Social capital, conflict and welfare

Abstract

This paper analyzes the role of external conflict as a force that can create social capital. Hostile inter-group interactions can help to resolve intra-group social dilemmas but these potential gains must be weighed against the insecurity of hostile relations with an out-group. Our central result is that the presence of an outside threat can induce higher levels of social capital either because a protective aspect of social capital comes into play and/or as a reallocation of investments from private to social capital. Given that social capital is potentially subject to free-riding, the threat, by promoting a greater level of social capital, can be welfare improving. When the threat is severe, social capital and welfare are more likely to fall. This effect of an external threat on social capital is stronger in poor economies. These results can shed light on the sometimes contradicting empirical evidence on the relationship between conflict and social capital.

Introduction

Conflictual group relationships are everywhere. As long as limited resources and opportunities exist and antagonistic identities persist, rival communities will clash. But hostile intergroup interactions can have ambiguous effects. Whilst conflict and its anticipation can be costly in terms of diversion, destruction and disruption of productive resources, out-group hostility may help to resolve in-group social dilemmas. An empirical literature has recently emerged which finds evidence of increased pro-social behavior and collective action in societies that have experienced conflict.¹ A number of field and laboratory experiments corroborate that in-group relations improve as a response to the existence of a rival out-group.² This might not be that surprising. As argued by Choi and Bowles (2007) and Bowles (2009), individually costly norms of pro-group behavior are evolutionary adaptive in hostile environments. Conflict can induce pro-social changes in preferences among members of affected communities (Voors et al., 2012). External threats also kickstart communal coping processes (Lyons et al., 1998), that is, mechanisms of cooperative problem-solving that emerge when a community must confront adversity. In sum, violence and conflict can enhance social capital and potentially compensate the costly diversion of resources and destruction that they bring.

Examples of the positive effect of external conflict on trust and social cohesion abound, from the local to the national level. During the decade-long civil war in Liberia, neighborhood watch schemes became a community response against burglary and related crimes (Sawyer, 2005). Bellows and Miguel (2009) describe how communities in Sierra Leone organized local fighting groups during the civil war; civilians volunteered to these groups which were supplied and funded through local contributions. Similar self-defence forces have emerged in villages in Afghanistan in response to Taliban insurgency (Jones and Muñoz, 2010). At a macro level, sociologists and political scientists have often argued that interstate war strengthens national identity. For instance, Smith (1981) postulated that Medieval France and Spain owed their sense of national unity to their wars against the English and the Moors respectively. In modern times, interstate war might have contributed to state-building processes such as the German unification of the 19th century (Sambanis et al., 2015).

With this evidence in mind, the present paper explores the role of conflict as a force that can create social capital. This is important because there is persuasive empirical evidence indicating that social capital contributes significantly to growth and development (Knack and Keefer, 1997, Zak and Knack, 2001, Sobel, 2002, Guiso et al., 2004). We build a model that focuses on investments in social and private capital. We analyze how conflict affects the investment decisions made by members of a community threatened by an external entity. Our main argument is that in the absence of conflict, the public good nature of social capital leads to free-riding and under-investment in social capital. Conflict can help overcome this collective action problem because social capital also has a protective facet that helps the community to confront the external threat. As a result, the external threat stimulates social capital as there now exists a protective reason to invest in it, in addition to the productive reason to invest which already existed under autarky.

For a relatively wealthy society, the protective facet of social capital also stimulates investment in private capital as it is made more productive by the increase in social capital. Supposing it is relatively small, the presence of an external threat can actually increase social welfare. To be clear, we are not arguing here that societies should engage in conflict just to increase their social capital and overall welfare. Our theory rather suggests that communities confronting an external threat can resist and in some cases develop relatively successfully. On the other hand, when the threat becomes relatively strong, welfare may fall below the autarkic level. In that case it is difficult to protect capital returns and as a result, capital investment falls in favor of non-expropriable consumption. Our model can thus help to reconcile the existence of the aforementioned evidence showing that conflict is linked to higher social capital together with the evidence showing that conflict can undermine trust, willingness to trade, and associational membership.³

We then move to the study of relatively poor societies which are constrained in terms of how much they are able to invest in all forms of capital. Now, increases in social capital stimulated by the threat come at the expense of a reduction in private capital. As a result, the virtuous knock-on effect for low levels of the threat experienced by unconstrained economies is lost for constrained economies. Poor economies become more social capital intensive than wealthy economies subject to a threat of the same intensity. But these constrained economies are also less able to protect themselves and, as a result, they are less likely to attain full security of property rights.

Our next result refers to the case where social capital has bonding elements, defined as forms of connectedness created within homogenous groups (Putnam, 2000). This type of social capital is potentially less productive but easier to form as it based on relations with similar individuals. Rather than modeling several types of social capital, we opt for parsimony and assume that social capital can have consumption-like returns. This is because the returns of the bonding facet of social capital are not easily expropriated and therefore may be very attractive to a community under threat. We show that if the bonding aspect of social capital is sufficiently strong, the level of social capital under the threat is always above the level under autarky. Not only that. The level of social capital is monotonically increasing in the scale of the threat. As the threat intensifies, members of the community divert their investments from expropriable private capital to partially expropriable social capital. Although it may make sense to invest in it because it is not subject to theft, the bonding aspect of social capital might not be productive enough to compensate the reduction in private capital investments and the insecurity that the threat provokes.

There are a number of papers which attempt to model social capital formally despite its dual nature as input and output of social interactions (Durlauf and Fafchamps, 2006). These attempts can be broadly divided into two perspectives. The first one is microeconomic and sees social capital mostly as an output.⁴ On the other hand, the macroeconomic perspective tends to see social capital as an input in production. We follow a similar approach in our analysis. In a pioneering contribution, Glaeser et al. (2002) model social capital as an individual characteristic which agents invest in and that has positive externalities for the rest of society. The closest paper to ours within this literature is Beugelsdijk and Smulders (2009) who consider several forms of social capital. Costly investments in one of these types can protect individuals from the expropriation efforts of other agents. In a similar line to our findings, they show that certain forms of social capital can crowd out economic growth.⁵

In the next section, we discuss various sources of evidence for our model, which we present in Section 3. In Section 4 the model is applied to expropriable social capital in a relatively wealthy economy. This represents the core part of the paper. In Section 5, we offer four extensions of the benchmark model. Section 6 contains some additional discussion and concluding comments.