed as a single entity; redistributions are undertaken separately for each State and Territory.

[2] The Commonwealth of Australia is composed of six States (New South Wales, Victoria, Queensland, Western Australia, South Australia and Tasmania) and two Territories, the Northern Territory and the Australian Capital Territory (which was formed as Australia’s seat of Parliament).

[3] While the Commonwealth Electoral Act 1918 makes provisions for a redistribution to be made even in cases where the augmented Electoral Commission is not unanimous in its view – four votes out for six, including at least two from members of the Australian Electoral Commission, suffice to carry the day – the lodging of a dissent by a member of a Redistribution Committee or augmented Electoral Commission has been a rare event.

[4] The Commonwealth Electoral Act 1918 was amended in 1983 to ensure that the final determination was not subject to veto by the Parliament. Prior to this date the redistribution was subject to approval by the House of Representatives, who on a number of occasions either delayed or rejected the boundary changes.

[5] Population projections are calculated using a demographic algorithm based on the widely accepted cohort-component method. This involves applying fertility, mortality and interstate migration rates to the base population to produce a cohort population 3.5 years from the assigned date.

[6] Michael Maley, Trevor Morling and Robin Bell, “Alternative ways of redistricting with single-member seats: the case of Australia,” Fixing the Boundaries: Defining and Redefining Single-Member Electoral Districts, edited by Iain McLean and David Butler, Dartmouth Publishing Company: England, 1996, p. 120.

Canada: Representation in Parliament

Two principles underlie the Canadian system of representation. The first is the principle of territorial representation: each elector is represented in the House of Commons on a territorial basis – the electoral district. The second is the principle of "one elector, one vote" – all voters are equal. This article discusses each of these fundamental aspects of the federal electoral system: the principle of representation in the House of Commons, that is, how the seats in the House of Commons are divided among the ten provinces and the territories; and how the electoral district boundaries are determined and periodically readjusted to reflect shifts in the population.

Representation

One of the crucial questions faced by the Fathers of Confederation in 1867 was how to ensure that all founding provinces were equally represented in the House of Commons of Canada, while at the same time guaranteeing that each region of the country had a fair say in the daily workings of the new federation. They adopted as a basic working principle the idea of "representation by population," and determined that each province was to be allotted a number of seats in the House of Commons that directly corresponded to its proportion of the total population in relation to the population of Quebec.

From the start, however, the Fathers of Confederation recognized the geographical, cultural, political and demographic diversity of the new provinces, as well as population size and rural and urban characteristics. As more provinces entered Confederation and as some regions grew and developed more than others, the diversity became more pronounced and a certain degree of compromise had to be built into the formula for allocating seats to the provinces. As a result, the basic principle of representation by population began to evolve and the allocation formula for seats has changed several times since Confederation to reflect this evolution.

In December 2011, Parliament passed the Fair Representation Act to address the under-representation in Canada’s fastest growing provinces, Ontario, British Columbia and Alberta.  The major provisions of the Act were to update the seat allocation formula (adopted in 1985) used for redistribution as well as update several provisions in the Electoral Boundaries Readjustment Act.

In 1985, the House of Commons had 282 seats. The 1985 formula, divided the total population of the ten provinces by a fixed denominator (279 seats which excluded the 3 seats allocated to Canada’s three northern territories) to obtain the electoral quotient, which was subsequently used to determine the number of seats for each province. 

The "senatorial clause," first adopted in 1915, states that a province cannot have fewer seats in the House of Commons than it does in the Senate. The "grandfather clause" (Representation Act, 1985) guarantees each province no fewer seats than it had in 1976 or during the 33rd Parliament.

The effect of the last provision was to add on extra seats to the House of Commons: by the end of 1992 process, the size of the House of Commons was 301, rather than 282 due to “add-on” seats. The redistribution based on the 2001 census saw the House of Commons at 308 seats, rather than 281, again due to ”add-on” seats. For both these redistributions, only Alberta, British Columbia and Ontario had the requisite number of seats according to their population; all the other provinces were given more seats than a strict application of the population equality criterion would suggest they were entitled. While the fixed denominator slowed the rate of growth in the House of Commons, it did so at the expense of provinces with faster-growing populations. The updated allocation formula presented in the Fair Representation Act will move every province towards representation by population. The Act introduces an updated seat allocation formula as follows:  

  • The electoral quotient for the 2011 readjustment is set at 111, 166, reflecting the average riding population prior to the last seat readjustment in 2001, increased by the simple average of provincial population growth rates.
  • The Senate floor and grandfather clause continue to apply.
  • A representation rule applies such that if a currently over represented province becomes underrepresented as a result of the application of the updated formula, additional seats would be allocated to that province so that its proportional representation according to population is protected.
  • For the purposes of calculating the provincial seat allocation, provincial populations are based on Statistics Canada’s annual population estimates.  These estimates correct for net undercoverage in the census and provide a more accurate representation of total provincial population.

Readjustment following the 2011 Census:

  • Step 1: Determine initial provincial seat allocation Provincial population / Electoral quotient (111, 166) = Provincial initial seat allocation
  • Step 2: Apply minimum seat guarantees
    Provincial initial seat allocation + Senate floor & Grandfather floor = Provincial second seat allocation
  • Step 3: Apply representation rule
    Provincial second seat allocation + Representation rule = Total provincial seats 
  • Step 4: Add territorial seats
    Total provincial seats + One seat per territory = Total seats

As set out below, for the 2021 readjustment and each subsequent readjustment, the electoral quotient will be increased by the simple average of provincial population growth rates since the preceding readjustment.

Readjustments following the 2021 Census and onward:

  • Step 1: Determine initial provincial seat allocation
    Population of the provinces / (Electoral quotient of last readjustment) X (provincial population average growth rates since last readjustment) = Electoral quotient
    Provincial population / Electoral quotient = Provincial initial seat allocation
  • Step 2: Apply minimum seat guarantees
    Provincial initial seat allocation + Senate floor & Grandfather floor = Provincial second seat allocation 
  • Step 3: Apply representation rule
    Provincial second seat allocation + Representation rule = Total provincial seats 
  • Step 4: Add territorial seats
    Total provincial seats + One seat per territory = Total seats

In the table below, the representation formula based on the 2001 census is displayed.

Table: Representation Formula for Canadian Parliament Following the 2001 Census

 

 

Province/Territory

Seats

33rd

Parl.

Population (2001 census)

Divide by National Quotient (107,220)

Rounded Result

Additional Seats (Senate Clause)

Additional Seats (Grandfather Clause)

Total Seats

Newfoundland and Labrador

7

512930

4.784

5

1

1

7

Prince Edward Island

4

135294

1.262

1

3

0

4

Nova Scotia

11

908007

8.469

8

2

1

11

New Brunswick

10

729498

6.804

7

3

0

10

Quebec

75

7237479

67.501

68

0

7

75

Ontario

95

11410046

106.417

106

0

0

106

Manitoba

14

1119583

10.442

10

0

4

14

Saskatchewan

14

978933

9.130

9

0

5

14

Alberta

21

2974807

27.745

28

0

0

28

British Columbia

28

3907738

36.446

36

0

0

36

 

 

 

 

 

 

 

 

Provincial Total

279

29914315

 

 

 

 

305

 

 

 

 

 

 

 

 

Nunavut

 

26745

 

 

 

 

1

Northwest Territories

2

37360

 

 

 

 

1

Yukon Territory

1

28674

 

 

 

 

1

 

 

 

 

 

 

 

 

National Total

282

30007094

 

 

 

 

308

Readjusting the Electoral District Boundaries

After the number of seats to be assigned to each province is established according to the constitutional formula, the most difficult part of the exercise begins: dividing the country into electoral districts to be represented by elected Members of Parliament.

The whole exercise is most properly known as "readjustment of electoral district boundaries," but is often referred to as "redistribution" and sometimes, particularly in other countries, as "redistricting." While the Constitution Act, 1867, specifies that a readjustment take place after each 10-year census, the rules for actually carrying out this enormous task are laid down in the Electoral Boundaries Readjustment Act (EBRA) which first passed Parliament in 1964.

Prior to passage of EBRA, the House of Commons was responsible for fixing the boundaries of electoral districts, through a committee appointed especially for that purpose. There were no rules to guide the exercise and research done on this period reveals a considerable amount of "political" interference in the readjustment process. This political interference was often referred to as "gerrymandering," or the manipulation of riding (electoral district) boundaries so as to ensure, as far as possible, the re-election of members of the governing party.

The Establishment of Independent Commissions

In the early 1960s, it was decided to assign the responsibility for readjusting electoral district boundaries to independent commissions, one for each province. Legislation to this effect was passed in 1964. Each commission was to be chaired by a judge designated by the chief justice of the province and there were to be three other members. One of these members was the Representation Commissioner, a public servant who was to sit on every commission. Initially, the other two members were to be political appointees, one from the governing party and one from the official opposition party. However, objections from the other opposition parties led to the legislation being amended such that the Speaker of the House of Commons was responsible for appointing the two remaining members of each commission.

The post of Representation Commissioner was abolished in 1979 and most of the duties transferred to the Chief Electoral Officer of Canada. Now each province has a three-member boundaries commission, chaired by a judge or by a person resident in that province and appointed by the Chief Justice of that province, with two additional members appointed by the Speaker. As the Northwest Territories, Yukon and Nunavut each constitute only one electoral district, no electoral boundaries commissions are required.

The goal of the readjustment process to be completely free of any political association is reinforced by a provision in the Electoral Boundaries Readjustment Act, which specifies that no sitting member of the Senate or of a federal, provincial or territorial legislature can be appointed to a commission. In practice, many commission members, aside from the chairmen, are university professors or non-elected officials of legislative assemblies.

Public Participation

When the rules for readjusting the federal electoral boundaries were laid down in 1964, Members of Parliament realized that, for the process to be completely fair, it not only had to be free of any political association but it also had to provide an opportunity for voters to express their views. Consequently, each commission publishes its proposals in the newspapers (and on the web) and invites the public to hearings, which are held at several different locations chosen to encourage the participation of as many interested people as possible.

Members of Parliament (MPs) are not excluded from this process of public involvement. Indeed, it is recognized that MPs will invariably have strong views on both the names and boundaries of the proposed electoral districts. Therefore, MPs are not only allowed to appear before a commission at the public hearings, but the legislation also provides them with the opportunity to object to the proposals of any of the boundaries commissions. The commissions must consider these objections, but are not obliged to make any changes as a result. In all cases, the final decisions as to where the boundary lines will be fixed rest with the commissions.

Criteria: Where to Draw the Lines

After receiving maps and documentation on the relevant population data from the most recent decennial census from the Chief Electoral Officer of Canada, commissions have 10 months to make proposals, hold public hearings and finalize their report. Guidelines for this enormous task are found in the Electoral Boundaries Readjustment Act, 1985. A look at the guidelines makes it clear that the readjustment exercise is not simply a mathematical computation but, rather, a delicate balancing act that must take into account human interests as well as geographic characteristics. In the course of their work, the commissions receive technical and administrative assistance from the staff of the Chief Electoral Officer at Elections Canada.

The commissions are charged with dividing the territory assigned to them into a specified number of electoral districts, so that the population of each one will correspond "as closely as is reasonably possible" to the predetermined average (or "quotient"). But, in fixing the electoral district boundaries, they must also take into consideration "the community of interest or community of identity in or the historical pattern of an electoral district ... and a manageable geographic size for districts in sparsely populated, rural or northern regions... "

To accommodate these human and geographic factors, the commissions are allowed to deviate from the average population figure when setting their boundaries. While generally restricted to a tolerance of plus or minus 25 percent from the quotient, a commission may exceed this limit "in circumstances viewed by the commission as being extraordinary."

The Process of Readjustment

While the work of the commissions is a crucial part of the readjustment exercise and may take up to 10 months or more to complete it is, in fact, only one part of the exercise. The whole process may take two years or more, from the time the Chief Electoral Officer of Canada receives the census data from the Chief Statistician to the time at which the new boundaries can be used at a general election. The main stages of the federal readjustment process are as follows:

Establishment of Commissions The members of the boundaries commissions are selected and appointed by the Speaker of the House of Commons and the appropriate chief justice. Commissions are established no later than six months following the census, or within 60 days of the census results being released, whichever comes first.

Allocation of Seats Prior to the release of decennial census data, the Chief Statistician provides the Chief electoral officer with estimates of the population of Canada and the provinces.  This allows commissions to begin their preliminary work. After the decennial census, the Chief Statistician of Canada gives to the Chief Electoral Officer of Canada population data for each province and the territories broken down by electoral districts and enumeration areas. Using these figures and the representation formula discussed above, the Chief Electoral Officer for Canada calculates the number of seats to be allocated to each province and the territories and publishes the results in the Canada Gazette.

Public Hearings Newspaper advertisements are published containing maps of the proposed electoral boundaries, as well as the time(s) and location(s) of public hearings, at least 30 days before the first hearing is scheduled. A commission must hold at least one public hearing before completing its report. During the hearings, interested individuals, groups and MPs may appear to express their views on the commission's proposals, after notifying the commission in writing of their intention to do so.

Completion of Reports No later than ten months after receiving the population data, each commission must complete its report on the new electoral districts. The Chief Electoral Officer of Canada may grant an extension of up to 2 months when necessary.

Review of Reports in Parliament Each commission's report is sent through the Chief Electoral Officer of Canada to the Speaker of the House of Commons, who must ensure that it is tabled and referred to a committee designated to deal with electoral matters. Written objections, each signed by at least ten MPs, may be filed with the committee within 30 days of the tabling of a report. The committee has 30 days, longer if the House of Commons is not sitting, to discuss any objections to a report and return it to the Speaker of the House of Commons.

Finalizing of Boundaries The reports are then returned to the commissions, accompanied by the minutes of the House of Commons Committee. The commissions then decide whether to modify their reports.

Representation Order The Chief Electoral Officer of Canada drafts a document called a "representation order" describing and naming the electoral districts established by the commissions and sends the document to the Governor in Council (Cabinet). Within five days of receiving the draft representation order, the Governor in Council must publicly announce the new boundaries in a proclamation which must then be published in the Canada Gazette within another five days from that date. The new boundaries cannot be used at an election until at least 7 months have passed between the date the representation order was proclaimed and the date that Parliament is dissolved for a general election.

Conclusion

Since Confederation, both the formula for calculating representation in the House of Commons and the process for adjusting electoral district boundaries have changed. The greatest changes have taken place in recent decades: Since the 1940s, we have seen four fundamental changes to the representation formula and one major change in the boundary readjustment process. The readjustments that followed the censuses of 1941, 1961, 1971 and 1981 were all delayed while such changes were made. Following the 1991 census, delays were caused by constitutional deliberations that resulted in the Meech Lake Agreement and the Charlottetown Accord and a review of the Electoral Boundaries Readjustment Act in general. The latest changes (to the formula and to EBRA), however, were accomplished prior to the beginning of the 2011 redistribution exercise resulting in the process proceeding as scheduled.

The redistribution process will continue to evolve in Canada as the population grows and the demands on the electoral system change.

Democratic Republic of Congo: Determining How Districts Might be Delimited in a Post-Conflict Society

The Democratic Republic of the Congo (DRC) is an ethnically/tribally divided society that has recently emerged from years of war (although sporadic fighting continues). Organizing and conducting elections in the DRC that will be viewed by all of the major stakeholders as free and fair presents a major technical and logistical challenge to the United Nations (UN)[1]. One of the many problems facing the UN in the DRC is the decision as to whether to delimit districts for the election of Members of Parliament (MPs), and if so, how.

Background

Since 1997, the DRC has been divided by ethnic strife and war. The hostilities, precipitated by a massive flow of refugees from the fighting in Rwanda and Burundi, led to the toppling of former dictator Mobutu Sese Seko by rebel leader Laurent Kabila in May 1997. Laurent Kabila’s regime was subsequently challenged by a Rwanda and Uganda-backed rebellion in August 1998. Troops from Zimbabwe, Angola, Namibia, Chad and Sudan intervened to support the Kabila regime. A cease-fire agreement was signed in July 1999 by the DRC, Zimbabwe, Angola, Uganda, Namibia, Rwanda, and Congolese insurgent groups, but sporadic fighting continued.

Joseph Kabila, who succeeded his father when Laurent Kabila was assassinated in January 2001, persuaded occupying Rwanda forces to withdraw from eastern Congo in October 2002. Two months later, an agreement (Global and Inclusive Agreement, 17 December 2002) was signed by all remaining warring parties to end the fighting and set up a transitional government. Ugandan troops officially withdrew from the DRC in May 2003. Localised violence (particularly in the Great Lakes region) continues, however.

Divided Society Battling ethnic groups (Tutsi, Hutu, Lendu, Hema and other ethnic groups) in the eastern portion of the country, supported by military forces from neighbouring countries, initiated much of the current conflict. Although the divisions within the DRC are not based solely on ethnicity, the large number of ethnic groups [2] – and the competition among them for limited resources – has certainly served to fuel clashes within society.

Lack of Resources Despite the vast potential of natural resources and mineral wealth, the DRC is one of the poorest countries in the world, with a per capita income of about US$90 in 2002. This is the result of years of mismanagement, corruption and war. In addition, the country has a high illiteracy rate (according to 2003 estimates, 41.7% of the population has had no schooling at all and an additional 42.2% has had only primary schooling) and limited technical expertise to draw on.

Little Intact Infrastructure The DRC is an enormous country (2,345,410 square kilometres; 905,063 square miles) with an almost complete lack of infrastructure. The existing roads (relatively few in number) have been badly damaged and many have yet to be tested for landmines. Although the UN Mission in the DRC, MONUC (Mission des Nations Unies en République Démocratique du Congo), has all existing airstrips operational, planes can reach a very limited number of areas in the DRC. In addition, the communication system is inadequate – although some radio and television broadcast stations are operating in the DRC. Conducting elections under these conditions will be quite challenging.

Lack of Legal Framework There is currently no electoral law in place. Nor has the transition parliament adopted laws on such issues as decentralisation and nationality – issues that are clearly controversial but must be decided before an election can be held. The transitional parliament, appointed by the signatories to the 17 December 2002 agreement, must reach agreement on these issues and promulgate laws and a new constitution before elections can proceed beyond the formative planning stage.

Time Constraints The transitional constitution, adopted on 2 April 2003, is of limited duration. It expires 24 months (with 2 six month extensions possible) after the inauguration of the transitional government, which occurred 30 June 2003. Elections must therefore be held by July 2005 (or July 2006 at the latest, if the two six month extension options are exercised). This is a very brief time period in which to organise a host of elections (including a referendum, general elections and local elections), even in optimal conditions – which certainly do not exist in the DRC.

Deciding on an Electoral System for the DRC

In newly emerging and transitional democracies, especially those where society is divided along ethnic, regional, or other factional lines, political institutions – particularly the electoral system – are especially significant. Political institutions in such societies are the most prominent channel of communication between antagonistic groups; if these institutions exclude significant groups from the “table” then differences between these groups cannot be resolved through negotiation and mutual accommodation. This could lead to the resumption of civil war and to the breakdown of the fledgling system.

Comparative experience suggests that the most important electoral requirement for democratic transition, particularly in a divided post-conflict state like the DRC, is an electoral system that maximises inclusiveness and is clearly fair to all parties. This goal is best achieved by a proportional representation (PR) electoral system, usually in conjunction with some form of overall power-sharing agreement within the government.

Given that some form of proportional representation is clearly in the best interest of the DRC, two obvious election system options exist. There are advantages and disadvantages associated with both of these systems which need to be considered when deciding which system to adopt in the DRC.

List Proportional Representation (PR) There is historical precedence for using List PR in the DRC: legislative elections held immediately following independence utilised this type of system (this was the system bequeathed to the DRC by the Belgium colonialists). Furthermore, List PR is the most common choice for newly emerging and transitional democracies so there is a wealth of experience to draw on for planning and conducting the elections.

Another advantage offered by List PR is that there would be no need to draw new electoral boundaries, although a decision would be required as to which set of existing administrative boundaries to use – province, district or territory – for regional List PR (assuming a national List PR system is not adopted).

A major drawback to using List PR is the current multiplicity of political parties – if this remains a feature of the DRC landscape, ballots could be far too long (especially if an open, rather than closed, party list is adopted). Political parties (most of which in the DRC have an ethno-regional basis and no ideological foundation) will have to unite, form coalitions or disappear for a List PR ballot to be manageable [3].

Another important disadvantage associated with List PR is that the geographical areas from which representatives are elected are usually quite large; hence the link between voters and their representatives is not as strong as it would be under a system offering single-member or small multimember constituencies. Many Congolese officials interviewed simply assumed that representatives would be selected from the level of the territory [4] (this has been the practice for the past 25 years at least) and felt that having representatives associated with specific territories was beneficial as it facilitated communication between voters and the government.

Mixed Member Proportional (MMP) The major advantage offered by an MMP system is the clear geographic link that is established between a representative and his or her voters. This specific concern was raised numerous times by Congolese interviewed. Not only would a distinct connection between a representative and his or her constituency facilitate the exchange of information, it would also empower Congolese voters. With time, Congolese voters would come to recognise their ability to hold their representative accountable (returning their MPs to office if they performed well on their behalf and voting them out of the legislature if their MPs failed them).

One important disadvantage with this system is the need to delimit a large number of constituencies. This could be done by either adopting administrative units smaller than provinces or districts [5], or by drawing unique electoral constituency boundaries. If such existing administrative units as territories were to be used, the population data associated with these units are even more problematic than the data at the provincial and district level. (The issue of data reliability is discussed at greater length below in “Construction of a Database”.)

Another disadvantage sometimes associated with an MMP system is the complexity of the ballot (MMP systems often require that two votes be cast, one for a constituency representative and one for a political party). In fact, however, the ballot can be designed so that voters cast only one vote; this single vote is used both to elect a constituent representative and to designate a party preference [6]. (Exercising the one-vote option would also limit the number of political parties to a manageable level, at least in the long run.)

Boundary Delimitation Options

Regardless of the electoral system adopted, some decisions will have to be reached concerning the delimitation of electoral boundaries. Although the size of the geographic area encompassed by the electoral constituency will vary depending on the type of electoral system (for example, List PR could adopt provincial lines as the electoral constituencies, while an MMP system would require smaller constituencies), some delimitation will almost certainly be necessary [7]. This delimitation could be as simple as adopting existing administrative boundaries (provincial, district or territorial boundaries, for example) as electoral constituencies and then allocating parliamentary seats to these constituencies on the basis of population; or it could be as complex as drawing new electoral constituency lines specifically for election purposes.

The importance of the delimitation process (and the rules that bind it), varies depending on the type of electoral system. Because plurality and majority systems can produce disproportional election results, the structure and rules established for the process are quite important. Although somewhat less important in the context of proportional representation systems, it is still essential that the law specify the process by which electoral constituency delimitation should occur.

Three alternatives exist for delimiting electoral boundaries in the DRC:

  • Use existing administrative boundaries (for example, provincial, district or territory boundaries) for electoral purposes.
  • New administrative boundaries could conceivably be drawn (and may in fact be proposed as part of a decentralisation package being debated by the transitional government in the DRC) and these could be utilised for electoral purposes.
  • Electoral constituencies could be drawn that are unique (separate from the administrative structure).

Current Administrative Units Used as Electoral Constituencies in the DRC

The DRC is subdivided into several layers of administrative units, many of these in existence since before independence in 1960. The largest of these units are provinces, next in size are districts, then territories; the smallest administrative unit (for which data is collected) are collectivities. According to the Administrative Divisions Directorate of the Ministry of Interior, the present number of administrative units for each layer is as follows [8]:

Provinces (including Kinshasa) 11 Districts 27 Cities (namely ±98 communes, ±1,249 districts and Kinshasa) 29 Territories 145 Communities (spread out over 476 sectors and 261 chiefdoms) 737 Groups (spread out over ± 60,000 villages) 5409

In addition, six urban groups having more than 100,000 inhabitants which do not have a “city status” have been identified. They are spread out in the provinces of Katanga (Kalemi, Kamina and Kipush), Nord-Kivu (Kayna and Kanya-Bayounga in a single group), Orientale (Bunia) and Sud-Kivu (Uvira).

These administrative units are interlinked and used not only for government administration purposes but also to conduct scientific and administrative censuses and to determine demographic projections. They also are employed to carry out various operations at the national, provincial, regional and local levels such as medical immunization campaigns.

Maps identifying the boundaries of these administrative units are readily available, but are out-of-date because centres of population have moved. (The administrative boundaries themselves have not changed, at least not recently, but because of massive population movements as a result of the war, administrative boundaries may cut through the middle of new population centres).

Population projections for these administrative units exist, although these projections are not particularly reliable. (This issue is discussed at greater length in this portion of the study, in a section entitled “Construction of a Database.”)

Co-opting a layer of administrative units (for example, provinces, districts, or territories) for use as electoral constituencies has several advantages:

  • Using already existing boundaries would negate the need to draw an entirely new set of electoral boundaries (which would be an extremely expensive and time consuming task).
  • There is population projection data (albeit less than reliable) associated with these existing administrative units, making the exercise of allocating seats to constituencies easier, and perhaps more accurate, than would otherwise be the case.
  • These administrative units have traditionally been used in the DRC for representational purposes (and many Congolese officials interviewed assumed that these units would be used again for the upcoming elections).

On the other hand, there are several disadvantages associated with the use of existing administrative boundaries:

  • Administrative units were not designed to encompass communities of interest, and in fact often cut across tribal lines and divide homogeneous groups that should be united in a single electoral constituency.
  • Some changes made to administrative boundaries (particularly at the territorial level) over time have been prompted by the desire to divide certain ethnic groups [9].
  • The population projections that do exist for these administrative units are clearly not reliable, especially given the war and other unanticipated events (i.e., the AIDS epidemic).

If the decision is reached to use current administrative units for electoral purposes, then the question remains as to which set of units to employ for electoral purposes: provinces, districts, territories, or some smaller unit. Of course, a large part of this decision is dependent on the type of electoral system adopted. For example, if a List PR system is selected, then the choice of administrative units is limited to provinces, districts or possibly territories since the electoral constituencies must be large enough to permit the allocation of several seats to each constituency. On the other hand, if an MMP system is adopted, then electoral constituencies would need to be much smaller in size – territorial units would probably be the largest possible unit that could be employed.

A series of simulations were performed for illustrative purposes only (using the 1984 census data and employing no projections or adjustments to this data) to determine what the seat allocation would be to each electoral constituency under three scenarios: using provinces as constituencies, using districts as constituencies, and using territories as constituencies. The Appendix contains the detailed results of these simulations.

If provinces were to be used for electoral constituencies (this would only be possible under List PR or for the PR portion of an MMP system), for example, the range in the number of seats allocated to the provinces would be from 11 (Maniema province) to 58 (Orientale province) using the 1984 census data and hypothesising a 400 seat parliament.

If electoral constituency boundaries were to coincide with administrative district boundaries, and no districts were combined, the range in seats allocated would be as few as 0 or 1 (district of Bandudu in the province of Bandudu) to as many as 26 or 27 (district of Kwilu in the province of Bandudu).

If territories were used as electoral constituencies, a great many territories would have to be combined in order to be accorded representation. This is true whether approximately half of the representatives were to be assigned to electoral constituencies (as would be the case with an MMP system) or all of the representatives were constituency-based (as would be the case with any plurality-majority system). On the other hand, some territories would form rather large multimember constituencies, with as many as 8 MPs assigned to a single constituency.

The simulations in the Appendix illustrate a number of important points:

  • The choice of which set of administrative units to utilize as electoral constituencies will have significant ramifications for the electoral system (how proportional the election results are likely to be), for the representation of communities (which communities would be assigned more and which less representatives, and which communities will have to be combined with others for representational purposes; and how many communities would be divided by electoral constituency lines), and for the electorate (how complex the ballot will be given the number of candidates competing and the number of seats to be filled).
  • Some boundary “delimitation” would likely have to be engaged in, at least to the extent of deciding which administrative units to combine for electoral constituency purposes – at least if territories (and possibly districts) are chosen as electoral constituencies. If a level below the territory is selected, then most certainly “delimitation” will have to occur.
  • The choice of a formula for allocating seats to electoral constituencies matters – an electoral constituency can receive more or less seats depending on what formula is used (i.e., compare our simple example of assigning seats when more than .51% of 74122 voters are assigned a seat, or when a seat is assigned for each 74122 voters)[10].
  • The population data upon with the seat allocation will be based is important. Obviously the more reliable and less controversial the data, the better: the process will be more accurate, and will be perceived as more legitimate, if the population data on which the allocation is based is not in question.

New Administrative Units Used as Electoral Constituencies

In the event of a decentralization of the government into administrative entities other than the existing ones, new administrative units would have to be created. If these new administrative units are also to be utilized as electoral constituencies, then the criteria for the effective representation of electors should prevail during the administrative delimitation process.

Of course, even if these electoral criteria are not considered when the new administrative units are drawn, it is still possible to use the new administrative units as electoral constituencies. However, the same disadvantages would hold for the new administrative units as for the current administrative units (see list above), with the additional problems of having even less reliable population data for these new administrative units, and less time to prepare for the elections.

If, on the other hand, the new administrative units are designed with such criteria as population equality and communities of interest in mind, then the adoption of these units as electoral constituencies would make a great deal of sense. This approach would require the collection of new demographic and sociological data, however – an expensive and time consuming operation. (This issue is discussed at greater length in the section below entitled “Conduct a New Census.”)

Delimitation of a Set of Unique Electoral Constituencies in the DRC

The delimitation of constituencies in the DRC specifically for the purposes of the election of representatives to parliament would be an enormous undertaking (both in terms of the time needed and the resources required). The process would involve a number of steps, including (1) the construction of a database minimally composed of maps and population data; (2) the formation of constituencies by allocating parliamentary seats to sub-regions of the country and creating unique electoral constituencies within these sub-regions; (3) the evaluation of the proposed redistricting plan and the adoption of a final redistricting plan.

Construction of a Database Delimitation requires the collection of several different types of information. The two essential pieces of information are population data and maps. The population data, which may be in the form of census enumeration data or voter registration data, provide the only means of creating constituencies that are relatively equal in population. Maps are needed to ensure that only contiguous geographic population units are assigned to constituencies and that constituency boundaries do not divide communities of interest unnecessarily.

Possible sources of delimitation data in the DRC include:

  • Using the currently existing census projections (based on 1984 census) and cartographic information
  • Updating and consolidating the census projections and cartographic information using local expertise and technical assistance
  • Conducting a new census (or, alternatively, a “light” census)
  • Using information obtained from voter registration process to update currently existing data
  •  

Using Currently Existing Population and Cartographic Data Much of the data needed for delimitation purposes in the DRC is out-of-date and, because of the drastic changes the DRC has undergone in the past ten years, unreliable.

The last scientific census of the population in the DRC was held in 1984, whereas the most recent administrative census dates back to 1996. Since then, the National Statistics Institute (INS) has produced demographic projections that estimate the number of electors aged 16 and over at ±25,600,000. These data on electors are broken down, by province, as follows:

Bandundu 2,915,000 Bas-Congo 1,555,000 Équateur 2,807,000 Katanga 3,429,000 Kasai Occidental 1,876,000 Kasai Oriental 2,160,000 Maniema 762,000 Nord-Kivu 1,982,000 Orientale 3,447,000 Sud-Kivu 1,784,000 City of Kinshasa 2,902,000 TOTAL 25,619,000

Although these population projections are estimated to be marginally reliable at the national and provincial level, they are deemed to far less reliable for the lower administrative levels (districts, territories, collectivities, etc.). Population projections in the DRC are problematic in large part because of the prolonged civil conflict and the displacement of the population and higher than presumed mortality rates that the war entailed [11].

The cartographic data currently available for the DRC suffers from the same defect: much of it is out-of-date because of large population shifts, particularly in the last five years. Although the geographical coverage index of the territory of the DRC, available to the specialists of the Geographical Information Centre of MONUC, is diversified, some of this information has not been updated for twenty years.

Updating and Consolidating Census Projections and Cartographic Data MONUC is equipped with a spatial reference geographical information system which, based on the information compiled and entered, can locate demographical and geographical elements, establish their exact number and distribute them in space. This system specifies the administrative boundaries to the smallest territorial entity; it also makes it possible to pinpoint hydrographical, rail, road networks, etc. This structured set of data makes it possible, at least in theory, to use a geographic information system to delimit electoral boundaries. However, the demographic and sociological data necessary to use this system for redistricting purposes is missing.

The main difficulty confronting MONUC specialists when maximizing the operation of the spatial reference geographical information system lies in the demographic and sociological data that are missing or incomplete or whose reliability must be validated. The collaboration of experts from the DRC would allow database specialists to obtain, verify and enter the required information.

Conducting a New Census The third option available for obtaining data for delimitation is to conduct a new scientific population census. However, this census operation would be quite costly and would require a minimum time period of two to three years to complete. Of course, the information obtained would be useful for endeavours far beyond the delimitation of electoral boundaries.

An enormous organizational and operational structure would have to be put in place to accomplish this task – the National Statistical Institute is ill-equipped to undertake a substantial statistical exercise at this point in time. For example, during the census held in 1984, use was made of 28,185 census-taking areas and 2,924 control areas spread out over the entire territory. In the event of a new census, these census-taking and control areas will first have to be checked and adjusted with respect to both the geographic territory encompassed and the number of anticipated respondents within them. The massive population movements in recent years, whether towards cities or elsewhere, require such a prior intervention.

Using Voter Registration Data The fourth option available is to use the information obtained during the voter registration process (assuming a voter registration procedure is conducted) to delimit electoral constituencies. A voter registration exercise designed to reach every household in the country could be used not only for a head count but could provide a geographical location for every potential voter in the country – invaluable information in a delimitation exercise.

The major drawback to using voter registration data is that much of the demographic and sociological data collected in a census would be missing from a voter registration database. Another problem is that the collection of this data would be completed rather late in terms of the election calendar, making delimitation on the basis of this data a challenge. Even if delimitation were to occur prior to the completion of the registration process, however, the voter registration counts could still be used to modify seat allocations if the need were to arise.

Formation of Electoral Constituencies Once a database has been prepared, the next step in the delimitation process is the formation of electoral constituencies. This is usually composed of one or two phases: the allocation, or apportionment, of parliamentary seats to regional entities such as provinces (this process is also referred to as “redistribution” in many countries); and the delimitation of electoral constituency boundaries within these regions.

The apportionment phase of the delimitation process is usually relatively mechanical [12], with the number of seats assigned to each sub-region usually dependent on the relative population of that sub-region. In countries that do not delimit single-member or smaller multimember constituencies, apportionment may be the only step taken to equalize population across electoral constituencies.

In countries that do delimit smaller constituencies, the second phase of the process is the creation of new electoral constituencies within the sub-regions themselves. (In countries that do not allocate seats regionally, this is the only phase in the delimitation process.) This is the step where the line drawers create a redistricting plan by assigning geographic units such as cities, towns and villages (or city blocks) to constituencies. A redistricting plan is complete when all geographic units within the given territory are assigned to a constituency and all constituencies in the plan meet the predetermined redistricting criteria.

Evaluation of Redistricting Plan Once the boundary authority has successfully completed a redistricting plan by assigning all geographic units in the territory to an electoral constituency, summary information for the plan should be produced. This information is used to evaluate the plan. A summary description of a redistricting plan should include information such as a description of the plan listing the geographic components of each constituency, map(s) of the plan showing the constituency boundaries, and a report summarising the most relevant statistical information for each constituency in the plan.

The summary information should allow the boundary authority, political parties, legislators and governmental officials, citizens, and other interested stakeholders to evaluate a redistricting plan according to the established criteria. Public hearing may be held to solicit the comments of these stakeholders. If, for example, the redistricting criteria adopted specify that constituencies be as equal in population as possible, information should be available regarding the population of each constituency the degree to which the population deviates from the electoral quota. The production of maps would allow interested parties to determine if communities of interest have been taken into account in promulgating the constituency boundaries.

After evaluating a proposed redistricting plan, including the solicitation of comments on this plan, the authority in charge of delimitation should endeavour to take these comments into account, and modify the redistricting plan accordingly. The final stage of the process is the adoption of the new redistricting plan; provisions for how this is accomplished should be described quite explicitly in the electoral law. In fact, the entire process (who should draw the constituencies, what criteria should be followed, etc.) should be mapped out as clearly as possible beforehand in the Election Act in order to guide authorities in charge of the process.

Problems with Delimiting A Unique Set of Constituencies The delimitation of constituencies in the DRC specifically for the purposes of the election of representatives to parliament would be an enormous undertaking both in terms of the time needed and the resources required. In fact, it is not likely to be technically feasible, given the lack of data and the current time constraints, to delimit a unique set of constituencies for the 2005 parliamentary elections. Furthermore, the delimitation of unique electoral districts – especially single-member constituencies – could well prove a political nightmare and is therefore not recommended for the DRC.

Conclusion

Only proportional representation (PR) electoral systems were discussed as viable options for the upcoming transitional elections in the DRC. This is because an essential condition for democratic consolidation in deeply divided, post-conflict countries such as the DRC is the inclusion of as many significant groups as possible, as proportionally as possible, in the parliament. A winner-take-all system that over-represents one of the political parties to the disadvantage of the others could only lead to more discord.

Regardless of what form of PR is adopted – regional List PR system or an MMP system – some delimitation of constituencies is likely to be required. Given the current status of population data in the DRC, it would be wise to consolidate and update census projections and maps for the currently existing administrative units (at least down to the territorial level) and use one of these sets of administrative units as constituencies for the 2005 parliamentary elections.

Germany: Delimiting Districts in a Mixed Member Proportional Electoral System

After the Second World War, the new Federal Republic of Germany adopted a mixed electoral system that combined party list proportional representation with single-member district representation.  Although mixed systems are becoming increasingly popular now, the German system was unique when first employed. 

Because mixed systems incorporate single-member districts, delimitation must occur periodically in order to ensure that electoral constituencies are relatively equal in population.  The importance of the delimitation process and the influence that constituency configurations have on the outcome of elections depends on whether the party list seats are used to correct any distortions in the relationship between votes cast and seats won in the single-member districts. 

In Germany, seats allocated under the party list component of the system are used to compensate for any distortions in the seats-to-votes ratio produced at the electoral district level.[1] This type of electoral system, often referred to as a Mixed Member Proportional (MMP) system, has been adopted by a number of countries in recent years (e.g., New Zealand, Hungary, Italy, Venezuela, Bolivia).  

In other countries with mixed systems, the party list seats are not used to compensate for any disproportionality arising from elections within the single-member districts.  Instead, seats allocated to the parties under the party list component of the election are simply added to the seats won at the electoral district level.  The partisan seats-to-votes ratio may, therefore, be distorted.  In this type of mixed system, often called a "parallel" system,[2]  the district delimitation process is more important because it can have a more pronounced effect on the partisan composition of the legislature.  

Even though Germany is an MMP system, the delimitation process does have ramifications for the outcome of elections.  This is because the allocation of compensatory seats (party list seats allocated to a party to compensate for any distortions in the seats-to-votes ratio) is done at the state (Länder) level rather than the national level.  Since German states vary in size and in political alignments, electoral districts that are malapportioned may lead to so-called overhang mandates (Überhangmandate), with strong parties in such states carrying more direct seats than have been allocated for that state.

Electoral System

The Parliament of the Federal Republic of Germany is bicameral: members of the lower house, the Bundestag, are directly elected; members of the upper house, the Bundesrat, are composed of representatives appointed by the states (Länder).

The composition of the Bundestag is determined by the MMP electoral system, combining elements of the single-member constituency plurality system with List Proportional Representation (PR).  Under this system, half the Bundestag members are directly elected from single-member constituencies (Wahlkreisen); the other half are elected by party list.

Constituency seats are filled via First-Past-the-Post (FPTP), under which the candidate obtaining the largest number of votes in each constituency (even if the candidate’s percentage constitutes only a plurality of the votes cast) is elected.  Between 1957 and 1987, there were 248 of these constituencies; from 1990 to 1998 there were 328; and in 2002 the number was reduced to 299.  These constituencies are allocated among the Länder in proportion to the relative populations of the Länder

The party list seats are distributed based on a party’s percentage of the popular vote.  For example, if a party wins 15 percent of the popular vote, it receives 15 percent of the seats in the Bundestag.

Each voter casts two ballots in a Bundestag election.  The first vote (Erststimme) is cast for an individual candidate running to represent a particular electoral district.  The second ballot (Zweitstimme) is cast for a political party, and it is this second vote that determines how many Bundestag seats each party will receive.[3]

To ensure that each party’s percentage of the combined district (first ballot) and party (second ballot) seats equals its share of the second vote, each party is allocated the requisite number of seats given its share of the total second vote.  The number of constituency seats won by each party is subtracted from the total number of seats allocated to that party, and the remaining seats are filled by candidates from the party list. 

If a party wins more constituency seats than it is entitled to, according to its share of the party vote, the party retains these seats (known as overhang mandates or Überhangmandate), and the size of the Bundestag is increased.  Every recent election has resulted in overhang mandates: after the 1990 election, the total number of seats in the Bundestag rose from 656 to 662; in 1994, another 16 seats were added for a total of 672 seats; in 1998 the election produced 13 overhang mandates for a total of 669 seats; and in 2002, despite the major re-delimitation to reduce overall seats to 598 and to avoid overhang mandates, the Bundestag was increased to 603 seats by 5 overhang mandates. 

Legal Framework for Delimitation 

At the beginning of each parliamentary term, the president of Germany – in  accordance with Article 3 of the German electoral law – nominates an independent standing Electoral Districts Commission (Wahlkreiskommission).  The task of the Commission is to report on population changes in electoral constituencies and to put forward recommendations on how to re-delimit constituencies, if necessary, to accommodate these changes. 

Boundary Authority     The independent Electoral Districts Commission (EDC) is established and selected by the German President at the beginning of each parliamentary term.  It consists of the President of the Federal Statistical Office, a judge from the Federal Court of Administration, and five additional members, usually high-level administrative state functionaries. 

Delimitation Criteria   The EDC is required to follow five rules (as established in Article 3 of the Federal Electoral Law, April 27, 2001) for electoral delimitation.  They are as follows: 

  • The Länder boundaries must be observed. 
  • The number of constituencies in the individual Länder must correspond to the population proportion as far as possible.
  • The population of a constituency should not deviate from the average population of the constituencies by more than 15 percent in either direction. Where the deviation is greater than 25 percent the boundaries shall be redrawn. 
  • Each constituency should form a coherent area. 
  • Where possible, the boundaries of the communes, districts, and urban districts should be respected. 

The Commission has 15 months to complete its report and present its findings to the Ministry of the Interior.  The report of the Commission contains the current population of the electoral districts and recommendations for reallocating district seats and modifying district boundaries. The EDC may put forward several alternative plans for redistricting so that Parliament has more than one option available when deciding how, if at all, to redraw the constituencies. 

Participation of Other Institutions   The EDC produces its report with input from a number of other institutions, including members of the Ministry of the Interior who take part in EDC meetings. The Commission is in contact with Länder governmental employees responsible for electoral procedures in all of the states.  Furthermore, each Länder has the opportunity to present its position on any of the Commission's recommendations. 

The Role of the Parliament    The Parliament makes the final decision as the whether to redistrict and what redistricting plan to adopt.  After reading the EDC report and debating the subject, Parliament votes on whether to accept any of the EDC recommendations.  Unless there are population deviations greater than 25 percent (previously 33 percent), Parliament is not required to accept any of the EDC proposals.  In fact, Parliament has frequently decided not to make any changes. 

Reforming the Redistricting Process in Germany 

The Commission Review that began in 1995 proposed major changes in the process for re-delimiting constituencies.  These changes were warranted for a number of reasons: 

  1. The allocation of districts had not changed since 1980 in the western Länder, and since 1990 in the eastern Länder, although major population changes had taken place.  Parliament not only rejected reallocation recommendations in 1983, 1987, and 1990 but also rejected the EDC constituency delimitation plan for a united Germany in 1994.  Parliament tended to approve only those EDC recommendations that were necessitated by law (that is, when the population variation was greater than 33 percent).
  2. The 1994 parliamentary elections produced 16 surplus seats (and the election in 1998 produced another 13 surplus seats).  There are several factors leading to surplus seats, but one of the major reasons is the unbalanced distribution of electoral districts due to substantial changes in populations across the Länder.
  3. As of the end of 1994, three electoral constituencies deviated by more than 33 percent from the average constituency size.  Several more were very close to this limit.
  4. County reforms initiated prior to 1994 in the new Länder also prompted a need to modify constituency boundaries.
  5. In addition, in June 1995, the Bundestag decided to reduce the size of the legislature by up to 100 seats starting with the 15th term (2002). 

Ultimately, a Reform Commission (Reformkommission zur Größe des Deutschen Bundestages) was established in September 1995 to make recommendations.  The Commission was composed of 20 members of the Bundestag (with the ruling CDU/CSU holding 11 seats on the Commission), and 14 experts (several of whom were former members of the Bundestag.) 

The task of the Reform Commission was to produce recommendations on all important questions relating to the reduction in size of the Bundestag, including the issues of delimitation of electoral districts and the overhang mandate rules, as well as a number of other related issues. 

On June 17, 1997, the President of the Reform Commission, Hans-Ulrich Klose, presented the final recommendations to the Bundestag. These recommendations led to the following changes in the electoral law related to constituency delimitation:

  • The number of Bundestag constituency seats was reduced from 328 to 299.
  • The number of electoral districts allocated to each state was required to be as proportional to each state’s relative population as possible. 
  • The permissible population deviation was reduced: previously population deviations of up to 25 percent were permitted, and it was not until the deviation reached 33 percent that the constituency boundaries were required by law to change.  The electoral law now requires that the population of a constituency not deviate from the average population by more than 15 percent, and if the deviation exceeds 25 percent, the constituency boundaries must be redrawn.

Seat Allocation by State: Before and After Bundestag Seat Reduction

The table below displays the allocation of constituency seats by state before and after the reduction of Bundestag constituency seats from 328 to 299.

States

2002

1998

Baden-Württemberg (Constituency 259 – 295)

37

37

Bayern (Constituency 215 – 258)

44

45

Berlin (Constituency 76 – 87)

12

13

Brandenburg (Constituency 56 – 65)

10

12

Bremen (Constituency 54 and 55)

  2

  3

Hamburg (Constituency 19 – 24)

  6

  7

Hessen (Constituency 169 – 189)

21

22

Mecklenburg-Vorpommern (Constituency 12 – 18)

  7

  9

Niedersachsen (Constituency 25 – 53) 

29

31

Nordrhein-Westfalen (Constituency 88 – 151)

64

71

Rheinland-Pfalz (Constituency 200 – 214)

15

16

Saarland (Constituency 296 – 299)

  4

  5

Sachsen (Constituency 152 – 168)

17

21

Sachsen-Anhalt (Constituency 66 – 75)

10

13

Schleswig-Holstein (Constituency 1 – 11)

11

11

Thüringen (Constituency 190 – 199)

10

12

History of Redistricting in Germany

The EDC makes recommendations with regard to redistricting approximately every four years.  The following indicates what actions the Parliament ultimately took when presented with EDC recommendations: 

  • First Report of the EDC in 1958: Parliament took no action. 
  • Report of 1962: Recommendations accepted by Parliament. 
  • Reports of 1966 and 1970: Parliament took limited action. 
  • Report of 1973: EDC recommended reallocation of seats to states, which was rejected by Parliament.  Redistricting within states was partially accepted by Parliament. 
  • Report of 1978: EDC recommendations partially accepted. 
  • Report of 1982: No major recommendation made by EDC because Parliament was dissolved.  There were two changes made due to population deviations in excess of 33 percent.
  •  Report of 1984: EDC recommendations to redistrict boundaries partially accepted.
  • Report of 1988: Minor changes only because census numbers were not yet available. 
  •  Report of 1992: EDC recommendations were not accepted, except for boundary changes required because of deviations in excess of 33 percent; major redistricting in Berlin. 
  • Report of 1996: Most of the 16 states did not accept changes proposed by EDC. Berlin was granted a new electoral district; Mecklenburg-Vorpommern lost a seat; and Lower Saxony gained a seat. There was some re-delimitation in Hessia due to deviations in excess of 25 percent. 
  • Report of 1999 and supplementary report of 2000: Major re-delimitation to reduce the number of electoral districts to 299. 
  • Report of 2003: Due to population shifts, the EDC proposed that Thüringen and Schleswig-Holstein each lose one seat, and that Bavaria and Niedersachsen each gain one seat. The Governments of Thüringen and Niedersachsen, however, rejected the proposal.

Conclusion

Delimitation in Germany is guided by firmly established legal requirements.  Under certain conditions, mandatory changes are required and there is very little leeway for political manoeuvring.  Because the EDC is an independent organ and there are set rules that it must follow when making recommendations for redistricting, gerrymandering is virtually impossible.   

However, Länder governments – particularly if they are due to lose a seat – do not relish a change in constituency boundaries.  Members of Parliament are also likely to object to changes in constituency lines: new constituency boundaries could make it more difficult for incumbent legislators to win and could lead to the loss of a party seat.  Therefore, Parliament often simply accepts only those changes that are mandated by law, retaining the status quo as much as possible.

Germany: Electoral District Populations and Population Deviations

[1] For example, if a political party were to win 55% of the total vote cast in a parliamentary election but win only 45% of the constituency seats, compensatory seats would be allocated to the party such that the percentage of seats held by that party would equal 55% of the parliamentary seats overall.

[2] Parallel systems are common among the former Soviet Republics.

[3] The political parties establish a party list for each Länder.

Fiji: Delimiting Communal Seats to Guarantee Ethnic Representation

The Pacific Island state of Fiji is an ethnically divided country with a history of political tension resulting in several coup d’etats. The two major ethnic groups are the indigenous Fijians and the Indo-Fijians. (The Indo-Fijians are primarily the descendants of indentured laborers brought by British colonists from India to work on sugar plantations during the 19th century.) These two groups are roughly equal in size – according to the 1996 census, 51.1 percent of the population is indigenous Fijian and 43.4 percent Indian.

Since independence in 1970, indigenous Fijians and Indo-Fijians have cast votes in separate communal constituencies, although under two of the three constitutions since 1970, “open” constituencies have also existed. Political parties tend to be ethnically identified and voting is, for the most part, ethnically-based. Because the population is approximately balanced, electoral victory tends to be dependent on the results in the open constituencies. The delimitation of constituency boundaries, especially open constituencies, is therefore of considerable significance in Fiji.

Historical Background

Following independence from Britain in 1970, Fiji appeared to make a relatively successful transition to multi-ethnic democracy. But in 1987, following the election of a government seen by the indigenous Fijian-dominated military as overly close to the Indo-Fijian community, two coups occurred. Eventually a new constitution, with provisions that weighed strongly in favor of the indigenous population, was promulgated.

The 1990 Constitution adopted a new electoral system based entirely on communal representation of ethnic groups, with separate electoral rolls for Fijians, Indians and “general electors”. The indigenous Fijian population was guaranteed primacy in most senior government and administrative positions, including the office of the prime minister; the Indo-Fijian population, on the other hand, was under-represented in government relative to its proportion of the population.

In 1994, following economic difficulties, international condemnation (including expulsion from the Commonwealth of Nations) and a high level of emigration by the Indian community, the Fiji government established a Constitution Review Commission to re-examine the constitution and recommend a less biased form of representation. The Commission ultimately recommended a new constitution containing an innovative package of electoral arrangements designed to encourage the development of multi-ethnic politics in Fiji. One of these recommendations was the adoption of an Alternative Vote System [1].

The 1997 Constitution as approved included the recommended preferential voting system. However, the constitution also retained communal seats: two-thirds of the seats in the new parliament were to be elected on a communal basis, the remaining one-third of the parliamentary seats were to be elected from an open electoral roll.

Elections under the new constitution were held in May 1999. These elections ended in a surprise victory for the Indo-Fijian party, the Fiji Labour Party (FLP), and Fiji’s first Indo-Fijian Prime Minister, Mahendra Chaudhry. One year later, on May 19, 2000, the Prime Minister and other high-ranking government officials were taken hostage in the parliament building by an armed group led by George Speight, a member of a radical group of indigenous Fijian nationalists. Chaudry’s government was dissolved, martial law was declared and the 1997 Constitution was abrogated.

In July 2000, the military transferred power to an interim administration (one that had not been elected) and indigenous Fijian Laisenia Qarase was sworn in as Prime Minster. Following significant pressure from the international community, the interim administration held elections in August 2001 under the provisions of the 1997 Constitution [2].

The electorate voted mainly along ethnic lines and the political party of the interim Prime Minister Qarase won the most seats (31) in the 71 seat parliament; the deposed Prime Minister’s Fiji Labour Party came in second with 27 seats. Mr. Qarase was sworn in as Fiji’s new Prime Minster on 10 September 2001. (The next election is to be held in September 2006.)

Electoral System

The bicameral parliament of Fiji is composed of an appointed Senate [3] and an elected House of Representatives. The 71 members of the House of Representatives are directly elected from single-member constituencies by a preferential voting system.

The 71 legislative constituencies are comprised of 46 “communal” constituencies and 25 “open” constituencies, with the members elected as follows: 

  • 46 members are elected by voters registered on one of 4 separate electoral rolls:
  • 23 are elected from a roll of voters who are registered as indigenous Fijians
  • 19 are elected from a roll of voters who are registered as Indians
  • 1 is elected from a roll of voters who are registered as Rotumans
  • 3 are elected from a roll of voters who are not registered as Fijians, Indians or Rotumans (this is the “general voters” roll)
  • 25 members are elected by voters from all communities registered on an open electoral roll

The outcome of an election depends heavily on the demographic distribution of ethnic groups and the way in which electoral boundaries – particularly the open seats – are drawn.

Legal Framework for Delimitation

Boundary Authority The 1997 Constitution dictates that delimitation (redistribution) be undertaken by a Constituency Boundaries Commission composed of three persons: 

  • A chairperson, who must be, or is qualified to be, a judge. The chairperson is appointed by the President, acting in his or her own judgment, following consultation with the Prime Minister and the Leader of the Opposition.
  • Two additional members, both appointed by the President, one on the nomination of the Prime Minister and the other on the nomination of the Leader of the Opposition. 

A person cannot be appointed as a member of the Commission if he or she is, or has been at any time during the past four years, a member of Parliament, or another representative body, or a member of a state service.

Final Authority Although the Commission must report its final determination to the House of Representatives, together with a summary of any objections made to it and the reasons for its final determination, the legislature does not vote on the determination and cannot veto it. Subject to the jurisdiction of a court to entertain an application for judicial review [4], the decision of the Commission is final.

Public Input into the Process Whenever the Constituency Boundaries Commission proposes to alter a boundary, it must publish a notice and invite submissions from political parties, members of the House of Representatives and any other person or body wishing to make a submission. The invitation for submissions must specify a period of not less than 60 days for the making of submissions.

Notice must be placed in the daily newspapers and played on local radio stations indicating where anyone interested may inspect maps showing the provisional boundaries and a summary of the reasons for the provisional boundaries. The notice must also indicate the last day on which the Constituency Boundaries Commission will receive written objections to the provisional boundaries.

If any objections are received, the Constituency Boundaries Commission must publish an announcement specifying the places where the objections are available for public inspection and the last day on which the Commission will receive counter-objections, being not less than 21 days after the date of publication of the notice.

After considering any objections and counter-objections, the Constituency Boundaries Commission makes a final determination and produces a report outlining the basis for its final determination and submits this report to Parliament.

Timing of Redistributions The Constituency Boundaries Commission must, in the year following each official census, review the boundaries and determine whether or not the boundaries should be changed to ensure compliance with established redistribution criteria.

Criteria for Redistributions The Constitution specifies a number of criteria the Commission must follow when drawing constituency boundaries, depending on the type of constituency [5].

In determining the boundaries of the 23 Fijian communal seats, the Commission must ensure that the boundaries for 17 of the constituencies are in accordance with the provincial boundaries, with the provinces of Ba, Tailevu and Cakaudrove comprising two constituencies each, and the other provinces each comprising one constituency. The remaining 6 Fijian constituencies are to be composed of predominately urban areas in which the number of voters is to be, as far as reasonably practicable, equal.

The boundaries for the remaining 23 communal seats (19 Indian, 1 Rotuman and 3 General seats) are to be drawn in such a manner that each seat has roughly the same number of voters within its own communal category.

The boundaries for the 25 open constituencies are to be drawn so that each constituency has roughly the same number of voters and “a good proportion of members of the different ethnic groups.”

In addition, the Commission must give due consideration to:

  • the constituency’s physical features
  • the boundaries of existing administrative and recognized traditional areas
  • means of communication and travel within the proposed constituency

1998 Delimitation Exercise

Both the 1999 and 2001 parliamentary elections were conducted using the constituency boundaries created by the 1998 Constituency Boundaries Commission [6].

The 1998 Commission divided the Fiji Islands into the requisite 71 constituencies. Of these, 25 constituencies are “open seats” where candidates and voters are from any ethnic group. The other 46 constituencies are communal seats in which a voter casts a vote only for candidates from his or her own ethnic community. In the Fijian communal constituencies, for example, only indigenous Fijians vote for indigenous Fijian candidates; in the Indo-Fijian constituencies, only Indo-Fijians vote for Indo-Fijian candidates. There are 23 communal seats for indigenous Fijians, 19 for Indo-Fijians, 1 for Rotumans and 3 for general electors (for those who are not on any of the other rolls).

Each elector appears on two rolls – one for the open seat and one for the communal, or racially-reserved, seat – and each voter casts two votes, one for an open seat representative and one for a communal seat representative. (There are actually two sets of constituency boundaries drawn for the country: a set of 46 communal constituencies that encompass the entire country and a set of 25 open constituencies that also encompass the entire country.)

Rural Fiji has high territorial segregation, and the outer islands are almost entirely indigenous Fijian. The main island and urban centers are more ethnically mixed and constituencies could be drawn in these areas that are, at least in part, more ethnically diverse.

Communal Constituencies According to the Constitution, the boundaries of most of the communal constituencies are to be drawn so that each constituency has roughly the same number of voters within its own communal category – for example, the 19 Indian constituencies should each have about the same number of Indo-Fijian voters. However, the Constitution places additional constraints on the 23 Fijian constituencies: the boundaries of 17 of the constituencies are to follow the provincial boundaries; the other six Fijian communal constituencies must be predominately urban.

As a result of these requirements, the populations of the constituencies vary quite dramatically. Table 1, below, lists the average number of voters for each type of communal seat, as well as for the open seats. (Appendix A  provides the populations of all 71 constituencies for both the 1999 election and the 2001 election.)

 Table 1: Average Population by Seat Type,1999 Electoral Constituencies[1]

Type of Seat

Number of Seats

Number of Voters

Average Number of Voters Per Seat

Fijian Reserved Seats

 

 

 

·  Provincial (Rural)

17

143889

8464

·  Urban

6

76375

12729

·  Total

23

220264

9577

Indian Reserved Seats

19

197621

10401

Rotuman Reserved Seat

1

5232

5232

General Reserved Seats

3

14029

4676

Total Reserved Seats

46

437146

9503

Open Seats

25

437146

17486

The population quota for the 46 communal seats was 9503 in 1999 [8], but the average number of voters per type of seat varied substantially from this quota. The provincial Fijian reserved seats were, on average, smaller than the population quota (the average number of voters in 1999 for these seats was 8464, but the actual range in seat population was from 2856 to 16051 voters), while the Fijian urban reserved seats were, on average, much larger than the population quota. The Indian reserved seats were also, on average, larger than the population quota. The general reserved seats, and the seat reserved for Rotumans, were, however, considerably smaller than the population quota.

The total percent population deviation for the 46 communal seats was very close to 140%. This is quite high; most consolidated democracies that have established tolerance limits for population deviations have set the limit at around plus/minus 10 percent, producing a total population deviation of no more than 20 percent. On the other hand, the total percent population deviation for the open seats was substantially smaller in 1999: only 29 percent.

Open Seats According to the 1997 Constitution, the Constituency Boundaries Commission must give due consideration to the principle that “voters should comprise a good portion of members of different ethnic communities.” However, the 1998 Constituency Boundaries Commission found that achieving ethnic parity in the 25 open constituencies was “impossible” and instead opted to “provide that the overall balance of the ethnic communities was maintained over the 25 open seats.” [9] This was done by distributing majority ethnic Fijian and majority Indo-Fijian constituencies roughly in proportion to their respective weights in the national population.

The Commission created 10 open seats in which Fijian voters composed over 55% of the total voters. Indian voters made up over 55% of the total voters in 8 open seats, and were a majority (between 50 and 55% of the voters) in an additional 3 seats. (Appendix B provides the ethnic composition – percent indigenous Fijian, Indian, Rotuman, and general – of the open constituencies when they were drawn.)

The Figure below, reprinted from an article written by Jon Fraenkel [10], illustrates the ethnic profile of the open seats:

 Reprint photo from Jon Fraenkel

Only a few of the open seats created by the Commission were truly ethnically mixed. These seats included:

 

 

  • Nausori / Naitasiri: 49% Fijian and 48% Indian
  • Nasinu / Naitasiri: 49% Fijian and 48% Indian
  • Laucala: 48% Fijian and 45% Indian
  • Samabula / Tamavua: 47% Fijian and 44% Indian
  • Suva City: 49% Fijian, 35% Indian and 11% General
  • Nadroga: 48% Fijian and 51% Indian

These ethnically mixed open seats proved pivotal in the 1999 and 2001 elections.

Results of the 1999 and 2001 Parliamentary Elections

Since independence in 1970, electoral contests in Fiji have been marked by ethnic conflict between indigenous Fijian-supported and Indian-supported political parties. The indigenous Fijian communal seats have consistently elected representatives from a separate set of parties as the Indian communal seats. As a consequence, the election has typically been decided by the open seats (when they have existed), and the most important open seats have been those with a heterogeneous population. The two most recent elections – 1999 and 2001 – are no exceptions to this rule.

1999 Parliamentary Elections Voting in 1999 was strongly along ethnic lines.

Table 2 displays the results of the election by type of seat. [11]

 

FLP

FAP

SVT

PANU

VLV

UGP

NVTLP

INDP

Open

18

2

3

0

0

1

0

1

Fijian

0

9

5

4

3

0

1

1

Indian

19

0

0

0

0

0

0

0

General

0

0

0

0

0

1

0

2

Rotuman

0

0

0

0

0

0

0

1

Total

37

11

8

4

3

2

1

5

FLP      Fiji Labour Party

FAP      Fijian Association Party

SVT      Soqosoqo ni Vakavulewa ni Taukei

PANU   Party of National Unity

VLV      Veitokani ni Lewenivanua Vakarisito

UGP     United General Party

NVTLP Nationalist Party                       

INDP    Independents

All 19 of the Indian reserved seats went to the Fiji Labour Party (FLP), but not a single one of the Fijian communal seats went to the FLP. Five major Fijian parties split among themselves 22 of the 23 reserved indigenous Fijian seats, but not one of these parties carried a single Indian seat. In fact, the largely Indian-supported parties (the FLP and the National Federal Party (NFP), which obtained a substantial portion of the Indian vote but no seats) obtained less than 2 percent of the indigenous Fijian vote, while the largely Fijian-supported parties received less than 1 percent of the Indian vote.

As Table 2 indicates, the Indian-supported FLP won the election with 37 seats, primarily because indigenous Fijian voters split their votes across five parties and the vast majority of open seats went to the FLP. The FLP not only carried the open seats in which Indian voters were a majority (which the FLP usually won on the first count), but also number of open seats in which the transfer of votes (Fiji has an Alternative Vote System) ultimately led to a victory for the FLP [12].

The Indian-backed FLP, after managing to secure an absolute majority of the seats, named their leader, Mahendra Chaudry, as the country’s first Indian Prime Minister. One year later, George Speight marched into Parliament and took the Prime Minister and most of his cabinet hostage. Although the coup was eventually defeated, a caretaker regime, composed largely of indigenous Fijians, was installed and new elections were called for August 2001.

2001 Parliamentary Elections The results of this election are listed in Table 3.

Table 3: Results of 2001 Parliamentary Election[13]

Type of Seat

Political Party

FLP

SDL

MV

NFP

NLUP

UGP

INDP

Open

8

13

1

1

1

0

1

Fijian

0

18

5

0

0

0

0

Indian

19

0

0

0

0

0

0

General

0

1

0

0

1

1

0

Rotuman

0

0

0

0

0

0

1

Total

27

32

6

1

2

1

2

Ethnically polarized voting in the 2001 election led to another divisive victory, but this time the winner was the indigenous Fijian-backed Soqosoqo ni Duavata ni Lewenivanua (SDL) party – a new party founded by a coalition of ministers serving in the post-coup caretaker regime – that emerged as the winner of the election.

As in 1999, no indigenous Fijian-supported party won a single Indian seat, and Indian-supported FLP did not win a single Fijian seat. In 1999, the victorious FLP received less than 2 percent of the indigenous Fijian vote; in 2001 the SDL received a mere .1% of the Indian vote.

All of the Indian communal seats went to the FLP. All of the Fijian communal seats, on the other hand, went to one of two indigenous Fijian-supported parties, the SDL or the Matanitu Vanua (MV). The open seats with large indigenous Fijian population also went to either the SDL or the MV. The other open constituencies obtained by the SDL were won on the basis of transfers of party preferences.

The 2001 election (as well as the 1999 election) produced very disproportionate results: the percentages of votes received by the parties were very different from their shares of total seats won. For example, the SDL won 27.5 percent of the vote, but received 45.1 percent of the seats. Table 4, below, reports the percentage of votes and seats won for the parties that actually secured seats.

 Table 4: Percentage of Votes and Seats Won,  2001 Parliamentary Election[14]

Party

Percent of Votes Won

Percent of Seats Won

FLP

26.5

38.0

SDL

27.5

45.1

MV

4.2

8.5

NLUP

1.3

2.8

UGP

.3

1.4

NFP

1.2

1.4

INDP

1.4

2.8

Total

62.4

100.0

 

Some commentators have argued that the reason the results were so disproportionate, and that the 1999 and 2001 elections were decided in the near-parity open seats, is that the proportion of open seats to communal seats is not high enough and the open seats are not well-designed:

“The way electoral districts were drawn…ensured that opportunities for genuine inter-ethnic cooperation were rare. Because only the 25 open electorates enabled multi-ethnic competition, and of these no more than eight were reasonably balanced in their mixture of indigenous Fijian and Indo-Fijian voters, the vast majority of electorate-level contests provided no opportunity at all for cross-ethnic campaigns, appeals or outcomes… The CRC’s recommendation for a “good” proportion of members of both major communities in all open seats was interpreted extremely loosely, to mean ethnic balances of up to 90:10 in some cases, which obviated the need for intra-communal vote swapping. In most seats, clear Indian or Fijian majorities prevailed.” [15]

Other writers (such as Jon Fraenkel in a series of articles discussing the Fiji electoral system) disagree with this assessment and contend that it is the Alternative Vote System that has failed. Regardless of which argument is correct, it is clear that the Fiji system has failed to foster the desired multi-ethnic cooperation.

Conclusion

Since 1970, when Fiji gained its independence, indigenous Fijians and Indo-Fijians have voted in separate ethnically based communal constituencies. Voting is ethnically polarized, and political parties tend to be aligned with one ethnic group or another. As a result, it is the votes cast in the open constituencies, where all voters – regardless of race or ethnicity – cast ballots, which have decided recent elections. This, and the fact that the electoral system depends on majoritarian voting in single-member constituencies that can distort the partisan votes to seats ratio, means that constituency boundaries have important implications in Fiji.

Notes:


[1] This table is based on data reported in Appendix A and obtained from the Fiji Elections Office.

 

[1] Under an Alternative Vote System, electors rank the candidates in order of choice. If no candidate has over 50 percent of the first-preferences, lower order preference votes are transferred until a majority winner emerges. (This system is used in Australia and some other South Pacific Island countries.)

[2] In November 2000, the High Court of Fiji ruled that the military’s abrogation of the 1997 Constitution was illegal.

[3] The 34 senate seats are appointed as follows: 24 are appointed by the Great Council of Chiefs, 9 are appointed by the president, and 1 is appointed by the council of Rotuma.

[4] Although the court can consider objections to a constituency plan, this has not happened to date – no one has brought a claim against a redistribution plan.

[5] The criteria are listed in Chapter 6, Part 2 (Article 52) of the 1997 Constitution.

[6] The next parliamentary elections are not scheduled until September 2006.

[7] This table is based on data reported in Appendix A and obtained from the Fiji Elections Office.

[8] The population quota is obtained by dividing the total population (437146) by the number of seats (46).

[9] Constituency Boundaries Commission, Final Report, September 1998.

[10] Jon Fraenkel, “The Alternative Vote System in Fiji: Electoral Engineering or Ballot-Rigging?” Journal of Commonwealth and Comparative Politics, volume 39 (2), July 2001, page 10.

[11] The data for this table was reported by the Fiji Elections Office.

[12] Had the parties supported by indigenous Fijians agreed to exchange preferences, the FLP would probably not have won nearly as many open seats. However, three ethnic Fijian parties (including the FAP and PANU) consistently put the FLP in second position in their preference lists rather than list each other.

[13] The data was reported by the Fiji Elections Office.

[14] The data for this table was reported by the Fiji Elections Office.

[15] Benjamin Reilly, “Evaluating the effect of the electoral system in post-coup Fiji,” Pacific Ecomonic bulletin, Volume 16 (1), May 2001, page 146.

The United Kingdom: Redistribution Process

Updated in 2011 by: Ron Johnston, David Rossiter and Charles Pattie

In February 2011 the UK Parliament, after extensive and sometimes heated debates, enacted the coalition government’s proposed major alteration of the country’s rules for redistributions. A new redistribution (the UK term for a redistricting) began in the following month. It will result in an almost total recasting of the country’s electoral map - a consequence of not only the new rules, with their emphasis on equality of electorates as the prime criterion in defining constituencies, but also the associated decision (enacted in the same legislation) to reduce the country’s MPs from 650 to 600.

The United Kingdom's previous system of redistribution operated, with some modifications, for just over sixty years. Six redistributions were completed, in 1947, 1954, 1969, 1983, 1995 and 2004/2007. The task was undertaken by four independent Boundary Commissions, one each for England, Scotland, Wales, and Northern Ireland, to a set timetable; this was changed twice after 1944, when it was set at every 5-7 years.

The four Commissions have been retained under the new legislation. Each is composed of three members appointed by the relevant Secretaries of State. In addition to those three, the chair is the Speaker of the House of Commons who neither attends nor participates; meetings are conducted by the deputy chair, a senior judge in each case. Each Commission has assessors representing departments, which supply vital information (for England and Wales these are the Registrar-General, responsible for statistical data, and the Director General of the Ordnance Survey, responsible for mapping). The Commissions make recommendations to Parliament, which can accept or reject, but not modify, them. (A Secretary of State can modify the recommendations before transmitting them to Parliament.)

The system for electing Parliament’s lower house, the House of Commons, was introduced in the thirteenth century, and went largely unchanged for the next six hundred years. Each shire (or county) and borough was invited to send two representatives, elected from among the landowners and the enfranchised burgesses respectively. When Scotland, Wales and Ireland were incorporated into the United Kingdom their members of Parliament (MPs) were similarly elected; the universities also had separate representation, as did the City of London. Changes to the system occurred largely as a by-product of three nineteenth century franchise extensions expanding the (all-male) electorate: the Reform Acts of 1832, 1867, and 1885. The franchise was extended to all males, and to females over 30, in 1918, when there was a further redistribution; total adult enfranchisement was completed in 1929, with the voting age reduced from 21 to 18 in 1969.

Before the 1832 Great Reform Act there were major variations in constituency electorates, a result primarily of eighteenth- and early-nineteenth-century industrial urbanisation. The three nineteenth century redistributions reduced these differences by removing seats from the small boroughs and reallocating them to the rapidly-expanding shires (many of the new industrial towns lacked borough status). Most of the small, rural two-seat boroughs lost their separate status, and the new constituencies allocated to the shires returned a single MP. By the turn of the twentieth century, most MPs were elected from single-member constituencies.

The nineteenth century redistributions were undertaken by the House of Commons, and were carefully constructed by the incumbent government to favour its electoral interests. The modern system of redistribution by independent commissions was not introduced until after the Second World War, in part as a response to requests for one during the 1930s.

Vivian Committee Recommendations for Redistribution

In 1942, the wartime coalition government established a committee, chaired by Registrar-General Vivian, to consider various aspects of the electoral system, including "the principles on which any [redistribution] scheme should be based". The Vivian committee identified equal representation as the basic principle for a Parliamentary democracy, with constituencies of equal population returning one member each, and set out four salient features to be taken into account during a redistribution:

  • the need for a constituency quota, or an average number of electors per electoral district;
  • the need for limits of toleration, indicating the allowable population variation around the quota;
  • the need for continuity of constituencies, with change being proposed only where necessary so that MPs could build lasting relationships with their constituents; and
  • the need for constituencies to conform to local government boundaries in order to provide community representation and ease of organisation for general elections (which is undertaken by local government officers).

Other recommendations suggested a time interval for redistributions and procedures for the four independent commissions to follow. Advice on whether each country should be guaranteed a minimum number of MPs was also offered.

The Redistribution Acts of 1944 and 1958

The first House of Commons Act (Redistribution of Seats) Act, approved in 1944, adopted many of the Vivian Committee's recommendations. It set the limit of toleration at plus or minus 25 per cent of the electoral quota, which was defined using each country’s total electorate (i.e. England, Northern Ireland, Scotland and Wales were treated separately) at the date when a redistribution commenced, rather than the total population. (The electoral roll is recompiled annually.) It guaranteed representation for Scotland, Wales and Northern Ireland at their 1944 levels, as well as indicating a desirable maximum number of MPs for Great Britain, thereby implying a maximum for England. The Initial Review of Parliamentary Constituencies, completed in 1947, was based on this Act.

Before the review was completed, however, the Boundary Commissioners claimed that they were unable both to meet the 25 per cent toleration limit and respect local government boundaries. The former requirement apparently dominated, since it came earlier in the Act's Schedule of Rules. Parliament, however, determined that the "organic" requirement to represent communities should take primacy over the mathematical requirement of (relatively) equal constituency electorates. The House of Commons (Redistribution of Seats) Act, 1949, removed the 25 per cent deviation rule and replaced it with one that constituencies should "be as near the electoral quota as is practicable". This new rule was placed after and, so it was assumed, thus considered subsidiary to the rule regarding local government boundaries; organic issues - relating to the representation of communities - took precedence over arithmetic concerns - equality of representation.

The Commissions’ First Periodical Reviews of all constituencies - following the Initial Review in 1947 - were reported in 1954. The 1944 legislation required them to be delivered within five to seven years of the previous review. This generated consternation among members of Parliament and party organisations, since constituencies were to be substantially changed soon after their creation, contrary to the Vivian Committee recommendation on continuity. Thus the government amended the Act in 1958, extending the time period between reviews to between ten and fifteen years (in 1992 it was changed again to eight to twelve years).

Although the Act was subsequently amended to take account of major local government changes in the 1970s and was then consolidated into the Parliamentary Constituencies Act, 1986, there were no further changes to the basic principles for redistributions: organic criteria retained their precedence over arithmetic. This situation was confirmed by the Court of Appeal in 1983 when four leading Labour party members unsuccessfully challenged the Boundary Commission for England’s recommendations in its Third Periodic Review (completed in 1982) on the grounds that the recommended constituencies were not as equal in electorates as they might be.

The pre-2011 Rules for Redistribution

The basic features of the pre-2011 rules were:

  • a guaranteed minimum number of seats for Scotland (71) and Wales (35), a maximum and minimum for Northern Ireland (16-18; this was introduced in 1978 after suspension of devolved government there), and a total number of seats for Great Britain (i.e. England, Scotland and Wales) that should not be substantially exceeded (613);
  • a requirement that, "so far as is practicable", constituency boundaries should not cross major local government boundaries - although this requirement was less stringent for Scotland and, especially, Northern Ireland because of changes in their local government systems;
  • a requirement that each constituency's electorate be as near the electoral quota as practicable, within the constraint of the previous rule (from which Commissions may depart to avoid disparities in electorates among neighbouring constituencies);
  • Commissions could depart from strict application of the previous two rules "if special geographical considerations, including in particular the size, shape and accessibility of a constituency" made that desirable;
  • Commissions should take into account the inconveniences that may be caused, and the local ties that may be broken, if they gave full effect to the "equal electorates" requirement.

 

Under these rules, each country had a separate electoral quota - determined by dividing the electorate when a redistribution began by the country’s then number of constituencies. Those quotas varied very substantially. For the redistribution reported in 1995, for example, they were: England - 69,281; Northern Ireland - 67,852; Scotland - 54,569; and Wales - 58,525. After devolution to Scotland, the English and Scottish quotas for the next review (completed in 2004 in Scotland and 2007 in England) were made the same, but there was no change to the Welsh situation.

The procedure for redistributions was also set out in the Acts. After determining the electoral quota for its country, each Commission then allocated a number of constituencies (its ‘entitlement’) to each local government area - or combination of neighbouring areas in some places, such as London where most of the boroughs are relatively small. It then published provisionally recommended constituencies for each area and invited written representations - both negative and positive. If there were substantial negative responses a public inquiry was held, chaired by a specially-appointed Assistant Commissioner who submitted a report on the proceedings and, if he saw fit, recommended changes. If the Commission accepted some or all of those recommendations it then published its revised set of constituencies and again invited submissions; a further public inquiry could then be held, but that was rare. When the public consultations were completed, all of the recommendations for the country were submitted to the relevant Secretary of State, for transmission to Parliament, where - after 1955 - they were either accepted or rejected en bloc. (In 1969 the Labour government placed the Commissions’ recommendations before Parliament but then ensured that they were rejected, because they feared a loss of seats were the new seats in place for the next election - which they lost in 1970; the incoming Conservative government immediately implemented the recommendations.)

The Parliamentary Voting System and Constituencies Act, 2011, and the New Rules

The procedure in place, with modifications, since 1944 was subject to considerable criticism by academic and other commentators, and the Commissions occasionally indicated the difficulties they had with applying a relatively incoherent set of rules. Parliament was not prepared to act on these, however. After the Labour party’s three general election victories in 1997, 2001 and 2005 (the first two by very large Parliamentary majorities), academic analyses showed that it had been very substantially advantaged by the operation of the system of translating votes into seats of which the rules for redistribution lay at the heart - largely because of its geography of support compared to the two other main political parties - the Conservatives and the Liberal Democrats. These showed that if the Labour and Conservative parties had obtained equal shares of the votes cast at those elections, Labour would have won many more seats (over one hundred more in 2001 and 2005); even in 2010, when the Conservatives defeated Labour by 7 percentage points in the national vote tally, if their vote shares had been equal Labour would have led the Conservatives by over 50 seats.

From 2004 on, the Conservative party identified the inequality in electorates as a major cause of the disadvantage from which it suffered. In 2001, for example, it obtained one MP for every 50,347 votes, compared to Labour’s one for every 26,031; the average electorate in a constituency returning a Conservative MP then was 72,137, compared to 67,544 for one returning a Labour MP. The Conservatives determined to change the rules to make equality of electorates the prime determinant in redistributions, and prepared a Bill ready to be introduced when it next gained power. This occurred in May 2010, when it was the major partner in a coalition with the Liberal Democrats. The Bill was tabled in July (it also contained the legislation for a referendum to be held on a possible change in the voting system to the Alternative Vote - which was lost in May, 2011 - hence the Bill’s title) and was enacted on 16 February 2011.

This Act, and especially the Schedule setting out the Rules for Redistribution (which was incorporated into a revised version of the Parliamentary Constituencies Act, 1986), sets out a clear and unambiguous sequence of rules in which the arithmetic criterion takes clear precedence:

  • There is a fixed number of constituencies - 600 - each returning a single MP;
  • There is a single UK electoral quota, obtained by dividing the total electorate when a redistribution commences by 596 (excluding the protected constituencies - see below);
  • The Sainte Laguë method is used to determine the number of constituencies for each of the four countries, with no constituency crossing a national boundary;
  • Each constituency must have an electorate within +/- 5 percentage points of the UK quota;
  • There are four ‘protected constituencies’ that are not subject to the previous rule - two are isolated, low density island areas in Scotland (the Orkney & Shetland Isles, and the Western Isles) and the other two are guaranteed for the Isle of Wight (previously England’s largest constituency with an electorate of c.110,000);
  • In addition, no constituency shall be larger than 13,000 square kilometres, and if it is necessary for a Commission to recommend one of 12,000 square kilometres or more it need not conform to the +/-5 per cent constraint, which will mean that remaining constituencies in the relevant country (it will only apply in Scotland) will on average be slightly larger than the quota;
  • If necessary, the +/-5 per cent rule can be modified with a slightly larger allowed variation in Northern Ireland (the smallest of the four countries) if its Commission considers that necessary because of difficulties in meeting the +/-5 per cent rule;
  • Within the size constraint, Commissions may take into account in determining constituency boundaries: special geographical considerations; the local government boundaries (of counties, county districts, London boroughs, and other local authorities, plus electoral wards and divisions) in place at the most recent local election prior to the start of the redistribution; boundaries of existing constituencies; any local ties that would be broken by changing constituency boundaries; and the inconveniences attendant on any proposed changes; and
  • The Boundary Commission for England could, if it wished, use the nine constituencies employed for elections to the European Parliament for the allocation of UK Parliament constituencies within England.

 

Redistributions are to take place every five years, with the first to be reported to Parliament by October 2013, approximately eighteen months before the scheduled date of the next general election on the first Thursday in May 2015, according to the Fixed Term Parliaments Act, 2011.

The Commissions have to recommend a name for each constituency (which can stimulate considerable local concern) and whether to classify it as a borough or county. Candidates are allowed to spend more money campaigning before a general election in county (rural) than in borough (urban) constituencies.

Public consultation

The 2011 Act also established a new system for public consultation. Initially the government intended there to be no public inquiries, in part because holding them extended the time taken by a redistribution and in part because the existing system had become, in one minister’s words, ‘not fit for purpose’ - the inquiries were confrontational and dominated by the political parties. The new system involves:

  1. An initial consultation period of twelve weeks after publication of provisionally recommended constituencies for an area, during which written submissions can be made. Publication involves announcing the recommendations in one or more newspapers circulating in the relevant area and sending notices to all affected MPs, political parties and local governments, giving details and indicating where maps showing the recommended constituencies can be viewed;
  2. A series of public hearings, each lasting no more than two days, at which oral cases for and against a Commission’s proposals and any alternative schemes can be presented; these are to be held during weeks 5-10 of the initial consultation period. A minimum of two and a maximum of five such hearings are to be held in each of Northern Ireland, Scotland and Wales and each of the nine English regions;
  3. A secondary consultation period of four weeks, after all of the written submissions and a transcript of the public hearings have been published, when further written comments can be made;
  4. Preparation of a report by the designated Assistant Commissioner analysing all of the evidence received (the written submissions and the oral evidence at the hearings) and making recommendations to the relevant Commission;
  5. In the light of that report, reconsideration of its provisional recommendations by the Commission and, where changes are made, publication of revised proposals with an eight-week period for further written representations only, after which the Commission will reach its final determination without further publication or consultation.

The new procedure in action

Within one month of the Parliamentary Voting System and Constituencies Act, 2011, receiving Royal Assent the four Boundary Commissions had undertaken the necessary initial tasks and commenced the redistribution that must be completed by October 2013.

Using data for December 2010, the UK electoral quota was determined as 76,641, which means that 596 constituencies must all have electorates between 72,810 and 80,473. (The electorates of the two Scottish protected constituencies - 21,837 and 33,755 - are outside this calculation, as is the electorate of the Isle of Wight - 110,924; the Boundary Commission for England has to propose two constituencies for the Isle of Wight.) The subsequent allocation of constituencies, using the Sainte Laguë rules was:

Review started in 2000
2011

Previous Allocated Preserved Total
England 533 500 2 502
Northern Ireland 18 16 0 16
Scotland 59 50 2 52
Wales 40 30 0 30

 

As a consequence of the introduction of a UK quota and its previous substantial over-representation Wales will lose one-quarter of its complement of MPs at the 2015 general election (the average electorate in a Welsh constituency at the 2010 general election was 56,545 compared to 71,882 in England); this compares to losses of 12 per cent for Scotland, 11 per cent for Northern Ireland, and 6 per cent for England.

At the start of its review, the Boundary Commission for England held a public consultation on whether it should subdivide the country into the European Parliamentary constituencies as the first stage of its allocation, and whether it should use the Sainte Laguë procedure for that process. Both proposals received general support and have been adopted.

Problems with the Review Process

Before the review started - indeed during the Parliamentary debates over the Act in 2010 and 2011 - a number of potential problems in implementing the new rules was identified.

The biggest of these was the difficulties that the Commissions will face building recommended constituencies using electoral data for smaller areas. At all of the post-1944 reviews they used local government electoral wards as the building blocks for constituencies; these are the smallest areas that have legal status. Some are relatively large, especially in urban areas (electorates of 5-10,000 in many cases and over 10,000 in some), and it will not be possible to use them and remain within the rigid +/-5 per cent limit in some places. The Commissions are making their own responses to this situation, recognising that splitting wards can have substantial impacts not only on the organisation of political parties and the conduct of elections but also on electors’ sense of local identity.

There are also problems with the electoral data because the Electoral Commission estimates that at least 3.5 million individuals (out of a total of some 49 million) eligible to register have not done so (despite a legal requirement that they should). Those ‘missing voters’ are concentrated among younger adults, including students, members of ethnic minorities, those living in rented properties and those who move frequently. A majority of them live in urban areas which means that the country’s (inner) cities may be under-represented in the next Parliament relative to suburban, small town and rural areas. This is not a new problem, and there have always been difficulties with the completeness of the electoral roll; however, the emphasis on equality of electorates in the new rules has increased its importance. The Commissions are not able to take this under-count (nor any changes in an area’s population/electorate during a review period) into account. The government intends to introduce a more complete registration process before the next review that will start after the 2015 general election.

The introduction of a new registration procedure is likely to exacerbate a further problem - that of continuity, which led MPs to change the frequency of redistributions in 1958 and was the cause of concern during the debates over the 2011 Act. Its proponents argued that conducting redistributions every five years should not lead to major changes in the pattern of constituencies for every Parliament, because population changes would be insufficiently large to cause most constituency electorates to fall outside the 5 per cent variation around the new quota. However, a significant increase or decrease in the electorate of one constituency requiring a change in its boundaries would impact on some, if not all, of its neighbours, potentially leading to them also falling outside the electoral tolerance limits, with impacts in turn on their neighbours. A necessary change to the boundaries of just a few constituencies could thus have a ripple effect across a considerable number of others, requiring a substantial redrawing of the constituency map.

Each redistribution could potentially involve a substantially changed electoral map, therefore, with consequences for MPs, party organisations, and electoral administrators but also the electors, who will lose continuity of representation. Introduction of a new system of electoral registration could exacerbate this for the review to be undertaken between 2015 and 2018. Currently in Great Britain the electoral roll is recompiled every year through a household canvass, although individuals can apply to be included on the roll at any time and many do (including recent migrants), especially in the weeks immediately prior to a general election. 

The household canvass was replaced by individual registration in Northern Ireland in 2002, with major changes in both the number of registered electors and their distribution. A similar outcome is likely in Great Britain should that system be introduced - as the government proposes by 2014. (For example, under the current household canvass all students living in University-owned and -operated accommodation are registered in that location by the relevant authorities; there are many thousands of these in a substantial number of city constituencies. This will cease under the proposed new system and unless the students register individually there - they can also legally register at their parental home, although they can only vote in one place - many urban electorates will be substantially reduced, with consequences for the subsequent redistribution.)

More constituency building?

A further constituency-building exercise was proposed by the coalition government in May 2011. Currently, the UK’s second chamber - the House of Lords - is unelected, comprising 92 hereditary peers (elected from among a much large number of them, who were removed from the House in 1999) and around 700 (mainly political) appointed life peers. The proposal is to replace this with a House of 300 members, 240 of them elected, in thirds, for fifteen-year terms and the remainder appointed, also for the same period (although the government indicated it was prepared to consider a fully-elected House).

It is proposed to elect peers using the Single Transferable Vote (STV) system from multi-member constituencies. The Electoral Districts would be Northern Ireland (which would return 9 members, three elected at each contest), Scotland (21 members, 7 per election) and Wales (12 and 4), plus a number defined in England by an ‘independent committee of experts’ returning 5-7 members each at every election. Those districts in England will be created by combining counties and should, wherever possible, be confined within one of the nine standard regions used both for statistical purposes and the election of members of the European Parliament. The ratio of electors to elected members should be ‘broadly equal’ across the country, averaging, according to current data, some 573,000 voters per member at each election - or 191,000 electors per member when the fully-elected House is completed in 2025.

The Parliamentary Voting System and Constituencies Act, 2011, can be found at http://www.legislation.gov.uk/ukpga/2011/1/contents

The revised Parliamentary Constituencies Act, 1986, is at http://www.legislation.gov.uk/ukpga/1986/56/contents.

Kosovo: Delimiting Electoral Districts for a Proportional Representation System

A number of Kosovar political leaders have urged the subdivision of Kosovo into electoral districts for the purpose of central elections. The current electoral system is a closed List Proportional Representation (PR) system with a single Kosovo-wide constituency. Reformers would like to retain the List PR system, but would like to see an open party list instituted in Kosovo and, in conjunction with this, the delimitation of electoral districts. [1]

Delimiting electoral districts is both a technically feasible and, particularly if an open party list is adopted, a desirable modification to the current electoral system in Kosovo. Delimiting districts would serve a number of commendable purposes: it would democratize the election process by decentralizing power within the political parties; it would provide geographic representation for many currently neglected areas of Kosovo; it would improve the accountability of representatives to their constituency voters; and it may decrease voter apathy and increase voter participation in Kosovo. But altering the electoral system to include the delimitation of electoral districts is not a decision to be taken lightly, and it is a decision that must be made well before – preferably at least six months before – the scheduled Election Day.

Electoral Systems that Delimit Constituencies

Traditionally, electoral systems have been categorized into three groups: plurality systems, majority systems, and proportional representation systems. [2] The most important element that differentiates these electoral systems from one another is the means by which seats in the legislature are allocated: (1) to candidates receiving a plurality of the vote, (2) to candidates obtaining a majority of the vote, and (3) proportionally on the basis of votes cast for political parties or candidates, respectively. A recent addition to these three categories is the mixed electoral system, which combines elements of both proportional representation and plurality or majority voting systems.

The significance of the delimitation process varies depending on the type of electoral system. Because most plurality and majority systems require the adoption of single-member districts and because these systems can produce disproportional election results, the delimitation process, and the decision as to which districting plan to adopt, is quite important. Although somewhat less important in the context of proportional representation systems, it is still essential that the decision on whether or not to delimit districts, and the process by which the delimitation might be accomplished, be given careful consideration.

Plurality and Majority Systems The delimitation of electoral districts is most commonly associated with plurality and majority electoral systems. Both systems tend to rely heavily, if not exclusively, on single-member electoral districts. These districts must be redrawn periodically to reflect shifts in the population.

Because of their reliance on single-member districts, the number of seats that a political party receives in these systems depends not only on the proportion of votes it received, but also on where those votes were cast. Under plurality and majority systems, minority political parties whose supporters are not geographically concentrated usually obtain fewer seats than their proportion of the vote would suggest they are entitled [3].

The major advantages associated with plurality and majority systems are that (1) they are usually quite simple to understand; (2) they offer voters a clearly identifiable representative (beholden to a specific geographic area) that can be held accountable and can be called on to provide information and services; and (3) they foster one-party government that can, in certain instances, enhance the opportunity for a stable and decisive government.

The primary disadvantage of these systems is that they can produce disproportional election results; a party with a small majority of the votes may win a disproportionately large number of legislative seats. Furthermore, smaller political parties and minority groups do not fare particularly well under these systems.

Proportional Representation Systems There are three major types of proportional representation systems: the List PR system, the Mixed Member Proportional (MMP) system and the Single Transferable Vote (STV) system. The MMP system, because it is a “mixed” system, will be discussed under the "mixed system" section below.

List PR This system is the most common PR system. Under the List PR system, if electoral districts are employed, they are relatively large multimember districts with boundaries that generally correspond to administrative divisions. To accommodate shifts in population, the number of seats allocated to individual constituencies is varied rather than redrawing the boundaries of the districts. List PR requires each party to present a list of candidates to the electorate. Electors vote for a party (or, in the case of an open list, for candidates within a certain party); parties receive seats in proportion to their overall share of the national vote. This system is widely used in continental Europe and Latin America. A closed list PR system, with a single constituency, is the electoral system currently in place in Kosovo.

Single Transferable Vote (STV) System This system, used in Ireland and Malta, is another type of proportional representation system. Under an STV system, voters are required to rank candidates in order of preference in the same manner as the Alternative Vote. After the first-place preferences are tallied, a “quota” of votes is established, which a candidate must achieve to be elected. Any candidate who has received more first preferences than the quota is immediately elected. If no one has achieved the quota, the candidate with the lowest number of first-preferences is eliminated, and their second preferences are redistributed among remaining candidates. Because voting is on the basis of candidates, not parties, these countries employ small multimember districts with only three to five members elected per district. (This makes the choices on the ballot far more manageable.) Electoral district boundaries must be redrawn periodically.

The strongest argument in favour of PR systems in general is that these systems avoid the anomalous election results of plurality and majority systems and facilitate a more representative legislature. For many newly emerging and transitional democracies, particularly those that face deep societal divisions, the inclusion of all significant groups in the parliament is an essential condition for democratic consolidation.

Other advantages include:

  1. These systems make it more likely that representatives from minority groups (and women) will be elected.
  2. Few wasted votes are cast in proportional systems. Almost all votes cast within a PR system go towards electing a candidate of choice, increasing voters’ perceptions that it is worth making a trip to the polls.
  3. Power sharing between parties and interest groups is more visible under these systems.

Some disadvantages of PR systems are: 

  1. PR systems usually lead to coalition governments, which can lead to legislative gridlock and the inability to carry out coherent policies.
  2. Some PR systems do not provide a strong linkage between a representative and his or her electorate. (This is not true of an MMP system, however.)
  3. PR systems offer a platform for small extremist parties (unless a high threshold is set for obtaining a seat in parliament).
  4. Some PR systems are criticized for leaving too much power in the hands of senior party officials (i.e., a candidate’s position on the party list, and therefore his or her likelihood of success, is often dependent on one or two party leaders). This is particularly true of a national closed-list PR system.

Mixed Electoral Systems Mixed electoral systems are becoming increasingly popular. They are called “mixed” because they employ both party list proportional representation and single-member (or small multimember) electoral districts, often with plurality or majority vote requirements.

Because mixed systems incorporate districts, delimitation must occur periodically in order to ensure electoral districts that are relatively equal in population. The importance of the delimitation process and the influence that district configurations have on the outcome of elections is dependent on whether the party list seats are used to correct any distortions in the relationship between votes cast to seats won produced by the single-member districts.

In countries such as Germany and New Zealand, seats allocated under the party list component of the system are used to compensate for any distortions in the seats-to-votes ratio produced at the electoral district level. Mixed systems that use party list seats in a compensatory manner are referred to as "Mixed Member Proportional" systems because the election results are proportional. (This system is used not only in Germany and New Zealand, but in Bolivia, Mexico, Venezuela, Hungary, and a number of other countries as well.)

In countries such as Russia, the party list seats are not used to compensate for any disproportionality arising from elections in single-member districts [4]. Instead, seats allocated to the parties under the party list component of the election are simply added to the seats won at the electoral district level. The partisan seats-to-votes ratio may therefore be distorted. In this type of mixed system, sometimes called a "parallel" system, the district delimitation process is more important because it can have a more pronounced effect on the partisan composition of the legislature. (Parallel systems are used in Russia, Japan, South Korea, Thailand and the Philippines, as well as other countries.)

Reforming the Electoral System in Kosovo There is no perfect electoral system; major design criteria often conflict with each other or are even mutually exclusive. For example, increasing the number of seats assigned to each constituency will enhance proportionality (“representativeness”) but will reduce the geographic link between a representative and his or her constituency (“accountability”). Careful consideration must be given to prioritizing the criteria that are most important in a given political context.

Comparative electoral experience suggests that the most important electoral requirement for transitional elections, particularly in a post-conflict situation like that of Kosovo, is a system that maximizes inclusiveness and is clearly fair to all parties. This goal is best achieved by a PR electoral system; no doubt this was the reason that a PR system was adopted in Kosovo. However, since its inception, some political leaders in Kosovo have advocated a change in the electoral system in Kosovo – a change that would not necessarily jeopardize the proportionality of the election results but would, at least in the opinion of many Kosovars, enhance the “democratic nature” of the election process in Kosovo.

The current electoral system is a closed list PR system with a single Kosovo-wide constituency. Reformers would like to see an open party list instituted in Kosovo and, in conjunction with this, the delimitation of electoral districts. (An open party list would be quite cumbersome to manage with 100 Kosovo-wide seats to fill; but with districts in place, the number of seats to fill within each district would probably be one-fifth to one-seventh that size, depending on the number and configuration of the districts employed.)

According to one prominent political leader, delimiting districts and opening the party list in Kosovo would [5]:

  • democratize the election process by decentralizing power within the political parties;
  • provide geographic representation for many areas of Kosovo currently unrepresented;
  • improve the accountability of representatives to their constituency voters; and
  • decrease voter apathy and increase voter participation.

These sentiments were echoed by leaders of Reform 2004, an association of well over 200 local non-governmental organizations (NGOs) in Kosovo [6]. The electoral system advocated by Reform 2004 is a variant of an MMP system, with seven multimember districts and 30 compensatory seats to ensure proportional representation. Representatives from the seven multimember districts would be elected via an open party list.

Regardless of what type of electoral system is chosen in Kosovo, assuming some system other than the current system (a single-constituency List PR system) is adopted, some delimitation of districts will be required. Although the size of the geographic area encompassed by the electoral constituency will vary depending on the type of electoral system (for example, regional list PR could adopt four or five large regions as the electoral constituencies, while an MMP system would probably require smaller constituencies), some delimitation will almost certainly be necessary. This delimitation could be as simple as adopting existing administrative boundaries (such as the current UNMIK regions) and then allocating parliamentary seats to these constituencies on the basis of population; or it could be as complex as drawing new electoral constituency lines specifically for election purposes.

Advantages and Disadvantages of Delimiting Districts

The major advantage of a change to a districted system (from a single-constituency system) is that districts link elected representatives to a smaller, geographically-defined, constituency. This allows voters to hold representatives accountable – voting them out of office if they do not act in accordance to voters’ wishes and returning them to office if they do. A geographic link also facilitates the exchange of information between voters and their representatives and promotes community services on behalf of constituents.

Another advantage to districts (assuming candidates are required to reside in the districts they represent) is that they ensure geographic diversity in the assembly. Of course, geographic diversity could also be mandated in a closed list PR system by establishing geographical distribution requirements on the candidate list, but this can be cumbersome, especially if there are already requirements for gender diversity or other forms of diversity placed on the list. It is even more cumbersome, perhaps even impossible, with an open list.

Districts would also permit the use of an open party list – something that is quite difficult, if not impossible, with a single constituency because the size of the candidate list would be unmanageable.

The most common argument against delimitation is that a districted system produces less proportional election results. This is not necessarily the case, however – an MMP system, for example, produces proportional election results (at least if enough compensatory seats are established).

Opponents of districting have argued that delimiting districts is a contentious and difficult process. Although drawing districts can be a contentious process (it is quite polarizing in the United States, for example), it does not have to be. Even in plurality or majority systems dependent entirely on single-member districts, the process can be quite routine and subject to little controversy. (This is true, in fact, of most countries that redistrict.) Moreover, in MMP systems like Germany and New Zealand, the process does not even register on the political radar. Delimitation need not be difficult, either; for example, if seats are simply allocated to already existing administrative units, the process can be managed quickly with little effort or resources required.

Delimitation may require an additional step in the voter verification process (potential residents will have to prove residence not just within the country, but within a specific district); it can also complicate absentee voting procedures and candidate eligibility verification. (On the other hand, if local elections have been incorporated into the election process and local administrative boundaries are not breached by electoral district lines, then it is not true that an additional step would be required.) Adding a district component may also make the ballot more complex by requiring two ballots rather than one. However, this problem can be alleviated with a single ballot MMP system.

A final drawback to delimiting districts, depending on the type of electoral system in which the districts are employed, is that the districting component may make the vote counting process more complex. While this is not true of most plurality and majority systems, it is true of mixed systems.

The Delimitation Process in Kosovo

Kosovo used a single-constituency closed list PR system for the assembly elections held in 2001 [7]. A number of political leaders and local NGOs are advocating a change in this system for the upcoming (and future) central elections. Of primary interest to these reformers is the institution of an open party list. However, because a Kosovo-wide single constituency would make an open ballot quite unrealistic, electoral districts have been also been promoted. Electoral districts offer the added advantages of guaranteeing geographic diversity and forging a closer link between voters and their representatives.

Two alternative approaches exist for delimiting electoral boundaries in Kosovo: 

  • Existing administrative boundaries can be used for electoral purposes.
  • Electoral constituencies can be drawn that are unique (and separate from the administrative structure).

The latter option would involve a great many resources, and would be a time-consuming and labor-intensive endeavor. This would not be the case, however, if existing administrative units are used for electoral constituencies.

Current Administrative Units Used as Electoral Constituencies

There are several different sets of administrative units in existence in Kosovo: 

  • The five UNMIK regions
  • The seven regions used by the Statistical Institute to collect and report data
  • The seven telephone exchanges (all identified with specific municipalities)
  • The 30 municipalities across Kosovo

Co-opting administrative units for use as electoral constituencies offers several important advantages:

  • Using already existing boundaries would negate the need to draw an entirely new set of electoral boundaries (which would be an extremely expensive and time consuming task).
  • Using existing administrative units would mean that election administrators and voters would already be familiar with the electoral district boundaries.
  • There is population data associated with these existing administrative units, making the exercise of allocating seats to constituencies easier, and more accurate, than would otherwise be the case.

On the other hand, there is at least one disadvantage associated with the use of existing administrative boundaries: the administrative units were not necessarily designed to encompass communities of interest, and could conceivably cut across ethnic lines and divide homogeneous groups that should be united in a single electoral constituency. In fact, in Kosovo, Serbian and other minority ethnic enclaves do not appear to cross municipal boundaries and therefore, so long as municipalities were assigned in their entirety to specific districts, minority ethnic communities of interest would not be divided by electoral district boundaries.

If the decision is reached to use current administrative units for electoral purposes, then the question remains as to which set of units to employ for electoral purposes. Of course, a large part of this decision is dependent on the type of electoral system adopted. For example, if a regional list PR system is selected, then the choice of administrative units is limited to the larger units (the five UNMIK regions, for example) since the electoral constituencies must be large enough to permit the allocation of several seats to each constituency. On the other hand, if an MMP system is adopted, then electoral constituencies could be smaller in size (i.e., perhaps as small as the municipalities).

The electoral system proposed by Reform 2004 includes seven electoral districts. While the districts in the current Reform 2004 proposal do not coincide exactly with either the seven regions defined by the Statistical Institute or the seven telephone exchange areas, this could easily be modified – and probably should be so that there is no question as to why certain municipalities have been assigned to certain districts.

Delimitation of New and Unique Electoral Constituencies

The delimitation of constituencies in Kosovo specifically for the purposes of the election of representatives to parliament would be an enormous undertaking both in terms of the time needed and the resources required. The process would involve a number of steps, including: (1) the construction of a delimitation database; (2) the creation of a districting plan by allocating territory to specific electoral districts; and (3) the evaluation of the proposed districting plan and the adoption of a final districting plan.

Construction of a Database Delimitation requires the collection of several different types of information. The two essential pieces of information are population data and maps. The population data, which is typically in the form of census enumeration data or voter registration data, provide the only means of creating constituencies that are relatively equal in population. Maps are needed to ensure that only contiguous geographic population units are assigned to constituencies and that constituency boundaries do not divide communities of interest unnecessarily.

Formation of Electoral Constituencies Once a database has been prepared, the next step in the delimitation process is the formation of electoral constituencies. This is the step in which the line drawers create a districting plan by assigning geographic units such as towns and villages (or city blocks) to constituencies. A redistricting plan is complete when all geographic units in the jurisdiction have been assigned to specific constituencies and the required number of electoral districts has been created.

Evaluation of Redistricting Plan Once the boundary authority has successfully completed a redistricting plan, summary information for the plan should be produced in order to evaluate the plan. A summary description of a redistricting plan should include information such as the geographic components of each constituency, maps of the plan showing the constituency boundaries, and a report summarizing the most relevant statistical information for each constituency in the plan.

This information should allow the boundary authority, political parties, legislators and governmental officials, citizens, and other interested stakeholders to evaluate the proposed redistricting plan according to established criteria. Public hearings may be held to solicit the comments of these stakeholders. After the solicitation process has been completed, the authority in charge of delimitation should endeavor to take these comments into account, and modify the redistricting plan accordingly.

The final stage of the process is the adoption of the new redistricting plan. Provisions for how this is accomplished should be described quite explicitly in the electoral law. In fact, the entire process (who should draw the constituencies, what criteria should be followed, etc.) should be mapped out as clearly as possible beforehand to guide authorities in charge of the process.

Conclusion It is not technically feasible (given the lack of sufficiently refined data and the current time constraints) for the delimitation of a unique set of electoral districts (especially single-member districts) for the 2004 central election. Furthermore, the delimitation of unique electoral districts could well prove a political nightmare in future elections and is not recommended for Kosovo.

Delimiting Districts in Kosovo for the 2004 Central Elections

If electoral districts are to be adopted for the 2004 central elections, these districts should be based on current administrative district lines. The basis for this assertion is at least threefold:

  • The existing population data (voter registration data) is insufficiently refined for the delimitation of unique electoral district boundaries (at least boundaries that cross municipal boundaries).
  • The risk of political tensions arising during an active delimitation exercise is certainly not minimal, and therefore the delimitation of a unique set of electoral districts is best avoided.
  • There is not enough time at this point (mid-February) in the election calendar to engage in a detailed delimitation exercise. 

However, there is sufficient information – and enough time – at this point to modify the electoral system to include electoral districts if these districts coincide with currently existing administrative boundaries.

Although the OSCE outlined a number of objections to changing the electoral system, they did concede that districting prior to the upcoming October 2004 elections was still “technically feasible.” [8] Some of the arguments offered by OSCE for not districting in Kosovo, and the reasons why these objections are not necessarily well-founded, are as follows: 

  • Districts could lead to less proportional election results While it is true that single-member districts could result in less than proportional election results, the electoral system proposed by Reform 2004, for example, would be no less proportional than the current Kosovo-wide List PR system. And this is true of any MMP or regional List PR electoral system that incorporates districts.
  • Existing data is insufficient for ensuring districts of equal size Although no accurate census data exists [9], there is up-to-date voter registration data that can be used to allocate seats to electoral districts. Voter registration data is, in fact, quite often used for this purpose [10].
  • Districts require complicated procedures for voter eligibility and for absentee voter assignment Since the voter registration process must already take into account voter residence for municipal elections, so long as municipal boundaries are not crossed by electoral districts lines (i.e., municipalities are allocated intact to a single electoral district), voter eligibility and absentee voter assignment will not be affected by the introduction of electoral districts.
  • Ballot creation and distribution would be more complicated It is true that the single party ballot would have to be supplemented with five to seven additional ballots (one for each district) and that all of these ballots would have to be distributed across districts. But this is still far less cumbersome than the ballot production process required for open list municipal elections in 2000.
  • Vote count would be more complex The vote count would, in fact, be more complex. With sufficient notice, however, training could commence on the vote count process and there is no reason to believe that Kosovars would be any less successful than, for example, Hungarians or citizens of any other transitional democracy with an MMP system, in counting the ballots correctly.
  • Election would be more expensive It is true that the election would be somewhat more expensive to administer. But the trade-off would be a more “democratic” election – one that Kosovars may well be more willing to claim ownership of and participate in. 

The OSCE also indicated that the administrative approval process within OSCE and UNMIK is time-consuming and may not be completed in time to institute districts prior to the upcoming election.

Choosing a Districting Plan and Allocating Seats

The following administrative boundaries have been identified as reasonable prospects for electoral district boundaries:

  • The five UNMIK regions
  • The seven Statistical Institute regions
  • The seven telephone exchange regions 

The 30 municipalities were rejected as a possibility because some municipalities are too small in population to be accorded their own representative – a political decision would have to be made whether to allocate these municipalities a representative regardless or to combine these municipalities with other municipalities to meet the electoral quotient.

Once a districting plan has been selected, parliamentary seats must be allocated to each of the electoral districts within the plan. The allocation process (also referred to as apportionment) is almost always based on population data, usually in the form of census enumeration data or voter registration data [11].

This phase of the delimitation process is relatively mechanical, although the decision as to what formula to use for apportioning seats to districts can be a controversial one. Depending on the size of the administrative units chosen to serve as electoral districts, it is also possible that some units (i.e., small municipalities) will have to be combined if districts of relatively equal population are to be created and the electoral quotient is higher than the population of a number of these units.

A series of simulations were conducting using the three possible delimitation plans identified above. The 2003 voter registration data was used to determine the seat allocation to each of the electoral districts in these plans [12]. The results of these simulations can be found in the Appendicies. The table below summarizes the results of the seat allocation exercise: 

 

UNMIK Districts

Telephone Exchange Districts

Statistical Institute Districts

Number of Districts

5

7

7

Range in Seat Allocation

9 to 19

6 to 18

7 to 18

Maximum  Deviation

4.9%

4.9%

4.9%

Minimum Deviation

-2.9%

-7.8%

-4.8%

Total Percent Deviation

7.8%

12.7%

9.7%

Delimitation Timeframe

The time required to delimit districts, and the cost associated with this endeavor, vary dramatically depending on how extensive the delimitation process is. If the system entails the drawing of an extensive and unique set of single-member constituencies, for example, the process can be quite expensive and time-consuming. On the other hand, if existing administrative units are used as constituencies, and legislative seats are simply allocated to these seats on the basis of population, then the delimitation process is straightforward, and not at all costly or time-consuming.

The delimitation of a restricted number of electoral districts (five to seven) that coincide with currently existing administrative regions will not require much in the way of additional time or resources. Assuming a fall Election Day, political parties (and potential candidates) will have to be informed of the change in the electoral system soon in order to prepare additional party lists and modify campaign operations – but certainly notification of such a change sometime in the spring of 2004 should be sufficient. A campaign to inform voters of the change need not begin until the summer of 2004. Ballot production will also be effected by the adoption of districts, but a decision to incorporate districts by mid-spring will not adversely affect the printing of ballots so long as the political parties are able to organize regional party lists in a timely manner. Finally, the vote count will be more complicated, but there is more than sufficient time if the decision is made in the spring to train vote counters. 

Notes:

[1] An open party list would be quite cumbersome to manage with 100 Kosovo-wide seats to fill; with districts in place the number of seats to fill within each district would probably be one-fifth to one-seventh that size, depending on the number and configuration of the districts employed.

[2] There are at least two other electoral systems that cannot be classified as plurality, majority, proportional, or mixed. These two systems, sometimes been referred to as “semi-proportional,” are the Single Non-Transferable Vote (SNTV) and Limited Vote (LV) Systems. In an SNTV system each elector has one vote but there are several seats in the constituency to be filled, and the candidates with the highest number of votes fill these positions. This system is used in Jordan and Vanuatu (and was used in Japan until 1993). A LV system is similar to SNTV, except that voters are permitted to cast more than one vote – but fewer votes than there are seats to be filled. This system is used in the Spanish upper house and in Gibraltar.

[3] There are four electoral systems commonly identified as plurality or majority systems: First-Past-the-Post (FPTP), Block Vote (BV), Alternative Vote (AV) and Two-Round System (TRS). 

  • First-Past-the-Post (FPTP): elections are held in single-member constituencies, and the winner is the candidate with the most votes, but not necessarily an absolute majority of the votes. Countries that use this system include the United States, Great Britain, Canada, India and many countries that were once part of the British Empire.
  • Block Vote (BV): this system is an application of FPTP in multimember rather than single-member constituencies. Voters have as many votes as there are seats to be filled, and the candidates with the highest number of votes fill the positions regardless of the percentage of the vote they actually receive. This system is used in some parts of Asia and the Middle East.
  • Alternative Vote (AV): in this system, electors rank the candidates in order of choice. If no candidate has over 50 percent of first-preferences, lower order preference votes are transferred until a majority winner emerges. This system is used in Australia and some other South Pacific countries.
  • Two Round System (TRS): has two rounds of voting, often a week or two weeks apart. The first round is the same as a FPTP election and, if a candidate receives an absolute majority in this round, then this candidate is elected outright. If, however, no candidate has received an absolute majority, then a second round of voting (with a more limited number of candidates) is conducted, and the winner of this round is declared elected. This system is widely used in France, many former French colonies, and some parts of the former Soviet Union.
  •  

     

[4] For example, if a political party were to win 55% of the total vote cast in a parliamentary election but win only 45% of the constituency seats, compensatory seats would be allocated to the party such that the percentage of seats held by that party would total 55% of the assembly seats overall.

[5] Adnan Merovci, CEO of the Central Election Commission, in an interview with the author of this report on 5 February 2004.

[6] This statement is based on the author’s interview (3 February 2004) with Leon Malazogu of KIPRED (Kosovo Institute for Policy Research and Development) and Burim Ejupi of The Forum, representatives of the two largest NGOs in Kosovo and leaders within Reform 2004.

[7] The Kosovo-wide district was used to elect 100 representatives from a general closed party list and 20 representatives from lists reserved for Kosovo’s smaller communities: 10 seats to Kosova Serbs, 4 to the Roma, Ashkali and Egyptian community, 3 to the Bosnian Community, 2 to the Turkish Community and 1 to the Gorani community.

[8] Interview conducted by the author with Lars Lagergren and Dennis Ennis, OSCE Division of Election Operations, 4 February 2004.

[9] Although a decennial census was routinely conducted until 1991, in 1991 Albanians in Kosovo boycotted the enumeration process (and the Yugoslavian government manufactured population estimates for Kosovo). No census has been undertaken in Kosovo since then, although plans for a census are currently before the SRSG.

[10] Almost half of the countries surveyed by the Epic Project use voter registration data for districting purposes. A list of these countries include: Albania, Armenia, Australia, Barbados, Bahamas, United Kingdom, Croatia, Iceland, Namibia, Slovakia, Ukraine, and Zimbabwe. (See the EPIC Project, a joint IFES, International IDEA and UN project that can be found at www.epicproject.org.)

[11] The choice of whether to use census data or voter registration data may be guided by either practical or theoretical concerns. For instance, census data may not be the best option if a general enumeration of the population is unavailable, outdated or inaccurate (as is the case in Kosovo). On the other hand, registration data may not adequate for redistricting purposes if it fails to include information that is essential given the specific country context. From a theoretical perspective, delimitation based on registration data is likely to produce districts that are more equal with respect to the number of voters contained within them, but a counter-argument could be made that representatives serve all persons, not simply voters.

[12] A model similar to Reform 2004’s suggested electoral system is utilized for the simulation exercise. However, rather than using the proposed 140 seat legislative (which is not recommended as it would dilute the value of the 20 set-aside seats) the current 120 seat legislature is retained: 20 set-aside seats, 30 compensatory seats (as proposed by Reform 2004), and 70 seats allocated to multimember districts.

Malaysia: Malapportioned Districts and Over-Representation of Rural Communities

Malaysia utilizes a simple plurality First Past the Post (FPTP) electoral system modeled on the British Westminster System, with 219 single member constituencies (SMCs) used for electing representatives to the House of Representative. In the most recent elections of March 2004, the ruling coalition won 90% of the seats in the House (198 out of the 219) with only 60% of the national vote. Opposition parties lost more than half of their 45 seats in the election, winning only 10% of the seats (20 seats) despite attracting nearly 40% of the votes nationwide.

Since independence, Malaysia has been governed by a coalition of political parties named the Barisan Nasional (BN) [1]. While striving to promote the multi-ethnic nature of the coalition, true power resides with the dominant ethnic Malay party, the United Malays National Organization (UMNO). Other coalition members also represent specific ethnic groups but retain very little autonomy from UMNO. Parties outside the coalition have never captured more than 40% of the seats in Parliament and under the current electoral framework, they never will.

The BN’s tight control over the election process has limited the ability of opposition parties to successfully contest elections. The Election Commission is seen as one of the primary instruments through which the BN has manipulated the election process for its own political gain [2].

Electoral System

Malaysia is technically a monarchy, although the “Paramount Ruler” (Yang di-Pertuan Agong) is elected every five years by and from the hereditary rulers of nine Malay states and plays a very limited role in governing the country. At the national level, the federation has a bicameral legislature consisting of the Senate (Dewan Negara) and the House of Representatives (Dewan Rakyat). Of the 69 members of the Senate, 43 are appointed by the king, with the remaining 26 elected from the state legislatures. The Senate is generally considered little more than a rubber-stamp for legislation passed by the House of Representatives.

The House consists of 219 members (increased from 193 in 2003) elected from single member constituencies throughout Malaysia’s thirteen states and three federal territories. The Federation of Malaysia utilizes a plurality first-past-the post electoral formula based on single-member constituencies. Elections are to be called at least once every five years and, since independence in 1957, elections have taken place every fifth year.

Legal Framework for Delimitation

The Election Law Malaysian elections are governed both by the constitution and by the “Elections Act of 1958,” both of which are subject to frequent amendment. The Constitution stipulates the FPTP formula and establishes criteria for the Election Commission.

Election Commission Established in 1957, the Malaysian Election Commission is charged with conducting elections for the House of Representatives and state legislatures. The Commission is also charged with recommending changes to constituency boundaries, which are then implemented by the federal government. The Commission is also responsible for the planning and oversight of all of the technical aspects of voter registration and elections. It also acts as a judicial body, hearing grievances from both candidates and electors about any aspect of the election process.

The Commission originally consisted of three members, a chairman and two subordinates. In 1963 an additional member was added to represent the states of Sabah and Sarawak. In 1981, the post of Deputy Chairman was established, bringing the total number of members of the commission to five, where it remains today. All members are appointed by the Paramount Ruler in consultation with the Conference of Rulers, an unelected body consisting of the executives of each state. The five members may serve until the mandatory retirement age of sixty-five, and may be removed from office only by a special tribunal called by the Prime Minister. Members of Parliament may not serve on the Commission [3].

The Election Commission is not a fully autonomous body. The Constitution stipulates, “so far as may be necessary for the purposes of its functions under this Article the Election Commission may make rules, but any such rules shall have effect subject to the provisions of federal law.” Therefore, any provision created by the Commission can be reversed by a federal law. In addition, any recommendations for changes to constituency boundaries proposed by the Commission must first go to the Prime Minister, who may make alterations as he sees fit. The Prime Minister then submits the proposal to the House of Representatives, which then approves or disapproves of the delimitation plan [4].

Criteria for Delimitation All criteria for the delimitation of electoral boundaries are contained in the Federal Constitution, as modified by periodic “Amendment Acts.” Core principles related to districting criteria include:

1) Delimitation may not take place more frequently than once every eight years;

2) No single delimitation exercise may take longer than two years to complete;

3) The recommendation of the Commission is submitted to the Prime Minister, who must then present it to the House of Representatives with or without amendment for a simple-majority vote [5].

The core issue confronting the Malay election system is the constitutional provision guaranteeing over-representation of rural constituencies. This principle was a product of negotiations held between the British colonial authorities and the two main Malay independence movements during the 1950s. In 1953, the British established a 46-member committee of the Federal Legislative Council to make proposals for a post-independence electoral system. The Committee recommended equality of population across the SMC districts but qualified this proposal by including an exception for rural areas. Since ethnic Malays predominated in the rural areas and non-ethnic Malays resided primarily in the urban centers, this “rural weightage” effectively ensured Malay dominance of the political system.

The Committee report held that: “the number of inhabitants within each constituency should be approximately equal except that, having regard to the greater difficulty of contacting voters in the country districts and the other disadvantages facing rural constituencies, a measure of weightage … should be given to the rural constituencies.” [6] The original 1957 Constitution contained a provision limiting the size discrepancy between any two districts to no more than 15%. This restriction, however, has since been eliminated by constitutional amendments in 1962 and 1973 [7].

Gerrymandering Districts to Benefit Ruling Party

Size Discrepancies in Districts One of the biggest complaints from the opposition has been that the ruling party, the BN, through electoral gerrymandering, has slowly eroded the principle of “one man, one vote.” Districts that have traditionally demonstrated strong support for opposition parties often have disproportionately large populations when compared to those districts that have traditionally supported the BN. For example, Penang – a state where opposition parties have done very well in the past – averaged 50,838 voters per district; on the other hand, Perlis, which has typically supported the ruling party, averaged only 33,032 voters per district in 1990 [8].

Gerrymandering by the BN also appears to favor the native Malay population, traditionally strong supporters of the party, at the expense of the large Chinese and Indian minorities. The Malay population tends to live in more rural areas, whereas the Indian and Chinese tend to live in the urban centers. The delimitation exercise in 1994 created rural districts with much smaller populations than their urban counterparts. For example, the constituency of Hulu Rajang, a rural district with a large Malay population in Sarawak, has an electorate of 16,085 and sends one representative to the House; Ampang Jaya, an urban constituency near the capital, with an electorate of 98,954 also sends only one representative to the House [9]. While the constitution does allow for increased weightage to be given to rural constituencies, the elimination of the 15% limit by the BN has allowed for extremely large discrepancies.

2003 Delimitation Proposal The most recent round of electoral boundary delimitation, which took place in early 2003, sparked numerous complaints from opposition parties. In 2002, the BN asked the Electoral Commission to develop a new proposal for electoral boundaries to reflect changing population demographics. In response, the Commission developed a plan to create 25 new seats in the House of Representatives and 53 new state assembly seats. The House of Representatives subsequently approved the plan on April 8, 2003.

From the very beginning, opposition members opposed the plan, declaring it unconstitutional and claiming that, “the EC had not acted fairly and professionally in accordance with the principles of democracy." [10] In response, the BN proposed that all criticisms could be aired during the parliamentary debate on the proposal. As angry opposition party members publicly vented their frustration, however, the BN refused to address the issues raised and ended debate after just two days. Forty-four opposition members walked out just before the final vote to protest against both the proposal itself and the BN’s abuse of parliamentary procedure [11].

The DAP (Democratic Action Party) also raised a challenge against the plan under the Election Commission’s grievance process. The Commission’s chairman refused to hear the grievance, however, claiming that “although [the DAP] did submit a personal letter asking me to reconsider the State's proposal on the delineation, [it] could not come up with a counter proposal which can grant a representation to be made and enable the EC to conduct [a] local inquiry to hear and consider appeals or objections.” The Commission effectively made it impossible for anyone to file objections to the plan unless they offered a full counter-proposal. The chairman of the Commission also refused to hear any objections filed on behalf of “an organization, political party, or certain communities.” [12]

Changes in the 2003 Delimitation The delimitation proposal passed by Parliament created 25 new constituencies – most of which emerged out of districts that had overwhelmingly supported the BN during the 1999 general election [13]. Many of the changes seem to blatantly ignore population trends. For example, the state of Selangor, with a population of 4.19 million and an annual growth rate of 6.1 % since the 1991 census, received five new seats. Johor, however, with a population of 2.74 million and an annual growth rate of 2.6%, was granted six new seats. Given the much larger total population and the much higher growth rate, Selangor should have received more new constituencies than Johor. The reason for the inconsistency is obvious: In the 1998 election, the BN only won 54.8% of the popular vote in Selangor, whereas in Johor they won 75.2% of the vote [14].

The 2003 delimitation produced the highest population variations of any previous delimitation exercise. For example, Johore Bahru now has an electorate of approximately 90,000 voters, while Lenggong has approximately 21,000 voters – a population variation of over 325 percent. The maximum population deviation created by the 1994 delimitation was 250 percent [15].

Conclusion

Although Malaysian voters are generally free from overt forms of intimidation during the voting process, more subtle forms of manipulation by the ruling BN party has created a system that is less than fair for opposition parties. Evidence of defective voting rolls, manipulation of postal votes, instances of vote buying through promises of lavish government programs in certain constituencies, and manipulation of the Electoral Commission have helped to maintain the BN’s control over the government.

The boundary delimitation process has been a primary tool in the BN’s manipulation of the electoral process for several reasons:

  • Elimination of constitutional safeguards protecting the independence of the EC: The original constitution contained a provision allowing no more than 15 percent deviation between constituency populations. Constitutional amendments have removed the 15% limit which, when coupled with the provision allowing for increase weight to be given to rural districts, have allowed for gross discrepancies in constituency populations.
  • Lack of independence of the Election Commission: The Government appoints all members of the EC, and all recommendations made by the EC must pass through the Government in order to take effect. The BN has been able to hastily push through delimitation proposals without serious debate in Parliament.
  • The Election Commission’s unresponsiveness to complaints: The commission has proven unwilling to answer grievances against delimitation plans brought by political parties or other groups. The difficulty in judicially challenging EC decisions, coupled with the questionable independence of the judiciary [16], has allowed the Commission to avoid any serious challenges.

Appendix A: Seats Won by State & Party in the 1999 General Elections [17]

 

State/Territory

Total Seats

BN

PAS

ADIL

DAP

PBS

MDP

Other

Perlis

 3

3 

 -

Kedah

15

7

8

-

-

-

-

-

Kelantan

14

1

10

3

-

-

-

-

Penang

11

6

-

1

4

-

-

-

Perak

23

20

2

-

1

-

-

-

Pahang

11

11

-

-

-

-

-

-

Selangor

17

17

-

-

-

-

-

-

Federal Territory

11

7

-

-

4

-

-

-

Negri Sembilan

7

7

-

-

-

-

-

-

Malacca

5

4

-

-

1

-

-

-

Johor

20

20

-

-

-

-

-

-

Terengganu

8

-

7

1

-

-

-

-

Sabah

20

17

-

-

-

3

-

-

Sarawak

28

28

 

-

-

-

-

-

TOTAL

193

148

27

5

10

3

0

0

* The BN won 148 out of 193 seats, but won only 56% of the popular vote.

Appendix B: Population by State [18] (in thousands) 

State

1991

2001

2003

 Perlis

 184.1 198.3 214.5

Kedah

1304.8

1572.1

1700.4

Kelantan

1181.7

1289.2

1394.4

Terengganu

770.9

879.7

951.5

Penang

1065.1

1225.5

1325.5

Perak

1880.0

2030.4

2196.0

Pahang

1036.7

1231.2

1331.6

Selangor

2289.2

3947.5

4269.6

Negri Sembilan

691.2

830.1

897.8

Malacca

504.5

602.9

652.1

Johor

2074.3

2565.7

2775.1

Sabah

173.6

2449.4

2649.2

Sarawak

1648.2

2012.6

2176.8

Appendix C: Change in Seat Allocation, 1994 to 2002 [19]

Increase in the percentage and number of voters by state between the 1994 and the 2002 delimitation exercises and under- or over- allocation of seats by voters for each state in these two delimitation exercises.  

State

Voters in 1994

Voters in 2002

% Change

Difference

Over/Under allocation in 1994

Over/Under allocation in 2002

Perlis

97,978

109,750

12.0%

11,772

1

1

Kedah

675,790

793,517

17.4%

117,727

1

-1

Kelantan

528,679

655,602

24.0%

126,923

3

1

Terengganu

337,918

411,453

21.8%

73,535

1

0

Penang

563,039

659,155

17.1%

96,116

-1

0

Perak

1,047,175

1,138,010

8.7%

90,835

1

2

Pahang

456,834

554,534

21.4%

97,700

1

3

Selangor

949,317

1,368,693

44.2%

419,376

-3

-5

Wilayah

591,806

664,233

12.2%

72,427

-3

-2

N. Sembilan

298,178

417,712

40.1%

119,534

1

0

Melaka

269,198

331,327

23.1%

62,129

-1

-1

Johor

982,484

1223,532

24.5%

241,048

-1

2

Total

6,798,396

8,327,518

22.5%

1,529,122

 

 

Notes:

[1] Malaysia is a federation consisting of thirteen states and two federal territories. Eleven states and the federal territories are contiguously attached on the Malay Peninsula, and two additional states (Sabah and Sarawak) are on the Island of Borneo. These latter states joined the federation only in 1963 and are accorded special representation rights under their ascension agreements.

[2] US State Department Annual Human Rights Report, 1999.

[3] Art. 114 (3) Malaysian Constitution.

[4] Thirteenth Schedule to the Malaysian Constitution, Part II (8&9).

[5] Other principles include constituencies not crossing state boundaries, availability of administrative facilities for carrying out elections, size of constituencies, and the desire to avoid excessive changes to constituencies; these are to “as far as possible be taken into account.” Ibid. Part I (2) (a-d).

[6] Ibid. Part I (2) (c).

[7] See, Lim Hong Hai, “Electoral Politics in Malaysia: Managing Elections in a Plural Society,” and “The Electoral Process,” available at http://www.malaysia.net/aliran/hr/js10.html.

[8] “The Electoral Process,” available at http://www.malaysia.net/aliran/hr/js10.html.

[9] Arjuna Ranawana, “The Maps to Power: Anwar’s Claims Fill the Court and the Media,” Asiaweek.com, 5 November 1999, available at http://www.asiaweek.com/asiaweek/magazine/99/1105/nat.malaysia2.html.

[10] “EC Chief: Constituency Delineation Exercise Constitutional,” New Straight Times, 6 March 2003, 2.

[11] “Parliament OK’s re-Delineation of Electoral Boundaries,” Financial Times, 8 April 2003.

[12] “EC Chief: Constituency Delineation Exercise Constitutional,” Malaysia Election Commission Online available at http://www.spr.gov.my/surat_khabar/2003/060303_nst.html.

[13] Lim Kit Siang, “DAP Will Challenge the Constitutionality of the 2002 Electoral Constituency re-Delineation Exercise,” DAP Media Statement, available at http://www.malaysia.net/dap/lks1804.htm.

[14] “Population Distribution and Basic Demographic Characteristics Report,” Malaysian Department of Statistics, available at http://www.statistics.gov.my/English/pressdemo.htm; “Re-delineation Exercise in Sabah Based on Current Needs, Says EC,” Financial Times Asia Africa Newswire, 4 April 2003; also see Appendix A for 1999 general election results and Appendix B for population figures.

[15] Siang, op. cit.

[16] See 2005 United States State Department 2005 Annual Human Rights Report.

[17] “Malaysian General Election 1999,” available at http://www.sadec.com/Election/parliment.html.

[18] The World Gazetteer, available at http://www.world-gazetteer.com/fr/fr_my.htm.

[19] http://www.malaysia.net/aliran/monthly/2002/8f.html

New Zealand: Drawing Electoral Districts to Guarantee Minority Representation[1]

When New Zealand adopted a new electoral system for parliamentary elections in 1993, the country retained its tradition of separate districts for the descendants of New Zealand’s aboriginal Maori population. This unique electoral feature has guaranteed Maori representation in the New Zealand legislature for more than 125 years.

Electoral System

In 1993, New Zealand replaced its first-past-the-post (FFP) method of plurality voting in single-member districts with a German-style, mixed-member proportional (MMP) electoral system for elections to its single chamber legislature, the House of Representatives. As in other MMP systems, each voter cast two ballots: one for a representative elected by plurality from a single member electoral district and one for a national party list. Following the German compensatory principle, seats that parties win in districts will be subtracted from a party’s list allocations, so each party’s overall representation in Parliament will be proportional to the vote for its list.

The single-member districts consist of two types of constituencies – General and Maori. The dual-constituency feature can be visualized as a map with two overlays – one dividing New Zealand into numerous General electorates[2], the other apportioning the same territory into a smaller number of geographically larger Maori electorates. Members of Parliament (MPs) elected from both types of electorates serve in the chamber with equal rights and privileges.

Maori Representation

Maori representation was guaranteed though the establishment of separate Maori electorates as early as 1867. These electorates, separate and distinct from the General electoral districts, are drawn overlaying the General electoral districts. The three maps appended to the end of this case study, showing the 2002 electorates, illustrate this point: the first map shows the 46 General electoral districts covering the North Island; the second map shows the 16 General electoral districts covering the South Island; and the third map shows the seven Maori electoral districts covering both the North and South Islands. (See Appendix A)

Although separate Maori electorates were established only 13 years after New Zealand’s first parliament met, the periodic review of their boundaries did not become the responsibility of the Representation Commission until 1983[3]. Even then, the number of Maori electoral districts remained fixed at four. It was not until the adoption of the MMP electoral system, in 1993, that it was determined that the number of Maori electoral districts should be calculated on exactly the same basis as the General electorates.

The number of Maori electoral districts largely depends on the number of Maori who choose, during the Maori Option period, to be registered on the Maori electoral roll. The Maori electoral option period begins shortly after the census is conducted, and continues for approximately four months. This option provides all Maori of voting age with an opportunity to declare which electoral roll they wish to be registered, either the Maori or the General. Once the option has ended, and the proportion of Maori electors opting for the General and Maori rolls is known, the Maori electoral population (MEP) can be calculated.

The Maori electoral option held during the latter half of 1994 resulted in a significant number of Maori opting to be registered on the Maori electoral roll, and the number of Maori electoral districts increased from four to five. The Maori options held before the 1998 and 2001 electoral distributions also led to increases in the number of Maori electorates: from 5 to 6 in 1998 and from 6 to 7 in 2001.

Legal Framework for Redistribution

Electoral redistribution[4] in New Zealand has had several distinguishing characteristics in additional to the separate Maori districts; three of which are particularly important to the New Zealand redistribution process, and all three of which are protected by reserved provisions [5] of the 1993 Electoral Act:

  • An independent commission is given sole responsibility for delimiting electoral boundaries. Once it publishes a final plan, the plan has the force of law and cannot be challenged.
  • Redistributions must take place following every five yearly census.
  • Electoral districts are based on total population and no electorate can vary by more than plus or minus five percent from the electoral quota.

Boundary Authority Redistributions are undertaken by an independent seven-member statutory body known as the Representation Commission. Four of the members are ex officio (that is, they are members by virtue of the positions they hold within New Zealand’s public service): the Surveyor-General, the Government Statistician, the Chief Electoral Officer and the Chairperson of the Local Government Commission. Two members are appointed by the Governor-General, following nomination by parliament; one represents the party or parties in Government and the other represents the party or parties in Opposition. The seventh member of the Representation Commission is nominated by the other member of the Commission and is appointed by the Governor-General. This person serves as the chairperson. Since the present Commission was first constituted in 1956, the chairperson has always been a member of the judiciary.

There are six voting members of the Representation Commission; the Chairperson of the Local Government Commission, who is appointed by the government, is a non-voting member of the Commission. When the Commission redefines the Maori electorates, it is augmented by three additional members: the Chief Executive of the Ministry of Maori Development ex officio, and two additional members appointed by the Governor-General, each of whom must be Maori. One represents the party or parties in Government, and the other represents the party or parties in Opposition. These three members are included on the Commission to provide a Maori community of interest perspective.

The four ex-officio members of the Commission provide expertise in the areas of topography and mapping, population distribution, electoral administration, and the relationship between proposed electoral district boundaries and local government boundaries. While the original intention was that the two “political” members would act largely as scrutinisers to satisfy themselves and their parties that the redistribution process had been conducted fairly and with the established rules, in more recent years these members have become much more active participates in the process. They (or their appointed deputies) must both be present at a commission meeting before the quorum requirement is met, and both have votes to cast. The independent chairperson commissions meetings and contributes, at a minimum, such skills as an interpreter of the law and meeting facilitator.

The Representation Commission has no more than six months after it commences formal deliberations to publish its final redistribution plan. Once published, the decision has the force of law and cannot be challenged. While the High Court accepts that it has a responsibility to ensure that the Commission operates within the powers granted to it by parliament, because the Commission is a creature of statute, the High Court has held that it has no jurisdiction “to inquire into the merits of the decisions of the Commission adjusting electoral boundaries.”

Frequency of Redistribution Redistributions are conducted every five years following the population census and the Maori Electoral Option. Since the length of the parliamentary term is restricted to a maximum of three years, each redistribution applies to, at most, two elections.

Census night (the first Tuesday in March of every fifth year) provides the trigger but the actual timing of the redistribution is dependent on when the next general election is scheduled. Redistributions must be completed within six months of the Commission’s first formal meeting so when an election falls in the same year as a census, the redistribution is delayed because there is insufficient time between the census and the last possible date that an election can be held for a redistribution to be completed.

Redistribution Criteria There is only one mandatory redistribution criterion: no electorate can vary more than plus or minus five percent from its appropriate electoral district quota. A further restraint, applying only to General electorates, is that no electorate can be constructed that is located partly in the North Island and partly in the South Island. [6]

Provided the electoral quota is met, the electoral boundaries are redrawn using the criteria that are outlined in Sections 35 (f) and 45 (6) of the 1993 Electoral Act. These criteria include taking account of:

  • Existing boundaries of the electoral districts
  • Communities of interest (including such factors as tribal affiliations for Maori electorates)
  • Facilities of communications
  • Topographical features
  • Projected variations in electoral populations including anticipated changes such as large increases or decreased in the size of the population over the next five years

Of these criteria, no order of priority is specified. It would appear, however, that existing electoral boundaries may take precedence – certainly an effort is made to modify the existing plan as little as possible given population constraints [7]. In the report issued by the 2001-2 Representation Commission, the Commission indicated that it “recognises the importance of equality of representation thorough the life of the Electoral Districts (in the view of this Commission, probably through to 2007) and the desirability of changing existing boundaries as little as possible. The proposed boundaries published in November 2001 made greater use of the quota tolerance than did the commission in 1998 in order to meet these specific considerations while best balancing the other criteria of the Act.” Because of the strict tolerance limits of +/- 5% [9], however, a large proportion of electorates require redrawing every redistribution. The table below lists the percentage of electorates that fell outside tolerance limits over the last few redistribution cycles:

 

Year

Total Number of electorates

Number of electorates outside the tolerance limits

% outside the tolerance limits

1977

83

50

60.2

1983

88

42

47.7

1987

93

 

 

1992

97

33

34.0

1998[10]

65

26

40.0

2002

67

29

43.3


The number of electorates that were changed, however, was actually much greater – at least in part due the “ripple” effect [11].

Allocating Seats and Calculating the Electoral Quota The boundaries are drawn based on the total number of ordinarily resident people in each electorate. This includes adults and children and is not based on the number of people who have enrolled as Parliamentary electors.

The General electoral population is the total number of people from the last census minus the Maori electoral population. The Maori population is calculated by taking the ratio of the number of people registered in the Maori electoral rolls compared to the total number of people on all electoral rolls (General and Maori) who said they were of Maori descent when they last enrolled. The Government Statistician then applies that proportion to the total number of people who said they were of Maori descent at the most recent population census.

The number of General electorate seats in the South Island is fixed at 16. The General electoral population (GEP) in the South Island is divided by 16 to give the General electoral district quota for the South Island. This quota is used to calculate the number of North Island General seats and the number of Maori seats. According to Section 35 (3) of the 1993 Electoral Act:

(a) The South Island shall be divided into 16 General electorate districts:

(b) The General electoral population of the South Island shall be divided by 16, and the quotient so obtained shall be the quota for the South Island:

(c) The General electoral population of the North Island shall be divided by the quota for the South Island, and the quotient so obtained shall be the number of General electoral districts in the North Island. Where that quotient includes a fraction, the fraction shall be disregarded unless it exceeds a half, in which case the number of such General electoral districts shall be the whole number next above that quotient:

(d) The quota for the North Island shall be ascertained by dividing the number of General electoral population of that Island by the number of General electoral districts in that Island, as ascertained under paragraph (c) of this subsection…

And Section 45 (3) of the Act:

(a) The Maori electoral population of New Zealand shall be divided by the quota for the General electoral districts in the South Island determined pursuant to section 35 (3)

(b) of this Act, and the quotient so obtained shall be the number of Maori electoral districts:

(b) Where the quotient includes a fraction, the fraction shall be disregarded unless it exceeds a half, in which case the number of Maori electoral districts shall be the next whole number above the quotient…

The following table lists the population figures for the North and South Island General electorate and the Maori electorate, as well as the number of electorates and the electoral district quotas as calculated by the 2001-2002 Representation Commission [12]:

 Elections New Zealand Table

Each electorate must have an electoral population that is within +/-5% of its appropriate electoral district quota. Appendix B lists the population of each of the 69 electorates (the 16 South Island electorates, the 46 North Island electorates and the 7 Maori electorates) in the 2002 Redistribution Plan, along with each district’s percent deviation from the quota.

Redistribution Procedures

There are ten clearly identifiable stages in the redistribution process in New Zealand:

  1. The number and distribution of the total population is derived from the five-yearly census. The analyses of this data can take up to a year after census night.
  2. The Maori option is held over a four-month period beginning shortly after the census. Persons of Maori descent are given the opportunity of deciding on which electoral roll, Maori or General, they wish to register.
  3. Data from the census and the Maori option are combined to allow the Government Statistician to calculate the General electoral population for each of the North and South Islands, the number of General electorates the North Island is entitled to (the South Island has a fixed number of seats), and the Maori electoral population and number of electorates.
  4. The Surveyor-General distributes the General and Maori population data across the existing electorates using small statistical units called mesh blocks. Each mesh block usually contains up to 200 people. Several alternative sets of “provisional” electorate boundaries are produced using these mesh blocks as the basic building units.
  5. The Surveyor-General convenes the Representation Commission and presents the Commission with the provisional electoral boundaries he has developed and explains and justifies his proposals. The Representation Commission has six months from the date of its first formal meeting to complete its work and publish its final plan.
  6. Before commencing its detailed scrutiny, the Representation Commission invites the political parties represented in parliament, and any independent MPs, to make submissions. Because the provision boundaries are confidential at this point, the submissions tend to focus on how each party thinks the rules of redistribution should be interpreted.
  7. Once submissions have been completed, the Commissioners examine the draft plan and vary it where necessary. Although confidentially is still demanded at this stage, the “political” appointees are permitted to discuss the proposals with a very small number of people from the parties they represent as the Commission develops its proposed plan.
  8. Maps of the Commission’s proposed plan are published along with a summary of the reasons for the Commission’s initial decisions, and public comment is invited. Any individual or organization has one month to object to the proposed boundaries and to suggest alternative boundaries. The Representation Commission publishes a summary of all of the objections it receives. Following this, the public then has two weeks to make counter-objections [13]. Objections come from a variety of sources: political parties, individual MPs, statutory and ad hoc authorities, community groups, individual electors and, occasionally, administrators involved in running elections.
  9. Once the counter-objection period has closed, public hearing are held wherever there are a sufficient number of objectors, usually in the larger population centers [14]. These hearings allow the public to put their objections directly to the Commission.
  10. The Commission’s proposed boundaries are then reconsidered in light of the objections and counter-objections, and the definitive electorates are determined. Detailed maps of the electorates covering all parts of the country and legal descriptions of each electorate are prepared to accompany the Commission’s report. Publication of the report marks the conclusion of the redistribution process. (The Commission remains in existence, however, until the night of the next five-yearly census.)

Although not formally part of the redistribution process, a complete re-registration of eligible electors takes place after the Representation Commission has announced its final decisions. Registration as an elector is compulsory and the State, through its agency, the Electoral Enrolment Centre, re-allocates all registered electors to their new electorates. A re-registration card is sent to each elector; its completion and return confirms re-registration.

Computerizing the Process Redistributions were traditionally carried out by physically drawing boundaries on paper maps and manually calculating electoral populations. In 1998, computer technology was used for the first time. All Commissioners were provided a laptop computer loaded with GIS software, and population and statistical data as well as the boundaries of existing electoral and administrative boundaries. Using the GIS software provided, Commissioners were able to develop various options interactively, with immediate feedback as to the effects of these changes. According to the report published by the Commissioners:

The use of this technology reduced the time that otherwise would have been taken by the Commission to formulate proposed boundaries and then to reach final decisions. It also enabled objectors and counter objectors to gain a clear appreciation of the matters being discussed by them at the hearing of objections and counter objections.

Conclusion

Overall, the timely, efficient and professional approach taken to the redistribution of electorates in New Zealand has resulted in general acceptance of the process and the outcome.

Even before replacing the FFP electoral system with an MMP system, redistributions in New Zealand were not considered particularly politically charged. The reason for this is at least threefold: 

  • New Zealand’s politicians have a very limited role in the redistribution process: they do not draw the electorate lines, nor do they have a vote on the plan to be enacted, nor can they prevent a redistribution from being implemented.
  • The membership on the Representation Commission is dominated numerically by non-political appointees. The “political” appointees to the Commission cannot outvote the non-political members.
  • The decisions of the Representation Commission have the force of law and cannot be challenged. 

On the other hand, while the redistribution process is designed to ensure that partisan influence does not dominate, the presence of political appointees on the Representation Commission guarantees that political input is not ignored. This is important if partisan bias, however unintentional, is to be minimized.

More generally, the success of the redistribution process can be attributed in large part to two important factors: the establishment of an independent Representation Commission – with a majority of the members being politically neutral public servants – and a detailed set of rules governing the mechanics of the process. As one commentator concludes:

The success of New Zealand’s redistribution procedures rests squarely on a number of inter-related pillars. The regularity and frequency of redistributions, over which a government has no control, limits keeps any distortions stemming from changes in population distribution to a minimum. These five-yearly revisions guarantee that nearly every election will be fought within electoral district boundaries that meet the prescribed population criteria. The rules governing redistributions are also clearly defined, and while there is some flexibility in applying the discretionary criteria, the mandatory arithmetic criterion and the accompanying narrow tolerance range severely restricts the opportunity to manipulate electorate boundaries in the interests of any particular group or party. Further, the opportunities for public and party input, and the requirement that the Representation Commission explains clearly the reasoning behind its proposals and its ultimate decisions, makes the redistribution process very transparent. And, at the conclusion of the redistribution process, the automatic application of the commission’s final decisions prevents any aggrieved party, community group, or individual, from challenging those decisions in an attempt to prevent their implementation [16].

Perhaps the most unique element of New Zealand’s redistribution process, the establishment of separate Maori electorates overlaying the General electorates, is also its most important element – at least with regard to “lessons to be learned.” This feature has served to promote fair and effective representation for the country’s indigenous minority population.

Notes:

[1] This case study was written by Alan McRobie for the Administration and Cost of Elections (ACE) Project. It was updated by Lisa Handley, in large part based on a paper by Alan McRobie entitled “An Independent Commission with Political Input: New Zealand’s Electoral Redistribution Practices” prepared for the conference “Redistricting from a Comparative Perspective” held at the University of California at Irvine, 6-8 December 2001. The website of Elections New Zealand (found at www.elections.org.nz) was also very useful in updating this case study. Updated materials provided courtesy of the USAID sponsored Delimitation Equity Project.

[2] In New Zealand the terms “electoral districts” and “electorates” are used interchangeably. These terms are the equivalent of “districts”, “constituencies” and “ridings” as used by other countries.

[3] While the non-Maori electorates were redistributed frequently and at regular intervals by the Representation Commission, any changes in the boundaries of the Maori electorates were made by government proclamation, and this occurred only rarely.

[4] “Electoral redistribution” is the term used in New Zealand to describe the delimitation of electoral districts.

[5] No amendments can be made to a reserved provision unless 75% of all MPs, or a majority of voters approve the change in a referendum.

[6] Outlying islands such as Stewart Island and the Chatham Islands are, however, included in one of the mainland electorates.

[7] But in producing the first set of electorate boundaries following the move the MMP electoral system, when the number of electorates declined from 97 to 65, other criteria played a more important role. For example, because of the rugged terrain – the South Island’s main axial ranges are nearly 400 miles long and have numerous peaks over 7,500 feet, and the North Island ranges, although lower, are still formidable – and the impact this landscape has on transportation and communication across the country, topography (and to a lesser extent, communication) played a significant role in redistributions.

[8] Report of the Representation Commission 2002, pages 6-7. The Report is posted online by Elections New Zealand under the section “How electoral boundaries are drawn.”

[9] A number of submissions from political parties and individual objectors have supported increasing the tolerance, and Parliament has considered this issue, but to date no change has been made; +/- 5% remains the electoral quota tolerance.

[10] The 1992 Representation Commission was re-activated in 1993 to reduce the number of electorates from 97 to 65. The 1998 Commission simply redrew the 65 districts created by the 1992 Commission.

[11] In 1977, the number of electorates left unchanged was five out of 88, in 1983 it was six out of 91, in 1987 it was 14 out of 93 and in 1992 it was 27 out of 99.

[12] This table is from the Report of the Representation Commission 2002 which was posted online by Elections New Zealand under the section “How electoral boundaries are drawn.”

[13] The Commission received 199 objections and 80 counter-objections to the proposed boundaries released in November of 2001.

[14] The Commission held hearings in Wellington, Christchurch, Auckland, Hamilton and Te Awamutu during the 2001-2002 redistribution.

[15] Report of the Representation Commission 2002, page 16. This report is posted online by Elections New Zealand under the section “How electoral boundaries are drawn.”

[16] Alan McRobie, “An Independent Commission with Political Input: New Zealand’s Electoral Redistribution Practices” paper prepared for the conference “Redistricting from a Comparative Perspective” held at the University of California at Irvine, 6-8 December 2001.

Singapore: Drawing Districts to Ensure Super-Majorities in the Parliament

 

Singapore is a parliamentary democracy modeled on the British Westminster system. The government has been controlled by the ruling People's Action Party (PAP) since independence from Britain in 1959. This uninterrupted reign of power stems from two key factors: First, PAP’s prudent economic management has moved Singapore from an economic backwater to one of the wealthiest countries in the world; second, the PAP has a low tolerance for opposition and micromanages elections to ensure resounding majorities in the Parliament. While voting is widely considered to be “fair, accurate, and free from tampering,” [1] the development of a robust opposition has been hampered by limits on basic freedoms associated with democratic practices and by the Government’s control and use of the electoral process as an instrument for political dominance.

Electoral System

Singapore’s parliament is unicameral. Until 1988, members were elected by universal and compulsory suffrage in single-member plurality constituencies (SMCs). Increases in population were accounted for by steady growth of the number of seats contested (from 58 in 1968 to 84 as of 2001).

In 1988, amendments to the Parliamentary Elections Act [2] established a mixed system in which some MPs are elected in SMCs while others are elected through a “Party Block Vote” in multi-member Group Representative Constituencies (GRCs) [3]. The purpose of the Party Block Vote is to promote better representation from minority communities. Parties contesting a GRC must propose a slate that includes at least one member of an official minority (listed as Indian, Malay, Eurasian, or Other) [4]. Within the GRCs, voters select from among closed party lists, with the party receiving a plurality of votes winning all seats in the district [5].

Revisions in the electoral law have increased both the number and size of the GRCs [6] and reduced the number of SMCs. As of 2001, nine seats were elected in SMCs, and 75 were elected from the GRCs. Table 1 details the size and composition of the constituencies.

Table 1: Single-member and Group Constituencies in 2001

 

Group Constituencies

 

 

 

Single-member Constituencies

Name

MPs

Electors

 

Name

MPs

Electors

Jalan Besar

5

100,268

 

Potong Pasir

1

16,616

West Coast

5

110,779

 

Ayer Rajah

1

18,475

Bishan-Toa Payoh

5

114,621

 

Joo Chiat

1

21,745

Jurong

5

115,113

 

MacPherson

1

22,010

Holland Panjang

5

118,834

 

Nee Soon Central

1

22,975

Aljunied

5

125,115

 

Hougang

1

23,320

Tampines

5

125,432

 

Chua Chu Kang

1

24,863

Hong Kah

5

129,073

 

Bukit Timah

1

26,951

Pasir Ris-Punggol

5

134,151

 

Nee Soon East

1

28,465

Marine Parade

6

140,174

 

 

 

 

Tanjong Pagar

6

141,150

 

 

 

 

East Coast

6

144,012

 

 

 

 

Sembawang

6

166,137

 

 

 

 

Ang Mo Kio

6

166,644

 

 

 

 

The GRC has several important political implications. First, whichever party wins the most votes in a GRC wins all five or six seats. As a result, elections routinely produce a disproportional seat allocation that always favors the PAP. In 2001, for example, the PAP won 75.3% of the national vote yet netted 82 of 84 parliamentary seats. As one analysis notes: “Over the last four general elections, the opposition has gained an average of 30 percent of the vote in contested seats, but this has led to only between 1.2 percent and 4.9 percent of the parliamentary seats.” [7]

Second, from the opposition’s perspective, the GRCs “dilute the force of personality of party leaders and also present the problem of fielding competent teams with minority representatives.” [8] As a consequence, opposition parties are often unable to field a full slate of candidates, and many of the GRCs are contested only by the PAP. In 2001, for example, PAP ran unopposed in nearly two-thirds of the constituencies. A related concern is the fact that the GRC also shields weaker PAP candidates (i.e., those who lack the dynamism or force of personality to compete strongly) through team membership [9].

Finally, the GRC system is routinely gerrymandered by the national elections department. Opposition parties complain that redistricting nearly always results in better prospects for PAP candidates as competitive districts are generally dismembered to ensure PAP dominance. Since the redistricting process (and in fact the elections department itself) is not independent from the government or overseen by the judiciary, no checks are in place to prevent abuse of the system. Following the 2001 elections, Amnesty International charged that “The small and poorly funded opposition parties complained that constituency changes and a range of regulations imposed by the PAP made it more difficult for them to win votes.” [10]

Table 2: Election Outcomes since introduction of GRCs

 

Evolution of the Electoral System in Singapore

 

Year

Total No. of Seats

No. SMCs

No. GRCs

No. GRC Seats

% PAP National Vote

% PAP Seats

 

 

 

 

 

 

 

1988

81

42

13

39

63.2

98.8

1991

81

21

15

60

61

95.1

1997

83

9

15

74

65

97.6

2001

84

9

14

75

75.3

97.6

Legal Framework for Redistribution

Singapore’s elections are governed by the Constitution and the Parliamentary Elections Act (last revised in 2001). All elections staff, including those responsible for redistricting, are appointed by the government [11]. Elections are administered by civil servants in the “Elections Department” which reports directly to the Prime Minister. There is no independent elections commission [12].

Boundary Authority

Prior to each election, the Prime Minister appoints a five-member “Electoral Boundaries Review Committee,” staffed solely by civil servants from the Elections Department. There are no restrictions concerning when the Prime Minister may call for the map to be redrawn. In the past, however, redistricting has always occurred immediately prior to an official announcement of an election.

Upon completion of the review, the Committee submits a report to the cabinet detailing modifications to constituencies and the drawing of boundary lines. The report is accepted by the government without need for debate or approval of the parliament or oversight by the courts [13]. Once approved, the new map is published in the Government Gazette.

Redistribution Criteria

The criteria and process for boundary review is not spelled out in the Parliamentary Elections Act or in any formal legal framework. The only constant across Singapore’s electoral history appears to be a 30% limit on district population deviation [14].

The basis for redistricting appears to be the ethnic distribution of the population. Given Singapore’s small size and high population density, the majority of Singaporeans live in government-built-and–managed, high-rise apartment buildings. The Housing Development Board (HDB) effectively determines where ethnic groups live by actively limiting the number of ethnic groups in each apartment complex. According to one analysis: “The HDB stipulates that only a certain percentage of each ethnic minority group (not more than the national percentage, that is, approximately 25 percent for Malays, 5 percent for Indians and 1 percent for Eurasians) can live in a particular housing estate. Buyers and sellers of the flats must seek permission from the HDB before a transaction can be made.” [15] As a consequence, the government effectively determines the ethnic distribution of the entire country, making it difficult for minority communities to form a plurality in any one electoral district.

Redistricting for the 2001 and 1997 Elections

2001 The general election of 2001 also brought accusations of gerrymandering by opposition parties. The Boundaries Commission submitted its recommendations for electoral district changes only a day before the announcement of the general election. The new map strengthened PAP electoral support in both GRCs and SMCs across the country [16]. It also eliminated four-person GRCs, leaving only five- and six-person GRCs, which had the dual effect of making it more difficult for opposition groups to come up with enough candidates to contest a district and guaranteeing PAP an extra seat in every district that had previously been a four- person GRC [17].

Protests against the 2001 elections were held almost immediately following the publication of the new districts. The Singapore Democratic Alliance, a coalition of four opposition parties, claimed that the new map guaranteed the PAP a victory even before voting took place. The protests spurred several small riots as demonstrators became increasingly indignant over the predetermined outcome [18].

1997 Three months prior to the general election of February 1997, Prime Minister Goh Chok Tong called upon the Electoral Boundaries Review Committee to redraw the constituency map. Chaired by Wong Chooi Sen, a Cabinet Secretary and PAP party loyalist, the Committee made sweeping changes to all but four electoral districts, despite census data indicating that significant population changes had occurred in only 10 SMCs and three GRCs. The changes were accepted by the Government and implemented shortly before the polls opened.

The Committee created six new GRCs, enlarged six, made two smaller, left one unchanged, and eliminated or renamed another six. Fourteen of the previous twenty-one SMCs were merged into new GRCs or absorbed by pre-existing GRCs. The election law was also amended to raise the maximum number of representatives in a GRC to six and reduce the minimum number of SMCs to eight seats. In the end, the total number of GRCs remained unchanged at 15, and the total number of SMCs shrunk from 21 to nine. The net result was an increase in the percentage of seats in parliament filled by GRCs to 89%, up from 74% in the previous election, and a decrease in the percentage of seats filled by SMCs to 10.8%, down from 25.9% in the previous election.

Some of the most politically significant changes occurred in the six districts that were either eliminated completely or chopped up and renamed. The Eunos district was eliminated completely, with its electorate split into the surrounding districts. In the previous general election, Eunos had experienced the closest race of any of the GRCs, with the PAP winning a narrow victory over the Worker’s Party (WP) by a vote of 45,833 to 41,673. Similarly, the Bedok constituency, where the PAP won over the WP by a vote of 49,109 to 30,121, was renamed East Coast and parts of five other districts were added to it. Three of those districts contained significant PAP support, thus weakening WP’s base. The tactic proved successful as the PAP ran uncontested in East Coast in the 1997 elections.

All four of the SMCs held by opposition party members prior to the election were left as SMCs. However, the boundaries of one of the SMCs were substantially changed. Nee Soon Central, an SMC held by the Singapore Democratic Party (SDP) prior to the 1997 elections, was added to the Ang Mo Kio GRC. As a result, the SDP lost the seat in Nee Soon Central to the PAP by a vote of 9,591 to 15,214. The Ang Mo Kio GRC, which absorbed a portion of Nee Soon Central, was uncontested. Thus, the PAP effectively eliminated a strong source of opposition support in one district by diverting a substantial portion of its population to a district where no opposition party even contested the election [19].

Conclusion

Singapore has a functioning democratic system in which voters are freely able to elect their preferred representatives. However, tight government control of the electoral process, combined with other tactics designed to harass opposition parties, results in these elections being less than free and fair. Given widespread popular support for PAP’s prudent management of the city-state, these anti-democratic practices do not serve any identifiable purpose except to provide the PAP with a consistent super-majority. Even without the obvious gerrymandering, the PAP would likely win a substantial majority of votes for the foreseeable future.

In terms of the boundary delimitation process, several key flaws are notable:

 

 

  • Lack of a legal framework: The redistricting process is not governed by a consistent legal framework. While the Parliamentary Elections Act mandates the creation and composition of SMCs and GRCs, it does not specify any criteria by which the districting process should occur.
  • Lack of independence of electoral officials: The government controls both the Elections Department and the Electoral Boundaries Review Committee. Redistricting is conducted in secret, with no public input or oversight, and the EBRC staff is appointed by the Prime Minister’s office, compromising its neutrality.
  • Lack of transparency: Opposition parties and civil society groups have no role in the delimitation process. Once delimitation is completed, no external authority is responsible for approval of the new electoral map. The Courts are excluded from the process entirely.
  • Time Constraints on the Opposition: The government routinely publishes updated electoral boundaries only weeks before an election. This hinders the ability of opposition parties to recruit candidates and propose slates in newly-created districts. As a result, nearly two-thirds of the GRCs are not even contested during the polling [20].
  •  

     

 

Appendicies

Notes:

[1] US State Dept. Annual Human Rights Report.

[2] (s8A) The full Act is available at http://agcvldb4.agc.gov.sg/

[3] The Party Block Vote is also utilized in Djibouti, Lebanon, Tunisia, Ecuador and Senegal. According to International IDEA, “[t]he advantages of the Party Block Vote are that it is simple to use, encourages strong parties and allows for parties to put up mixed slates of candidates in order to facilitate minority representation. However, a critical flaw of the Party Block is the production of super-majoritarian results, where one party can win almost all of the seats with a simple majority of the votes.” See http://www.idea.int/publications/esd/esd-blockvote.html

[4] Each GRC is categorized based on whether the minority member represents the “Malay” or the “Indian and Other” minority communities PEA, 8A1b, http://agcvldb4.agc.gov.sg/

[5] Constituent responsibilities in the GRCs are left up to the discretion of the group. Generally, a district is divided into several wards, with each member of the elected group being assigned a ward to specifically represent. However, despite dividing the districts into wards, each member of the group is legally responsible for representing the entire district.

[6] In 1991, for example, the number of GRCs was increased to 15 and SMCs reduced to 21. In 1997, 15 GRCs were scaled into four-, five-, or six-member constituencies, and SMCs were reduced to only nine. For 2001, all GRCs were scaled to five or six members.

[7] Mauzy, Diane K., “Electoral Innovation and One-Party Dominance in Singapore.” In John Fuh-Sheng Hsieh and David Newman (eds.), How Asia Votes. (London: Chatham House Publishers, 2002): 235 – 254.

[8] Mauzy: 244.

[9] http://www.fesspore.org/pdf/Electoral%20Politics/Singapor.pdf p. 206

[10] http://www.singapore-window.org/sw02/020528ai.htm

[11] PEA 3(1).

[12] http://www.fesspore.org/pdf/Electoral%20Politics/Singapor.pdf p.209

[13]http://www.fesspore.org/pdf/Electoral%20Politics/Singapor.pdf p.210

[14] Warren Fernandez, “15 GRCs, 9 SMC’s for Election,” The Straits Times, 22 November 1996, http://ourstory.asia1.com.sg/dream/politics/ref/grc2.html; According to previous accepted practice, there can be no more than a 30% deviation in population among GRCs and no more than 30% deviation in population for SMC’s.

[15] http://www.thinkcentre.org/article.cfm?ArticleID=1435

[16] John Burton, “Why Bother Voting? The Government is Almost Unopposed,” The Economist, 1 November 2001, available at http://www.sfdonline.org/Link%20Pages/Link%20Folders/01Pf/econ011101.html; see also Appendix A.

[17] “Opposition Faces New Obstacle after Authorities Slap New Election Rules,” The New Straight Times, 21 October 2001, available at http://www.sfdonline.org/Link%20Pages/Link%20Folders/01Pf/nst211001.html.

[18] “Riots break out on eve of Singapore General Election,” Japan Newswire, 2 November 2001.

[19] Fernandez, “15 GRCs, 9 SMC’s for Election”; “Changes to the Electoral Map,” available at http://ourstory.asia1.com.sg/dream/politics/ref/images/elecmap.jpg; “History of Singapore Elections.” eCitizen, available at http://www.ecitizen.gov.sg/inc_frame.htm?link=http://www.elections.gov.sg. See also Appendix C.

[20] In 2001, 55 out of the 84 parliamentary seats ran uncontested, thus guaranteeing a PAP majority regardless of the outcome of the elections; see also Appendix B.

[21] “2001 General Election Results,” eCitizen, available at http://www.ecitizen.gov.sg/inc_frame.htm?link=http://www.elections.gov.sg

 

The United States of America: Reapportionment and Redistricting

The redistricting process in the United States can be distinguished from redistricting elsewhere in the world in at least two very fundamental ways: the extent to which the process is overtly and acceptably political – legislators still have the responsibility for drawing electoral districts in most states – and the degree to which the American courts have intervened in the process. These two characteristics of the process are interconnected and have meant that redistricting is often contentious and that the result may be biased in favor of one political party over the other party.

Although redistricting is often driven by partisan concerns, the overall result in the U.S. Congress does not necessarily favor one party because the process is quite decentralized – each of the fifty states is responsible for drawing its own electoral district boundaries. The federal government allocates congressional seats to the states (a process referred to as reapportionment) on the basis of total population, but has only a limited role to play in the redistricting process after this point. For example, the U.S. Congress can pass laws to regulate the process to an extent and has done so on rare occasion – the Voting Rights Act of 1965 is one notable example. On the other hand, the federal courts (as well as the state courts), play a major role in the redistricting process in the United States. They have been called upon to develop redistricting standards, to arbitrate redistricting conflicts, and even to draft redistricting plans [1].

Apportioning Congressional Seats to the States

The number of members of the U.S. House of Representatives is prescribed by law and has been set at 435 members since 1912. Each state is entitled to at least one representative and the remaining members are apportioned among the states in accordance to their relative populations. The population used for this process is the total population (as compared to, for example, the voting age population or the number of registered voters) which is determined by a census that the Constitution requires be taken every ten years.

The apportionment of congressional seats to states is made pursuant to a statutory formula. This population-based formula has changed several times over the course of the last two hundred years.

Reporting the Census The U.S. Constitution stipulates that there shall be a national census conducted every ten years. This census is conducted at the beginning of each decade (1990, 2000, 2010). The total population of each state is reported to the President by December 31 in the year of the census. These enumeration figures are then used to allocate congressional seats to the fifty states. 

  • Formula for Reapportioning Congressional Seats to the States* Over the years, four different apportionment formulas have been used to apportion seats to the states on the basis of population:
  • From 1790 to 1840, Congress used a method proposed by Thomas Jefferson, sometimes called the "method of greatest divisors." This method divided the total population by the number of seats and assigned each state its quota, disregarding any fractional remainder. The number of members was adjusted so that every state was awarded exactly the number of seats it was entitled to on the basis of its quota.
  • From 1842 to 1850, Congress used a formula proposed by Daniel Webster, sometimes called the "method of major fractions." This method gave an additional member to any state whose quota included a fraction greater than one-half. The number of members was adjusted accordingly.
  • From 1850 to 1910, Congress used a formula that had originally been proposed by Alexander Hamilton for the apportionment of 1790. Under that formula, members were first apportioned according to each state's quota, disregarding any fractional remainders, and then any leftover seats were assigned to the states with the largest fractional remainders.
  • Between 1911 and 1930, Congress reverted to using the Webster method.

After the 1930 census, in accordance with a report from the National Academy of Sciences, Congress adopted the "method of equal proportions." The formula uses the state's population divided by the geometric mean of that state's current number of seats and the next seat (the square root of n(n-1)). This formula allocates the remainders among the states in a way that provides the smallest relative difference between any pair of states in the population of a district and in the number of people per representative. Congress's choice of this method over the other possible methods has been upheld by the U.S. Supreme Court and it remains in use today.

Drawing Congressional Districts within States

Each state's quota of representatives must be elected from single-member districts of equal population. Who is tasked with drawing the boundaries, and the criteria that guide the process, varies somewhat from state to state.

Boundary Authority

Since the earliest days of the republic, redrawing the boundaries of congressional districts after the decennial census has been primarily the responsibility of the state legislatures. Only a few states assign the responsibility for redrawing congressional district boundaries to a body other than the legislature. Those states that do not permit the state legislature to draw the congressional boundaries usually delegate the task to a commission established solely for the purpose of redistricting.

Final Authority for Redistricting Plans

Except for the few states that use redistricting commissions, the new districts are drawn by the state legislature and enacted in the form of a bill. The enactment of the bill is subject to whatever public hearing requirements may apply in the state. In almost every state legislature, enactment of the bill is an intensely partisan issue, with the majority party attempting to gain a political advantage through the way the lines are drawn. Partisan gerrymandering is thus a fact of life in most American congressional redistricting. Equal population requirements and other "traditional districting principles" are limits the federal courts and state constitutions have imposed to restrain this tendency to gerrymander (draw districts that favor one political party at the expense of the other).

Role of the Courts

When partisan differences between the two houses of a state legislature, or between the legislature and the governor, look like they may prevent the legislature from enacting a redistricting bill in time for the general election, or when some stakeholder in the process believes the plan drawn by the legislators is illegal or unconstitutional, residents may bring suit in state or federal court and ask the court to correct an enacted plan or adopt a plan if none has been enacted. A federal court must defer to a state court, and both must defer to a legislature that is actively engaged in adopting a plan. However, if the legislature fails to meet reasonable deadlines imposed by the court, the court may impose a redistricting plan of its own, to be effective until adoption of a valid plan by the legislature.

Criteria for Redistricting

Each state has its own constitution and laws, and the constitutional requirements for redistricting vary considerably from state to state. What little there is in the way of national law on the subject has been developed over the years in a series of cases decided by the U.S. Supreme Court. The Voting Rights Act, passed by congress in 1965 and amended in 1982, also places some limits on what states are permitted to do when redistricting electoral boundaries.

The federal constraints that have been imposed on the redistricting process include:

  • A redistricting plan must create districts that are relatively equal in population
  • A redistricting plan must not dilute the strength of minority voters
  • A redistricting plan must not be a “racial gerrymander”
  • A redistricting plan must take into account traditional redistricting criteria such as compactness, contiguity, and respect for political subdivision lines and communities of interest

The courts have played a major role in the development and interpretation of all five of these redistricting criteria.

Equality of Population Following World War I, as the nation's population began to shift from rural to urban areas, many legislatures lost their enthusiasm for the decennial task of redistricting and failed to carry out their constitutional responsibility. Over time many urban areas were denied the political representation their populations warranted because of this failure to redistrict. For decades, the U.S. Supreme Court declined repeated invitations to enter the "political thicket" of redistricting and refused to order the legislatures to carry out their duty.

In 1962, however, in the case of Baker v. Carr, the Court for the first time held that the federal courts had jurisdiction to consider constitutional challenges to redistricting plans. The next year, in Gray v. Sanders, Justice Douglas declared: "The conception of political equality from the Declaration of Independence, to Lincoln's Gettysburg Address, to the Fifteenth, Seventeenth, and Nineteenth Amendments can mean only one thing--one person, one vote." In 1964, in Wesberry v. Sanders, the Court held that congressional districts must be redrawn so that "as nearly as is practicable one man's vote in a congressional election is ... worth as much as another's." Finally, in 1983, in Karcher v. Daggett, the Court developed a standard of equality for congressional districts that required them to be mathematically equal, unless justified by some "legitimate state objective." As a consequence of these court decisions, congressional districts within a single state vary minimally (inevitably less than one percentage point) – although there is still considerable population variation across states because every state, regardless of population, is guaranteed at least one congressional seat.

Equality of Opportunity for Minorities Once the courts took the initial step of striking down redistricting plans for inequality of population, Congress instigated the next step. In 1965, it enacted the Voting Rights Act to provide equality of opportunity for racial minorities to vote. Section 2 of the Act prohibited any state or political subdivision from imposing a "voting qualification or prerequisite to voting, or standard, practice or procedure to deny or abridge the right to vote on account of race or color." Section 5 of the Act required a covered jurisdiction to preclear any changes in its electoral laws, practices, or procedures with either the U.S. Department of Justice or the U.S. District Court for the District of Columbia before it could take effect. The Justice Department began to use this new authority to require that redistricting plans be precleared before they could take effect.

In 1980, in City of Mobile v. Bolden, the Supreme Court said that a redistricting plan would not be found to violate the Fourteenth Amendment or Section 2 of the Voting Rights Act unless the plaintiffs could prove that its drafters intended to discriminate against them. Congress was swift to react to this new limitation on how to prove racial discrimination. In 1982, after most of the redistricting plans based on the 1980 census had already been enacted, Congress amended Section 2 of the Voting Rights Act to make clear that it applied to any plan that results in discrimination against a member of a racial or ethnic minority group, regardless of the intent of the plan's drafters.

How were the courts to determine whether a redistricting plan would have discriminatory results? In the 1986 case of Thornburg v. Gingles, the Court set forth three preconditions a minority group must prove in order to establish a violation of Section 2:

  • The minority group must be sufficiently large and geographically compact to constitute a majority in a single-member district;
  • The minority group must be politically cohesive, that is, it must usually vote for the same candidates; and
  • In the absence of special circumstances, block voting by the White majority usually defeats the minority's preferred candidate.

If the minority group establishes these three preconditions, it is entitled to proceed to the next step: proving a Section 2 violation by "the totality of the circumstances." Those circumstances would have to show that the members of the minority group had "less opportunity than other members of the electorate to participate in the electoral process and to elect representatives of their choice."

What did that mean, "less opportunity?" In North Carolina, where the Gingles case arose, it meant that multimember districts where blacks were in the minority and had been unable to elect candidates to office had to be replaced with single-member districts where blacks were in the majority. To the rest of the country, and to the state legislatures and commissions who were going to be drawing new districts after the 1990 census, it meant that wherever there was a racial or ethnic minority that was "sufficiently large and geographically compact to constitute a majority in a single-member district," the state would have to draw a district for them or risk having the plan thrown out.

Being forewarned of the effects of Section 2, drafters of redistricting plans after the 1990 census went to great lengths to draw majority-minority districts wherever the minority population counts seemed to justify it. In states where redistricting plans could not take effect until they had been pre-cleared by the Justice Department, the Justice Department encouraged the state to draw districting plans that created new districts where members of a racial or language minority group (primarily blacks or Hispanics) were a majority of the population. These new "majority-minority" districts were intended to protect the states from liability under Section 2 for failing to draw districts that the minority group had a fair chance to win.

As states drew and submitted their redistricting plans to the Justice Department, they discovered that the Justice Department had little concern that majority-minority districts be compact in shape. A number of states took advantage of this and redrew plans that included districts with rather bizarre shapes. These districts were labeled "racial gerrymanders" by those opposed to their creation.

No Racial Gerrymandering A number of redistricting plans that included oddly shaped districts were attacked in federal court as “racial gerrymanders” that denied white voters their right to equal protection of the laws under the Fourteenth Amendment to the U.S. Constitution. The U.S. Supreme Court ultimately held that districts that did not conform to traditional redistricting principles could be construed as unconstitutional racial gerrymanders. Such districts had to be subjected to "strict scrutiny" to determine whether they were "narrowly tailored" to achieve a "compelling state interest" in complying with Section 2 – and only if they were narrowly tailored could such districts be declared constitutional. As a result of this holding, a number of majority minority districts were struck down by the federal courts because their drafters had not followed "traditional districting principles," such as keeping districts compact, not splitting political subdivisions, and preserving communities of interest. These states then had to redraw the districts to conform to these traditional districting principles.

Traditional Districting Principles For the 2000 round of redistricting, each state was left to decide for itself what "traditional districting principles" might be and how they were to be applied. Federal law does not identify these principles. However, some “redistricting principles” are required by individual state's constitutions, and other redistricting principles have been adopted by state law.

The districting principles used by each state are shown in the table below.

(This table is a copy of a table that appears as Table 5 in Redistricting Law 2000, a publication of the National Conference of State Legislatures, Denver, Colorado, 1999.) They include requiring that districts be composed of contiguous territory, making districts geographically compact, respecting the boundaries of political subdivisions, preserving communities of interest, preserving the cores of prior districts, and avoiding contests between incumbent representatives.

Table: Districting Principles (in addition to Population Equality) Used by Each State 

 

State

 

Compact

 

Contiguous

 

Preserve Political Subdivisions

 

Preserve Communities of Interest

 

Preserve Cores of Prior Districts

 

Protect Incumbents

 

Voting Rights Act

 

Alabama

 

C, L

 

C, L

 

C, L

 

C, L

 

C, L

 

 

 

C, L

 

Alaska

 

L

 

L

 

L

 

L

 

 

 

 

 

 

 

Arkansas

 

 

 

 

 

C, L

 

 

 

C, L

 

YC, YL

 

C, L

 

Arizona

 

C, L

 

C, L

 

 

 

 

 

 

 

 

 

C, L

 

California

 

 

 

L

 

L

 

 

 

 

 

 

 

 

 

Colorado

 

L

 

 

 

L

 

L

 

 

 

 

 

L

 

Connecticut

 

 

 

L

 

L

 

 

 

 

 

 

 

 

 

Delaware

 

 

 

L

 

 

 

 

 

 

 

NL

 

 

 

Florida

 

 

 

L

 

 

 

 

 

 

 

 

 

 

 

Georgia

 

 

 

C, L

 

C, L

 

 

 

C, L

 

YC, YL

 

C, L

 

Hawaii

 

L

 

L

 

L

 

L

 

 

 

NL

 

 

 

Idaho

 

C, L

 

C, L

 

C, L

 

C, L

 

 

 

NC, NL

 

C, L

 

Illinois

 

L

 

L

 

 

 

 

 

 

 

 

 

 

 

Indiana

 

 

 

L

 

 

 

 

 

 

 

 

 

 

 

Iowa

 

C, L

 

C, L

 

C, L

 

 

 

 

 

NC, NL

 

C, L

 

Kansas

 

C, L

 

C, L

 

C, L

 

C, L

 

C

 

NL

 

L

 

Kentucky

 

 

 

C

 

C

 

C

 

C

 

 

 

C

 

Louisiana

 

L

 

L

 

L

 

 

 

L

 

 

 

 

 

Maine

 

L

 

L

 

L

 

 

 

 

 

 

 

 

 

Maryland

 

C, L

 

 C, L

 

C, L

 

C, L

 

C, L

 

YC, YL

 

C, L

 

Massachusetts

 

 

 

L

 

L

 

 

 

 

 

 

 

 

 

Michigan

 

L

 

L

 

L

 

 

 

 

 

 

 

 

 

Minnesota

 

C, L

 

C, L

 

C, L

 

C, L

 

 

 

 

 

C, L

 

Mississippi

 

C, L

 

C, L

 

C, L

 

 

 

 

 

 

 

C

 

Missouri

 

C, L

 

C, L

 

C

 

C

 

C

 

 

 

 

Montana

 

L

 

L

 

L

 

L

 

 

 

NL

 

L

 

Nebraska

 

C, L

 

C, L

 

C, L

 

C, L

 

 

 

NC, NL

 

C, L

 

Nevada

 

C, L

 

L

 

C, L

 

L

 

 

 

 

 

C, L

 

New Hampshire

 

 

 

L

 

L

 

 

 

 

 

 

 

 

 

New Jersey

 

L

 

C, L

 

L

 

 

 

C

 

 

 

C

 

New Mexico

 

L

 

L

 

L

 

 

 

 

 

 

 

 

 

New York

 

L

 

L

 

L

 

 

 

 

 

 

 

 

 

North Carolina

 

 

 

C, L

 

C, L

 

 

 

C

 

YC

 

C, L

 

North Dakota

 

L

 

L

 

L

 

 

 

 

 

 

 

 

 

Ohio

 

L

 

L

 

L

 

 

 

 

 

 

 

 

 

Oklahoma

 

L

 

L

 

L

 

L

 

 

 

 

 

 

 

Oregon

 

 

 

C, L

 

C, L

 

C, L

 

 

 

NC, NL

 

C, L

 

Pennsylvania

 

L

 

L

 

L

 

 

 

 

 

 

 

 

 

Rhode Island

 

L

 

 

 

 

 

 

 

 

 

 

 

 

 

South Carolina

 

C, L

 

C, L

 

C, L

 

C, L

 

C, L

 

YC, YL

 

C, L

 

South Dakota

 

L

 

L

 

L

 

 

 

 

 

 

 

L

 

Tennessee

 

 

 

L

 

L

 

 

 

 

 

 

 

L

 

Texas

 

 

 

L

 

L

 

 

 

 

 

 

 

C, L

 

Utah

 

C, L

 

C, L

 

C, L

 

C, L

 

 

 

NC, NL

 

 

 

Vermont

 

L

 

L

 

L

 

L

 

 

 

YL

 

 

 

Virginia

 

C, L

 

C, L

 

L

 

 L

 

 

 

 YL

 

L

 

Washington

 

C, L

 

C, L

 

C, L

 

C, L

 

 

 

NL

 

 

 

West Virginia

 

C, L

 

C, L

 

C, L

 

 

 

 

 

 

 

 

 

Wisconsin

 

L

 

L

 

L

 

 

 

 

 

 

 

 

 

Wyoming

 

C, L

 

C, L

 

C, L

 

L

 

 

 

NL

 

L

 

C = congressional districts                 NC = not congressional districts

L = state legislative districts              NL = not state legislative districts

 

Conclusion

The United States is one of very few democracies in which legislators are still permitted to draw electoral boundaries – most countries have reformed the process and reassigned the task of redistricting to independent, non-partisan commissions. Because legislators are responsible for redistricting, the process is often quite partisan and the results politically biased. But there are signs that this could change in at least a few states: a number of states are currently contemplating the adoption of a commission approach to redistricting.

What is less likely the change, at least in the near future, is the very active role that the courts play in the redistricting process. Subjecting a redistricting plan to court scrutiny is one of the very few options opponents to a redistricting plan have to challenge a plan and very few Americans would advocate giving up this right.

Notes:

[1] The courts may draw district boundaries when a state legislature fails to adopt a plan, or adopts a plan that the court has determined does not satisfy legal or constitutional requirements.

Yemen: Assessing the Feasibility of Computer-Assisted Delimitation

 

The Yemeni Supreme Commission for Elections and Referendum (SCER) is considering the use of Geographic Information Systems (GIS) computer technology for constituency delimitation prior to the 2006 elections. Because the Central Statistics Office (CSO) is in the process of producing the necessary electronic database for Yemen, utilizing GIS for delimitation is feasible.

Background

The Republic of Yemen has a First-Past-the-Post electoral system with a bicameral legislature consisting of the Shura Council (a consultative body with 111 members appointed by the President) and a House of Representatives. The 301 members of the House of Representatives are elected from single-member constituencies in plurality elections. The most recent parliamentary elections were held in April 2003[1].

According to official 2003 results, the ruling General People’s Congress (GPC) received 58.2 % of the vote, and won 230 (76.4%) of the 301 seats[2]. In addition to the disparate seats-to-votes ratio, an examination of the election results indicates that, despite a population quota of slightly over 60,000 persons per constituency – the equivalent of approximately 26,700 eligible voters per constituency in Yemen – some constituencies were found to contain more than 50,000 voters[3].

Among the recommendations offered by such international non-governmental organizations as National Democratic Institute (NDI) and IFES following the 2003 parliamentary elections was that constituency boundaries be redrawn following the 2004 Yemeni census. This is necessary to bring constituency populations into compliance with Yemeni constitutional and electoral law[4].

The 2002 Delimitation Process

In May and June of 2002 (in anticipation of the 2003 parliamentary elections), the Supreme Commission for Elections and Referendum (SCER) delimited Yemen into the requisite 301 parliamentary constituencies. The Constitution (Article 63) requires constituencies[5] “equal in number of population with a variation of not more than 5% plus or minus.”[6] In addition, the electoral law obliges the SCER to take into account “geographic and social considerations” when creating parliamentary constituencies[7].

According to the SCER, the factors considered during the 2002 delimitation – besides population – included geographical features that formed natural barriers, such as mountains and rivers, and the following social factors:

 

 

  • administrative subdivisions referred to as ozal (singular: ozla) were not divided between constituencies, if possible;
  • villages with an historical animosity towards one another (tha’ar) were not included in the same constituency, if possible[8].
  •  

     

The SCER began the delimitation process by dividing each of the 332 administrative divisions (modiriya) into local constituencies [9]. The only established legal criterion for the delimitation of local constituencies is that the populations of the constituencies within an administrative district should not vary by more than plus or minus 5% from the population quota for that district[10].

The local constituencies usually perform the dual purpose of electing representatives to the local government councils (there are 332 local government councils – one for each administrative division) and serving as voting centers for casting and counting ballots. However, it appears that in some administrative divisions, local constituencies have been combined to create a single voting center, while in other administrative divisions local constituencies have been divided into more than one voting center[11]. See Table 1 below for the number of local constituencies and the number of voting centers by governorate.

Table 1: Number of Administrative Divisions, Parliamentary and Local Constituencies, and Voting Centers by Governorate

 

Governorate

Number of Administrative Divisions

Number of Parliamentary Constituencies

Number of Local Constituencies

Number of Voting Centers

Sana’a City

10

19

179

179

Aden

8

10

130

130

Taiz

23

39

494

488

Lahj

15

12

280

282

Ibb

20

36

488

490

Abyan

11

7

197

194

Al-Baidha’a

20

10

285

291

Shabwa

17

6

211

199

Al-Mahra

9

2

74

74

Hadhrmout

30

18

365

365

Al-Hodeida

26

34

515

515

Thamar

12

21

297

297

Sana’a

21

20

423

414

Al-Mahweet

9

8

167

167

Hajja

30

20

521

515

Sa’ada

15

9

181

181

Al-Jawf

12

5

160

160

Ma’areb

14

3

138

138

Amran

20

15

354

369

Al-Dhale’a

9

7

161

172

TOTAL

332

301

5,620

5,620

The 5,620 local constituencies/voting centers created by the SCER were then used as building blocks for drawing the 301 parliamentary constituencies. These 301 single-member constituencies were used to elect representatives to Parliament in April 2003.

The SCER used paper maps (topographical maps of the administrative divisions) with acetate overlays to draw the local and parliamentary constituency boundaries. Estimates of constituency populations were based on projections of the 1994 census. As a result of employing these manual techniques, the only maps that exist of the current local constituency boundaries are the acetate overlay maps that are housed at SCER headquarters. Moreover, the population estimates for the constituencies are only rough approximations of the actual population contained within the constituency boundaries [12].

Once provisional maps of the constituencies were completed, field work was conducted to ensure that the constituency boundaries took into account local geographic and social considerations. Local officials were asked to review the maps, solicit comments and offer suggestions if the boundaries were problematic. As a result of this field work, some changes were made to the provisional constituency boundaries. The final constituency boundaries were established by the SCER in the summer of 2002.

The entire delimitation process took approximately 3 months and involved about 60 SCER staff members and as many as 900 officials in the field. The cost of the 2002 delimitation was estimated by the SCER to be approximately US$2 million.

The 2005 Delimitation Process

The SCER anticipates redrawing constituency boundaries following the release of new census data in mid-2005 [13]. Local elections are to be held in September 2006; therefore the delimitation process must be completed by the end of 2005 (before voter registration for the 2006 election commences).

The SCER hopes to have to make only “minor” or “technical” adjustments to the 2002 constituency boundaries to comply with the requirement that constituency populations deviate no more than plus or minus 5%. Assuming only minor adjustments, the SCER has indicated that it does not plan to go back into the field to solicit comments on the constituency boundaries.

There are at least two reasons to be sceptical about only minor adjustments to constituency boundaries being required: First, relating the 2004 census data to the 2002 constituency boundaries is liable to uncover substantial population deviations across constituencies since the SCER had to use less-than-reliable (especially at lower levels of geography) projections of the 1994 census for their population data. Second, if the Ministry of Local Authorities revises the boundaries of the 332 existing administrative divisions, at minimum, local constituencies will have to be redrawn for any reconfigured administrative divisions [14].

The SCER would like to utilize GIS for the upcoming delimitation process. There are a number of reasons for supporting the SCER in this effort:

 

 

  • GIS technology could generate a more efficient, accurate, and cost-effective (at least in the long term) delimitation process.
  • GIS technology could assist the SCER in meeting such constitutional and legal delimitation requirements as equal population and consideration for geographic and social factors when drawing constituency boundaries.
  • GIS technology could foster greater transparency in the delimitation process by permitting the easy production of maps and reports that can be used by interested stakeholders to evaluate and comment on proposed constituency plans.
  • GIS offers the potential for producing a “fairer” constituency plan – one that optimizes established delimitation criteria such as population equality and consideration of geographic and social factors at the expense of other, less appropriate, factors such as the potential political consequences of the plan.
  • GIS may have additional uses in election administration: for example, GIS can be used to assign eligible voters to the correct voting center.
  •  

     

Using GIS for Delimiting Constituencies in Yemen

The SCER utilized manual techniques – paper maps with acetate overlays, colored markers, and calculating machines – for the 2002 delimitation exercise. Although the process was completed in a timely manner, it required a very large staff to accomplish this. Furthermore, very little information about the constituencies was produced, and even less information was made available to interested stakeholders (i.e., members of Parliament, political parties, NGOs and interested voters).

Adopting GIS technology would expand the information available to the SCER. It would allow the SCER to:

 

 

  • create constituency plans much more rapidly: a plan would be created interactively by assigning geography to constituencies piece by piece and seeing the results of the assignment displayed on the computer screen immediately;
  • produce maps – both on the screen and, if desired, on paper – of the constituency plan as each piece of geography is assigned and, of course, once a plan is completed;
  • generate statistical reports summarizing delimitation plans for evaluation purposes.
  •  

     

Creating an Electronic Database If GIS technology is to be used for drawing district lines, an electronic database must be created. This database must include, at a minimum, population data (i.e., census enumeration data or voter registration data) and the maps associated with the geographic units for which the population is reported. As the Yemeni constitution requires the “population” of constituencies to be equal (and not, for example, the number of voters or registered voters), census data and maps of the census enumeration areas will have to be included in the electronic delimitation database.

The most common obstacle to using GIS for delimitation purposes is the lack of computerized maps for the relevant geographic units. This will not pose a problem in Yemen, however: the Central Statistics Office (CSO) is currently in the process of digitizing census geography down to the smallest unit for which census data will be reported, the census enumeration area. If the SCER is willing to use enumeration areas as the building blocks for creating constituencies, then computerized maps and the associated population data will presumably be available from the CSO for use by the SCER. In addition to maps of the census enumeration areas, other maps that would prove useful in delimiting constituencies in Yemen include: administrative boundaries (such as governorate boundaries, administrative division and subdivision boundaries), major physical features such as mountain ranges and rivers, and existing constituency boundaries.

Some of the administrative boundaries have been embedded in the CSO database; for example, governorate and administrative division boundaries can be found within the database [15]. In addition, some physical features will be demarcated in the CSO database.

Incorporating existing constituency boundaries into the delimitation database, however, will be one of the most challenging tasks facing the SCER. The boundaries of the current local constituencies can be added to the electronic CSO database by either (1) assigning entire census enumeration areas to constituencies in a manner that approximates current local constituency boundaries but does not follow the constituency boundaries exactly, or by (2) re-creating the local constituency boundaries precisely by electronically “splitting” census enumeration areas and estimating the associated population for the split portions of the census enumeration area [16]. The latter approach is more time-consuming, and will produce only estimates of the population encompassed within a constituency. On the other hand, only the latter approach will produce an exact delineation of the current constituency boundaries.

Drawing New Constituency Boundaries Once the electronic delimitation database has been prepared, including the delineation of the current constituency boundaries by one of the means described above, a new delimitation plan can be drawn. This is accomplished by moving geographic units (census enumeration areas) from one constituency to another until all of the constituencies in the plan meet the predetermined districting criteria. GIS technology would speed up this process enormously by re-tabulating the population automatically, and instantly, each time a new assignment of territory is made and by displaying the results of the tabulation, along with the new constituency map, on the computer screen.

Evaluating Constituency Plans If the SCER employs GIS to delimit constituencies, the software will facilitate the production of maps and reports, which can then be used to determine compliance with such criteria as:

 

 

  • population equality;
  • geographic considerations such as mountains, rivers and other physical features that form natural barriers;
  • respect for existing governmental units such as administrative divisions, as well as administrative subdivisions if these are digitized.
  •  

     

If GIS software is used to create a constituency plan, then producing a statistical report listing the population of each constituency, as well as the percent by which that constituency’s population deviates from the population quota, is a very simple matter. GIS technology also permits the overlay of maps displaying administrative division boundaries and physical features such as mountain ranges and rivers on the map of the constituency boundaries. These reports and maps could serve as tools for the SCER to evaluate any proposed constituency plans. Furthermore, if the SCER released these reports and maps, interested stakeholders could also evaluate proposed constituency plans.

Advantages and Disadvantages of Using GIS Using GIS offers a number of important benefits, all of which have been discussed above. But there are drawbacks to using GIS – and it is important to consider both the advantages and the disadvantages of GIS when contemplating using GIS for delimitation. The table below lists some of the major advantages and disadvantages associated with GIS:

Advantages of GIS

 

 

  • GIS technology may produce a more efficient, accurate, and more cost-effective delimitation process.
  • GIS technology could assist the SCER in meeting such constitutional and legal delimitation requirements as equal population and consideration for geographic and social factors when drawing constituency boundaries.
  • GIS technology could foster greater transparency in the delimitation process by permitting the easy production of maps and reports that can be used by interested stakeholders to evaluate and comment on proposed constituency plans.
  • There is potential for producing a “fairer” constituency plan – one that optimizes established criteria such as population equality – by using the evaluation tools found in most GIS packages.
  • GIS may have additional uses in election administration: for example, GIS can be used to assign eligible voters to the correct voting center.
  •  

     

Disadvantages of GIS

 

 

  • The financial cost associated with acquiring GIS capabilities for delimitation may be prohibitively high.
  • It may be difficult to find and train qualified personnel to operate the GIS software – which has important implications not only for setting up the system but for sustaining it as well.
  • Poorly managed GIS could result in a delimitation process that is actually less efficient, less effective and less timely than a process employing manual techniques (at least in the short run).
  • Improperly used, GIS could allow users to manipulate the delimitation process, and ultimately, the outcome of elections.
  •  

     

Cost of Using GIS: Hardware, Software and Staffing The cost of using GIS for delimitation varies dramatically depending on the availability of electronic maps and the associated population data. If electronic data and digitized maps of the entire country are available – as they are in Yemen – the costs associated with adopting GIS are much lower. In fact, it is only because the CSO is in the process of creating the necessary database that the SCER can even contemplate using GIS technology for the upcoming delimitation exercise.

The SCER must still purchase the necessary hardware and software, and train qualified staff to use the GIS system, however – and this will not be inexpensive. The SCER should plan on spending in the neighborhood of US$150,000 to $200,000 for hardware, software (including license) and training [17].

Potential Mismanagement of GIS GIS technology can be mismanaged, resulting in a disorganized, inefficient and delayed delimitation process. Detailed planning, adequate training and ample time and resources must be devoted to the endeavor if GIS is to be successfully incorporated into the delimitation process.

Potential Misuse of GIS Not only can GIS software be mismanaged, it can be misused: GIS technology could, at least in theory, make it easier for a ruling party to manipulate constituency boundaries so as to retain control of the Parliament even after the majority of the voters have ceased to support the party. In the United States, for instance, state legislatures assigned the responsibility for redrawing constituency boundaries often include political data (i.e., election results) in the delimitation database so that the political implications of proposed constituency configurations can be taken into account when drawing constituency boundaries [18].

However, political considerations are not included among the factors which the law requires the SCER to consider in drawing constituency boundaries. Moreover, including political data in the GIS database in Yemen would be very difficult because the units for which political data are available (election results are reported for voting centers) are not the same geographic units that would be employed for delimitation (census enumeration areas will have to be used to draw constituencies). But to ensure that the insertion of political data is not even contemplated, the electoral law could be revised to expressly prohibit the use of political data during delimitation. Yemen would not be unique in adopting this approach; many countries have expressly forbidden boundary authorities from considering political data when drawing constituency boundaries [19].

Article 159 of Yemen’s Constitution establishes the SCER as an ‘independent and neutral’ body. GIS would assist the SCER to demonstrate that it undertook boundary delimitations in a non-partisan manner if it:

 

 

  • made the GIS-produced maps and statistical reports associated with provisional constituency plans readily available;
  • instituted a public hearing process to allow interested stakeholders to comment on provisional plans; and
  • took stakeholders’ comments into account when modifying provisional plans to produce a final constituency plan, and published its reasons for modifying the provisional plans.
  •  

     

Conclusion

GIS technology offers the SCER a tool for implementing a more efficient, accurate and cost-effective delimitation process in 2005. GIS technology will produce a constituency plan that is far more likely to meet such constitutional and legal delimitation requirements as equal population and consideration for geographic and social factors; furthermore, this technology could be used to promote greater transparency in the delimitation process. The extent to which GIS technology can “democratize” the delimitation process, however, depends on how much information the SCER is willing to share with interested stakeholders and how open the delimitation process will be to public input.

If GIS is to be utilized for the 2005 delimitation, then plans for its use must begin immediately. The Appendix outlines, in very broad strokes, the steps that would have to be taken to implement GIS and proposes a tentative schedule. At least two possible impediments to the schedule have been identified, however:

(1) If the Ministry of Local Authorities redefines the 332 administrative divisions for which local constituencies are devised, and if this process is not completed by mid-2005, then proceeding with delimitation on schedule would be impossible. (Also, if administrative division boundaries were to change substantially, the delimitation process would take longer because the SCER would, in essence, have to begin from a blank slate, rather than simply modifying existing local constituency boundaries.)

(2) The Central Statistics Office must be willing and able to release the digitized census enumeration area maps by April 2005, and the population data associated with these enumeration areas no later than July 2005. The failure of the CSO to release this information in a timely manner would make proceeding on schedule difficult, if not impossible. It should be noted that the CSO considers it essential to have updated versions of the Oracle and ESRI ArcView software currently being used to meet its proposed release dates; if the CSO is unable to acquire these updates, the release date could be moved back several months.

Recommendations

 

 

  • The SCER should be supported in its effort to employ GIS technology in the upcoming delimitation exercise. Utilizing GIS in conjunction with the electronic database currently being constructed by the Central Statistics Office offers the SCER an opportunity to delimit constituencies more efficiently and more accurately. As a result, the process is likely to produce a constituency plan that meets such constitutional and legal delimitation requirements as equal population and consideration for geographic and social factors.
  • The SCER should be strongly encouraged to use GIS technology to generate a more transparent delimitation process. The SCER should release statistical reports and maps associated with any provisional plans and should hold public hearings to solicit comments on the provisional plans. The process could be managed by permitting only comments directly related to the established delimitation criteria (population equality, geographic and social considerations).
  • Consideration should be given to expanding the electoral law on constituency delimitation. In particular, a stipulation that delimitation occur on a regular schedule and that census data be utilized for the task would be appropriate (i.e., a delimitation of constituency boundaries should follow every decennial census and should rely on census enumeration counts for population data). Other supplements to the electoral law might include an express prohibition against incorporating political/partisan information in the delimitation database, and a requirement that the SCER consider stakeholders’ comments on provisional boundaries before producing the final delimitation.

Appendix

     

     

 

Proposed Stages for Implementing GIS

and Delimiting Local Constituencies

 

 

Proposed

Schedule

Planning Stage

September – December 2004

  • Assess hardware/software/training needs

 

  • Prepare budget and schedule

 

 

 

Acquisition and Training Stage

January – March 2005

  • Acquire necessary hardware/software

 

  • Train staff on GIS (possibly hire additional staff)

 

 

 

Database Development Stage

 

  • Phase I: obtain electronic census enumeration area maps from Central Statistics Office

April 2005

  • Phase II: draw boundaries of current local constituencies using census enumeration areas

April – June 2005

  • Phase III: obtain population data associated with census enumeration areas from Central Statistics Office

July 2005

 

 

Provisional Map Drawing Stage

August – September 2005

  • Modify current constituency boundaries to create provisional map that meets established criteria

 

  • Produce paper maps and statistical reports for provisional map

 

 

 

Public Hearing Stage

October 2005

  • Organize public hearing schedule (locations, dates)

 

  • Disseminate maps and statistical reports (provisional map)

 

  • Hold public hearings to solicit comments

 

 

 

Final Map Drawing Stage

November – December 2005

  • Modify provisional map based on public hearing comments

 

  • Create final map of local constituency boundaries

 

  • Produce written report, with maps and statistics

 

 

Notes:

[1] The 2003 elections were the third parliamentary elections held since the unification of Yemen in 1990; previous parliamentary elections were held in 1993 and 1997.

[2] The General People’s Congress (GPC) holds 240 seats (79.7%) in total because ten of the independents who ran and won in 2003 later affiliated themselves with the GPC.

[3] See “IFES Post-Electoral Assessment: Yemen April 27, 2003 Parliamentary Elections,” prepared by the IFES office in Yemen. Later figures from the Supreme Commission for Elections and Referendum (SCER) show 10 constituencies had 40-50,000 registered voters, seven had 50-60,000 and one had 70,109. The lowest number of registered voters in a constituency was 9,980. The SCER’s official estimate of the population in 2002 (based on a projection of the 1994 census) was 18,192,000; therefore, the population quota was 60,439 (18,192,000 ÷ 301). The average number of eligible voters per constituency should have been much less: the percentage of the population 18 years and older, according to UNICEF, was approximately 44% in 2002; hence, the average number of eligible voters per constituency was about 26,700. [UNICEF reports that 55.8% of the Yemeni population was less than 18 years of age in 2002; therefore, somewhat less than 44.2% of the total population should have been eligible to vote (assuming there are some non-citizens included in the population count). On the other hand, UNICEF estimates a total population of 19,315,000 in Yemen in 2002; if this estimate is more accurate than the population estimate employed by the SCER, then the number of eligible voters per constituency could be as high as 28,400 – still nowhere near the number of registered voters in some constituencies.]

[4] See “April 27, 2003 Parliamentary Elections in Yemen: Final Report” prepared by the National Democratic Institute for International Affairs and “IFES Post-Electoral Assessment: Yemen April 27, 2003 Parliamentary Elections,” prepared by the IFES office in Yemen.

[5] Article 24 of the General Elections and Referendum Law (2001) assigns the responsibility for determining the boundaries of electoral constituencies – both the parliamentary constituencies and the local constituencies – to the Supreme Commission for Elections and Referendum.

[6] The Constitution requires equality of total population rather than, for example, equality of voting age population or registered voters.

[7] Article 24 (a) of the General Elections Law provides that constituencies shall be “based on the principle of equal population and taking into consideration social and geographic factors.” [Unofficial English translation]

[8] Although ozal are not necessarily unified by tribal/clan ties, if an ozla was united in this manner, the SCER attempted to keep the ozla intact within a single constituency.

[9] The 20 governorates of Yemen are divided into 332 administrative divisions. According to the local authority law passed in 2000, the number of local constituencies per administrative division is to range from 18 to 30, depending on the total population of the administrative division:

population of administrative division

number of constituencies

35,000 or less

18

35,000 – 75,000

20

75,000 – 150,000

26

150,000 or more

30

It appears, however, that the number of constituencies allocated to administrative divisions is substantially less than the minimum requirement of 18 local constituencies in many instances: if all 332 administrative divisions had been assigned even the minimum number of constituencies (18), there would have to be at least 5,976 local constituencies created. In fact, however, only 5,620 local constituencies currently exist.

[10] Article 24 (b) of the General Elections Law requires a population variation of not more than plus or minus 5 percent.

[11] The number of voting centers is less than the number of local constituencies in the governorates of Taiz, Abyan, Shabwa, Sana’a, and Hajja; and the number of voting centers exceeds the number of local constituencies in Lahj, Ibb, Al-Baidha’a, Amran, and Al-Dhale’a.

[12] Constituency population estimates often had to rely on projections of the 1994 census to the local (i.e., village) level – projections that are particularly likely to be unreliable, especially given the lack of data on population migration.

[13] A census of the population is conducted every ten years in Yemen. The last census was done in 1994; the next census is scheduled for December 2004. The Central Statistics Office (CSO) anticipates releasing the census data in mid-2005, assuming they are successful in obtaining the updated versions of the software (ESRI ArcView and Oracle) they are currently seeking.

[14] Information on the possible redrawing of administrative boundaries (i.e., the number of divisions likely to be affected and the time table for the project) was not available at the time this report was prepared. The SCER should keep in mind that a revision of administrative district boundaries could impact on the delimitation process.

[15] However, other boundaries (such as administrative subdivision boundaries) would have to be digitized if they are to be included in the delimitation database.

[16] The boundaries of the census enumeration areas are unlikely to follow the boundaries of the currently existing constituencies since no reference was made to the electoral constituency boundaries when the CSO established the census enumeration areas.

[17] If the SCER works cooperatively with the CSO to purchase the required GIS software and training, the overall financial cost is likely to be higher (i.e., the software license will be more expensive), but the cost, as well as the benefit accrued, will be shared by both agencies. Furthermore, cooperation with the CSO is likely to yield additional benefits – at a minimum it is likely to provide the SCER with access to the GIS expertise the CSO has already developed.

[18] In the United States, the constituency plans for most states are drawn by state legislatures despite a very clear conflict of interest. Moreover, the inclusion of political data in the delimitation database, and even the outright manipulation of boundaries for political benefit, has been deemed legal by the U.S. courts.

[19] For example, boundary commissions in the United Kingdom (England, Northern Ireland, Scotland and Wales), Canada, and Australia have all been prohibited from considering political factors when promulgating a constituency plan.

 

The United Kingdom: System of Redistribution

 

The United Kingdom’s system of redistribution has operated in its current form, with some modifications, since the end of the Second World War. As of 2005, there have been five redistributions completed: 1947, 1954, 1969, 1983 and 1995. The sixth is underway and due to be completed in 2006. The task is undertaken by four independent Boundary Commissions, one each for England, Scotland, Wales and Northern Ireland.

Historical Background

The system for electing the lower house of the Parliament, the House of Commons, was introduced in the thirteenth century, and this system went largely unchanged for the next 600 years. Each Shire (or County) and Borough was invited to send two representatives: the former were elected from among the landowners, the latter by the enfranchised burgesses. When Scotland, Wales and Ireland were incorporated into the United Kingdom their Members of Parliament (MPs) were similarly elected.

Parliamentary boundaries were not controversial because no one expected constituencies to have even approximately equal populations – MPs were conceived of as representing places (or property) rather than people. Changes to the system occurred largely as a by-product of three nineteenth century franchise extensions expanding the (all-male) electorate: the Reform Acts of 1832, 1867, and 1885.

Before the 1832 Great Reform Act there were major variations in constituency electorates, a result primarily of nineteenth-century industrial-urbanisation. The three nineteenth century redistributions reduced these differences by removing seats from the small boroughs (mostly “rotten boroughs” with negligible electorates) and reallocating them to the rapidly-expanding shires. Most two-seat boroughs lost their separate status, and the new constituencies allocated to the shires returned a single MP.

The nineteenth century redistributions were undertaken by the House of Commons, and were carefully constructed by the government to favour its electoral interests. The first twentieth century redistribution in 1917 marked the earliest occasion when commissioners were explicitly non-partisan and the Commissions had clearly stated principles and freedom to recommend constituencies. It was also the first time that the goal of approximate equality in district population was accepted. Following the 1918 redistribution, however, constituency boundaries were left alone for a generation and large divergences in populations developed. The modern system was introduced after the Second World War, in part as a response to the major anomalies in the size of constituencies (as indicated by the number of electors).

Vivian Committee Recommendations for Redistribution In 1942 the wartime coalition government established a committee (chaired by the Registrar-General, Vivian) to consider various aspects of the electoral system, including ‘the principles on which any redistribution scheme should be based’. The Vivian Committee identified equal representation as the basic principle for a Parliamentary democracy – with constituencies of equal population returning one member each – and set out four salient features to be taken into account during a redistribution:

 

 

  • The need for an electoral quota, or an average number of electors per electoral district
  • The need for limits of toleration, indicating the allowable population variation around the quota
  • The need for continuity of constituencies, with change being proposed only where necessary so that MPs could build lasting relationships with their constituents
  • The need for constituencies to conform to local government boundaries in order to provide community representation and ease of organisation for elections (which are conducted by local government officers)
  •  

     

Other recommendations suggested a time interval for redistributions and procedures for the four independent Commissions to follow. Advice on whether each country should be guaranteed a minimum number of MPs was offered by the Vivian Committee as well. The latter subject would become the focus of debate for the next fifty years.

The Redistribution Acts of 1944 and 1958 The first House of Commons (Redistribution of Seats) Act, enacted in 1944, adopted many of the Vivian Committee’s recommendations. The Act set the limit of toleration at plus or minus 25 per cent of the electoral quota. It guaranteed representation for Scotland, Wales and Northern Ireland at their 1944 levels, as well as indicating a desirable maximum number of MPs for Great Britain (thereby implying a maximum for England). The Initial Review of Parliamentary constituencies, completed in 1947, was based on this Act.

Before the Initial Review was completed, however, the Boundary Commissioners claimed that they were unable both to meet the 25 per cent toleration limit and respect local government boundaries. The former requirement apparently dominated, since it came earlier in the Act’s Schedule of Rules. Parliament, however, determined that the ‘organic’ requirement to represent communities should take primacy over the ‘mathematical’ requirement of equal constituency population. They removed the 25 per cent deviation rule and replaced it with a rule that constituencies should ‘be as near the electoral quota as is practicable’. This new rule was placed after and, it was assumed, subsidiary to the rule regarding local government boundaries.

The Commissions’ First Periodical Review of all constituencies were reported in 1954 (the 1944 legislation required them to be delivered within 5-7 years of the previous Review). This generated consternation among MPs and party organisations, since constituencies were to be substantially changed soon after their creation, contrary to the Vivian Committee recommendation on continuity. Thus the government amended the Redistribution Act in 1958, extending the time period between Reviews to 10-15 years (in 1992, the allowable time period between Reviews was amended again – Reviews must now be conducted every 8-12 years).

Although the Act was subsequently amended to take account of major local government changes in the 1970s and was then consolidated into a new one – the Parliamentary Constituencies Act, 1986 – there were no further changes to the basic principles for redistributions.

Legal Framework for Redistribution

Boundary Authority The delimitation of constituencies is undertaken by four independent and non-partisan Boundary Commissions, one each for England, Scotland, Wales and Northern Ireland. The Commissions are composed of four members appointed by the relevant Secretaries of State. The nominal Chair is the Speaker of the House of Commons who neither attends nor participates; meetings are conducted by the Deputy Chair (a senior judge). Each Commission has assessors representing departments which supply vital information (for England and Wales these are the Registrar-General and the Director General of the Ordnance Survey). Each Commission also has a small civil service staff.

The Commissions make recommendations to Parliament, which can accept or reject, but not modify, them. (A Secretary of State can modify the recommendations before transmitting them to Parliament, but this has never happened.) In this century Parliament has never significantly revised the constituency boundaries prepared by the Boundary Commissioners [1].

Frequency When in 1944 it was decided to institute regular redistributions of seats, the time limit envisaged was “not less than three or more than seven year” – that is, once in the life of every five-year Parliament. But the dislocations involved in the first routine redistribution in 1954-55, only five years after the general redrawing of boundaries had taken effect in 1950, produced an outcry. The time limits were therefore altered in 1958 to “not less than ten or more than fifteen years” and in 1992 the limited was again changed – a redistribution of seats is now required every 8 to 12 years.

In each review, the process has been slow. The Third Periodical Review actually took seven years to complete. Although some of the delay was due to the two-year legal tangle over ward boundaries (and therefore could not be blamed on the delimitation process), a large portion was the result of the elaborate appeal procedures laid down in the 1958 Act.

Allocation of Seats to Component Parts of UK For the First through the Fourth Reviews, the electoral law guaranteed a minimum number of seats for Scotland (71) and Wales (35), a minimum and maximum number of seats for Northern Ireland (16-18), and a total number of seats for Great Britain (i.e. England, Scotland and Wales) that should not be substantially exceeded (613) [2]. As a consequence of these rules, England has become under-represented relative to its population and Scotland, Wales and Northern Ireland have become over-represented. For example, if seats were allocated to the component parts of the United Kingdom on the basis of population alone, prior to the Fourth Review, England would have additional seats, and Scotland, Wales and Northern Ireland would have far fewer seats:

 

 

Actual Number of Seats at start of Fourth Review

Number of Seats based on Population, start of Fourth Review

England

529

549

Scotland

72

59

Wales

40

34

Northern Ireland

18

17

This seat allocation will change rather dramatically for the Fifth Review because of the Scotland Act 1998, which removes the guarantee of a minimum number of parliamentary seats for Scotland. The Scotland Act also requires that the electoral quota for England be used to determine the appropriate number of Scottish seats to Westminster. The mathematical effect of applying the English electoral quota to Scotland is likely to be a reduction in the number of Scottish seats from the current 72 to fewer than 60 seats. (In fact, the Boundary Commission of Scotland, which has now completed its Fifth Review, proposes 59 constituencies.)

Allocation of Seats to Counties and Boroughs Each commission begins the delimitation process assuming that it will have the same number of seats in Parliament that it currently holds. The Fourth Periodic Review therefore began, in 1991, with England assuming an allotment of 529 seats, Scotland 72 seats, Wales 40 seats and Northern Ireland 18 seats.

The commissions establish the new “electoral quota” by dividing the latest annual register of the electorate by the number of seats authorized at the last redistribution (i.e., the current number of seats). In the Fourth Review, the quotas were as follows:

 

 

Number of Seats

Electoral Quota

England

529

69281

Scotland

72

58525

Wales

40

54569

Northern Ireland

18

67852

The English and Welsh Commissions then allocate the appropriate number of seats to each shire country (ranging from 1 in the Isle of Wight to 17 in Kent and Hampshire); to each metropolitan country (ranging from 13 in Tyne and Wear to 29 in the West Midlands); and to each of the 32 London boroughs.

Once seats have been allocated to counties and boroughs, the Commissions draw up provisional maps. It is possible, in the process of creating these maps, to award extra seats when “special geographic considerations” merit it. As a result, the size of the House of Commons has consistently grown: from 625 to 630, then to 635, then to 650, and to 659 in the Fourth Review [3].

Public Consultation Process The Commission’s provisional maps are published once they are completed. If objections to the boundaries are raised, local inquiries are scheduled [4]. The stages involved in the public consultation are as follows:

1) The Commission publishes its provisional recommendations for a local government unit in one or more newspapers circulating in the area, and sends notices to all affected MPs, political parties and local governments, giving details of the recommendations and indicating where maps showing the recommended constituencies can be viewed. Representations are invited within one month of the publication.

2) After the closing date for representations, if objections have been received from either at least 100 local electors or one interested local authority, then a public Local Inquiry must be convened. This is chaired by a specially-appointed Assistant Commissioner (AC), who is invariably a senior lawyer. ACs must have no political affiliations and in England (though not elsewhere) they must have no detailed knowledge of the area they are assigned.

3) Before the Inquiry, a document is produced summarising the representations received; a full list of those making representations and the grounds for the recommendations are appended.

4) At the Local Inquiry, the AC invites those who made written representations to make oral submissions, where they may be questioned by the AC and cross-examined by others who have made representations. Some of those objecting to the provisional recommendations (mainly the political parties) offer alternative configurations for one or more constituencies. Partisan considerations cannot be discussed, but the proceedings are invariably dominated by the political parties and their representatives (including local governments, most of which are politically-controlled). They use the criteria in the rules, especially those concerning community ties and the inconveniences of change, to influence the AC to recommend constituencies to the Commission which are in their own electoral interest.

5) On the basis of what has been read and, especially, heard, plus site visits when chosen, the AC’s report summarises local opinion on the provisional recommendations, discusses any counter-proposals presented to the Inquiry, evaluates the evidence, and recommends whether the Commission should change its provisional recommendations.

The boundary commissioners usually – but not always – accept the recommendations of the ACs. The reports and the reasons for accepting or rejecting offered recommendations are made public.

If the commissioners redraw the constituency boundaries based on the inquiry process, then further inquiries may be required [5]. The whole inquiry process can be quite long-winded and repetitive. The English commissioners started work on their Third Periodical Review in 1976 and did not produce their final report until late in 1982.

Commission can also conduct Interim Reviews to take account of either major local government changes or substantial population changes. Only one significant Interim Review has been undertaken: in 1990 the English Commission recommended an additional constituency for the rapidly expanding new town of Milton Keynes.

Delimitation Criteria Population equality, compactness, established geographic links, natural and administrative boundaries, and community of interest are the only factors that commissioners are to take into account in creating constituency boundaries. A summary of the rules are as follows:

 

 

  • “So far as is practicable’, constituency boundaries should not cross major local government boundaries - although this requirement is less stringent for Scotland and, especially, Northern Ireland.
  • Each constituency’s electorate should be as near the electoral quota as practicable, within the constraint of the previous rule. Commissions may depart from that previous rule to avoid disparities in electorates among neighbouring constituencies.
  • Commissions may depart from strict application of the previous two rules “if special geographical considerations, including in particular the size, shape and accessibility of a constituency” make that desirable.
  • Commissions should also take account of the inconveniences that may be caused, and the local ties that may be broken, if they give full effect to the “equal electorates” requirement.
  •  

     

The rules include two that give the Commissions considerable flexibility: the Commission can over-ride the requirement not to create constituencies crossing designated local government boundaries if this is necessary to avoid major disparities in constituency electorates; and the use of ‘special geographical considerations’ (of which size, shape and accessibility are cited as particular examples) to justify over-riding both the local government boundary and the equal electorates requirement.

The following table lists the number of constituencies that deviated from the electoral quota, and by how much, after the Fourth Review:

 

 

 

England

Wales

Scotland

N. Ireland

TOTAL

Within 10%           

444

29

50

14

537

10 -15% (plus or minus)

60

8

12

0

80

15 - 20% (plus or minus)

21

1

5

4

31

over 20% (plus or minus)

4

2

5

0

11

Non-Partisan Boundary Commission Commissioners are not supposed to have any cognizance of political party considerations. In fact, no one has seriously suggested that the commissioners have deviated from this apolitical ideal. On the other hand, during the public inquiry process, political parties and MPs may attempt to persuade the commissioners, using supposedly objective arguments about natural communities, etc., to act in the self-interest of a particular political party.

Court Challenges to Redistributions

In 1982, the Labour Party brought suit against the English Boundary Commission, challenging the Commission’s newly completed redistribution plan. This was the first such legal challenge brought against a boundary commission in the United Kingdom.

In the complaint, the Labour Party argued that the Commission had given too much weight to “natural communities” and county boundaries in the plan and too little weight to ensuring equal electorates. There were, in fact, large disparities in population across constituencies: for example, both the Isle of Wight and the London suburb of Surbiton were designated as single seats, but the Isle of Wight had an electorate of 95,000 and Surbiton had only 48,000 electors.

The court, however, in its decision in R. v. Boundary Commission for England ex parte Foot, found no evidence that the Commission had failed to undertake its statutory obligation to ensure equality of numbers. Furthermore, the court indicated a decided reluctance to query a Commission’s sovereignty – it suggested it was willing to do so only in the very special circumstance that a Commission has clearly acted unreasonably. The court in Britain has not been asked to consider the fairness of a redistribution plan since this 1983 decision.

Steps in the Commission Process

Each Commission operates in the following way:

(1) The Commission decides when to initiate a Periodic Review and announces its intention. (The four Commissions are not required to act together, but do, although English redistributions normally take much longer to complete.)

(2) Each Commission calculates its electoral quota using the 1986 Act formulation: the country’s registered electorate on the ‘qualifying date’ (when the review was publicly announced) is divided by its current number of seats.

(3) In England, Scotland and Wales, the Commissions determine each major local governmental unit’s ‘theoretical entitlement’ to seats, dividing its electorate by the electoral quota. (This is not done in Northern Ireland because local government units are not identified in the Act.)

(4) If some ‘theoretical entitlements’ would produce constituencies that are very large or very small relative to the quota, two contiguous local government units can be combined to achieve greater equality. (This has rarely happened.)

(5) The Commission staff prepares a number of optional schemes for constituencies in each local government unit. Local government electoral wards are always used as the ‘building blocks’: this is not legally required but has become the accepted modus operandi.

(6) The Commission evaluates the options offered and decides which one to put out to public consultation as its provisional recommendation.

(7) After the public consultation (see below), the Commission assesses the additional information and advice provided by the Assistant Commissioner who held the Inquiry, and decides whether to modify or confirm its provisional recommendations. If it takes the latter course, the provisional proposals become the final recommendations: they are published and included in the final report to Parliament. If the Commission decides to change any or all of its provisional recommendations (including a proposed constituency name), however, the changes are published and a further round of public consultation is initiated.

(8) When all of the recommendations have been made final, the reports are submitted to Parliament through the relevant Secretary of State.

For each constituency, the Commission has to recommend a name (which can stimulate considerable local concern) and whether to classify it as ‘borough’ or ‘county’. Candidates are allowed to spend more money campaigning in ‘county’ (rural) than in ‘borough’ (urban) constituencies.

Problems with the Review Process

The Review process can be quite time-consuming; the Fourth Review took four years to complete in England, with consideration of one County (Devon) taking 1,028 days to complete. An even greater problem is that the rules are ambiguous, with standards that are subjective (such as “as far as practicable”) and no indication of the relative salience of the various criteria. Some the particular problems that the rules and procedure create are as follows:

 

 

  • The guarantee of a minimum number of seats to three of the four countries ensures that they are over-represented relative to England, whose population is growing more rapidly. (This will change for one of the three countries – Scotland – once the Fifth Review has been completed.)
  • The method of calculating the electoral quota produces an in-built bias towards an increase in the number of seats. This is because constituencies that vary considerably from the electoral quota (because of ‘special geographical considerations’) are included in the denominator, increasing the likely allocation.
  • The allocation of theoretical entitlements to local government areas also tends to inflate the number of seats. This is because fractional entitlements are often rounded -up rather than -down.
  • Different Commissions can give different weight to the various criteria. For example, in the Fourth Review, the Scottish Commission determined not to create any additional seats (after Parliament expressed a desire for no growth in the number of MPs), whereas the Welsh Commission created an additional two seats, even though Wales was already substantially over-represented.
  • The same Commission (especially the English Commission, which has the largest task) can weight the criteria differently in different areas, giving an impression of inconsistency.
  • The use of the registered electorate rather than the population, although beneficial because the electorate is enumerated annually, means that 2 to 3 million people may not be included in the count. The Commissions cannot take this undercount into account when allocating seats, which may disadvantage areas with high under-enrolments (mainly inner cities), nor can the Commissions take an area’s projected population growth into account.
  • The Local Inquiry system allows the political parties to employ the various criteria to press cases which favour their electoral interests, without being transparent in their reasons. Thus the strength of the advocacy may convince the AC, rather than the merits of the case.
  •  

     

Notes

[1] However, the 1948 Labour government, believing urban areas to be under-represented, asked the commissioners to devise 17 extra seats. Although the request may have been politically motivated, the ultimate result was that the seats divided almost evenly at the next election. In 1969, the Labour government, worried about the seats it would lose under redistribution, postponed implementing the boundary commissions’ proposals on the ground that the impending change of local government boundaries would render the constituency boundaries obsolete.

[2] The 1944 Redistribution of Seats Act guaranteed that the representation of Scotland should be maintained at a minimum number of 71 seats and Wales at a minimum of 35 seats. Northern Ireland seats were increased from 12 to a proportionate 17 by the Labour government in 1979.

 

 

First Review

Second Review

Third Review

Fourth Review

England

511

515

523

529

Scotland

71

71

72

72

Wales

36

36

38

40

Northern Ireland

12

12

17

18

Total

630

634

650

659

[4] Public consultation was included in the nineteenth century redistribution process but its nature was only formalised in the 1944 Redistribution Act. The 1958 Act specified the circumstances in which a Local Inquiry is mandatory.

[5] Although the Local Inquiry process may be repeated if a Commission publishes revised boundaries after receiving an AC’s report, this does not happen often since the Commissions will not allow issues already fully covered in the previous proceedings to be reconsidered. (There were only two second Local Inquiries in the most recent Review, which included 83 first Inquiries.)

 

Federal Redistribution in Australia

This case study deals with the determination of the boundaries of electoral divisions for the House of Representatives of Australia's Federal Parliament, a process known as 'redistribution' in Australia10.

The Australian Constitution requires that the numbers of members of the House of Representatives elected from each Australian state be proportional to the populations of the states. It also stipulates that a federal electoral division cannot be formed from territory from more than one state. A consequence of these requirements is that separate redistribution processes must be undertaken to determine the boundaries of the federal electoral divisions in each state and in each territory represented in the Parliament.

The basic line of thinking that underlies the current redistribution provisions in the Commonwealth Electoral Act of 1918, which were substantially implemented in 1984, is that the legitimacy of the process can be guaranteed by attention to the following considerations:

Timing of redistributions. Provision is made for redistributions to be conducted with sufficient frequency to limit malapportionment, i.e., the inequality in the voter populations of divisions within a state or territory. In addition, the timing of redistributions is determined by law and cannot be manipulated for political advantage.

Constitution of bodies conducting redistributions. Redistributions are undertaken by politically neutral and independent bodies.

Provision for public input. The redistribution process is very public, and extensive scope exists for the views of interested individuals and bodies to be taken into account.

Criteria governing the drawing of boundaries. The bodies undertaking redistributions are required to work in accordance with well-defined and reasonable criteria which are broadly supported across the political spectrum.

Automatic implementation of redistributions. Once a redistribution has been made, it is not subject to veto at the political level, nor by Parliament.

Now consider each of these in more detail:

Timing of Redistributions

There are only three situations in which a redistribution of a state or territory represented in the Parliament can be initiated:

     

     

  • when there is a change in the number of members of the House of Representatives to be chosen in the state or territory at a general election, as determined approximately two years before each general election;
  •  

     

  • when a prescribed level of malapportionment is achieved and sustained in the state or territory;
  •  

     

  • when seven years have elapsed since the state or territory was last redistributed.

 

Of the three possible 'triggers' for redistributions, change in representation entitlements has been by far the most important and has accounted for the majority of redistributions initiated since 1984. There have been no redistributions triggered by malapportionment.

The rules governing the timing of redistributions are based on objective criteria, and do not permit the government of the day any discretion with regard to the scheduling of redistributions. The rules also clearly ensure that there is a limit on the length of time that can elapse between redistributions, which has tended to limit the extent to which malapportionment can arise.

Constitution of Bodies Conducting Redistributions

A redistribution in Australia occurs in two main stages. During the first stage a proposed redistribution is produced; during the second stage objections to the proposal are dealt with and a final determination is made. Different bodies, though with overlapping memberships, are involved in the two stages.

The proposed redistribution in a state is made by the redistribution committee for the state, which consists of:

     

     

  • the electoral commissioner, who is a statutory officer appointed under the Commonwealth Electoral Act of 1918, the administrative head, and one of the three members of the Australian Electoral Commission
  •  

     

  • the Australian electoral officer for the state, who is a statutory officer appointed under the Commonwealth Electoral Act of 1918, and who heads the Australian Electoral Commission's administrative structures in the state
  •  

     

  • the surveyor-general from the state government of the state (or, where the state has no surveyor-general, a person nominated by the relevant state minister as holding an office equivalent to that of surveyor-general)
  •  

     

  • the auditor-general from the government of the state

 

The second stage in the redistribution process commences after the proposed redistribution has been published. Objections to the proposed redistribution are considered and a final determination is made by the augmented electoral commission. This body consists of the redistribution committee for the state or territory, plus the two members of the three-member Australian Electoral Commission who were not members of the Redistribution Committee, i.e., the chairperson of the commission and the so-called 'non-judicial appointee'.

There is a consensus among Australia's main political parties that the membership of the redistribution committees and augmented electoral commissions is appropriate and enables them to operate in an independent and politically neutral way.

Provision for Public Input

Extensive provision is made for public input into the redistribution process. Redistribution committees are required by law to call for public suggestions at the outset of the redistribution process, and any suggestions received are made publicly available and can be the subject of further public comment. All suggestions and comments are required to be considered.

After a proposed redistribution has been published, objections to it may be lodged. The augmented electoral commission is required to conduct a public inquiry into an objection unless the objection covers matters which were already substantially raised in earlier suggestions or comments, or is frivolous or vexatious. Submissions regarding the objection may be made at the public inquiry by the person or organisation objecting, and any person who or organisation that made suggestions or comments.

The augmented electoral commission may also invite other witnesses to appear. The augmented electoral commission is not bound by the legal rules of evidence, and in general has considerable flexibility in determining how inquiries will be conducted. Inquiries are typically held in the capital city of the state, but on some occasions in the past an augmented electoral commission has chosen to hold an inquiry in a provincial city, particularly where such an approach facilitated the examination of a controversial proposed boundary.

Having held such public inquiries as are necessary, the augmented electoral commission is required to make a further proposed redistribution. As part of that process, the augmented electoral commission must determine whether, in its view, the proposed redistribution is 'significantly different' from that put forward earlier by the redistribution committee, and then make an announcement to that effect. If the proposed redistribution is not regarded as significantly different from that of the redistribution committee, there is no provision for further objections, and the proposed redistribution is in practice reflected in the final determination made by the augmented electoral commission. If the proposed redistribution is regarded as significantly different, further objections can be lodged by individuals who or organisations that had previously made suggestions or comments or lodged objections, and a further public inquiry or series of public inquiries must be held.

The net effect of all of these changes has been to make the redistribution process a far more open and accountable one than was previously the case. The redistribution committees and augmented electoral commissions have tended in their statements of reasons to give substantial explanations of their proposals, addressing in some detail the suggestions, comments and objections they have received. As a result, those who made suggestions, comments or objections have been better placed than before to appreciate why their own preferred positions may not have been accepted. Participants at inquiries into objections have typically been given extensive opportunities to expand on their submissions, so that the augmented electoral commission gains the maximum benefit from their views, and so that the participants can be satisfied that they have received a proper hearing. As a result, argument over the substance of a particular redistribution has tended to be talked out within the framework of the redistribution process, rather than forming the basis for on going questioning of the legitimacy of a particular redistribution.

It is also notable that notwithstanding the extensive provision made for public input, the redistribution process is one that can be undertaken relatively expeditiously.

Criteria Governing the Drawing of Boundaries

The criteria governing how boundaries are to be drawn by the redistribution committees and augmented electoral commissions are fully spelled out in the Commonwealth Electoral Act of 1918, and are as follows:

(a) The State or Territory must be distributed into a number of electoral divisions equal to its representation entitlement in the House of Representatives as most recently determined.
(b) A quota is determined by dividing the State or Territory's total electoral enrolment as at the end of the period for the lodging of comments by its representation entitlement in the House of Representatives. An electoral division which deviates in enrolment from the quota by more than 10 percent cannot be proposed.
(c) Subject to rules (a) and (b), the Redistribution Committee or augmented Electoral Commission must, as far as practicable, endeavour to ensure that the boundaries of the electoral divisions are such that the number of electors enrolled in each division will not, at the point three-and-a-half years after the redistribution has come into effect, be less than 98 percent nor more than 102 percent of the average divisional enrolment for the State or Territory at that time.
(d) Subject to rules (a), (b) and (c), due consideration must be given, in relation to each proposed division, to:
(i) community of interests within the proposed division, including economic, social and regional interests;
(ii) means of communication and travel within the proposed division;
(iii) the physical features and area of the proposed division; and
(iv) the boundaries of existing divisions.

 

These criteria are heavily numerical and to a considerable extent objective. The relative importance of the main criteria is clearly defined in the legislation, and is not a matter in which the bodies conducting redistributions have any significant discretion.

It should be noted that those bodies are not required to take any account of the possible partisan implications of redistributions, and in fact they have not done so.

Automatic Implementation of Redistributions

Under current provisions, the determination made by an augmented electoral commission is final and cannot be vetoed at the political level. It is also not subject to judicial review, except on very limited constitutional grounds.

Concluding Comments

The emphasis in the current scheme is very much on the legitimacy of processes rather than specific outcomes. It is generally believed that if the mechanisms leading to a specific redistribution are acceptable, then the redistribution itself will be accepted, even by participants in the electoral process who see themselves as being disadvantaged by it. The history of the last ten years has tended to bear out that point of view.

Federal redistributions have largely ceased to be subject of partisan debate. The Australian approach can be contrasted with that which is adopted in the United States. In the United States, the concept of the apolitical civil servant is not always regarded as credible, and redistributions have tended to be conducted on an overtly partisan basis. The legitimacy of redistributions in the United States has, therefore, been analysed by reference to postulated outcomes much more than by reference to processes.

In addition, the legal provisions governing redistributions are precisely drafted, are detailed, and give relatively little scope for discretion in their implementation. Some might see this as a disadvantage; but a contrary argument can be made, to wit, the current provisions represent a political and community consensus reached after painstaking consideration some fourteen years ago. Particular outcomes can be seen to be the product of the application of enduring principles rather than less well-defined ad hoc considerations. These factors have arguably enhanced community acceptance of the process.


Boundary Delimitation