New Zealand Spectrum Auction
In 1990, the New Zealand government tried to auction off their radio spectrum through a second price auction, where the highest bidder pays the second-highest bid price. In theory, this sounds like a great way to do things – a bidder can be upfront about what they are willing to pay, but do not necessarily have to pay that much. In practice, things did not go too smoothly. The government was expecting to raise NZ$250 million from the auction but only ended up raising NZ$36 million. Clearly, a second price auction was not the best way to do things.
The first issue was that sometimes there would be a huge discrepancy between the highest bid and the second bid. For example, Sky Network TV had the highest bid of over NZ$2.3 million for lot 1, but the second-highest bid was only NZ$401,000. This works out great for Sky Network TV, who only had to pay $401,000, but serves as an example of how inefficient the system was. Another issue was that companies who were willing to pay more for a lot ended up paying less than companies who were willing to pay less. BCL was willing to pay $255,124 for a lot, and the second-highest bid was $200,000. Totalisator Agency Board was willing to pay $401,000 for a lot, but the second-highest bid was $100,000. There were winners and losers in this process.
This is a real-world example of the second price auction that we learned about today in lecture. Following their dominant strategy worked out great for Sky Network TV and Totalisator Agency Board. It also helped BCL, but not as much as it could have. It did not help the New Zealand Government. This serves as a reminder of how things that sound great on paper do not always behave as optimally in the real world.
Source: http://www.eecs.harvard.edu/~parkes/cs700/patw.pdf