When an Auction Fails: New Zealand TV Frequencies Case
For hundreds of years, people searched for ways to distribute goods as efficiently as possible. A good way to do so when we do not know the values others have for an object is running an auction. Auctions have been around since 500 B.C., when Ancient Greek historian Herodotus reported that “each year the beautiful women in each village were auctioned off to the highest bidder”(McNeal). However, despite hundreds of years of research in the field, some auctions fail, creating scandals or losing their organizers a lot of money. Let’s look at an example of a notoriously botched auction and try to understand what lead to its failure, and how it could have been prevented.
In 1990, the New Zealand government hired an economic consulting firm, NERA, to design the best auction for Television frequencies. NERA proposed an auction that was expected to raise $250 million in revenue (and only raised $36 million instead), the second-price sealed bid auction.
Originally described by Vickrey, this auction implements the ascending bid auction, without the need to assemble bidders together. Moreover, it presents each bidder with an easy way to determine their bid, without having to consider the number of others’ values: the dominant strategy is to always bid your true value.
Despite auctioning off identical licenses, NERA advised to not adopt the “highest rejected bid rule,” which was sometimes used in the sale of U.S. Treasury bills. Such rule would mean one auction would be held for all n licenses, with the highest bidders paying a uniform price equal to what the (n+1)th highest bidder was willing to pay. Instead, they decided to conduct an independent second price sealed-bid auction for each of the licenses. Bidding your true value is not a dominant strategy is this case. In fact, it becomes extremely hard to give a good piece of advice to any of the bidders. Should you pick a random auction and bid your true value? If so, it is possible that one of the auctions will not receive any bids at all. Instead, maybe it would be a good decision to distribute your value equally by making small bids in every auction? If everyone does so, maybe it’s better to bid a bit more in one of the auctions and secure a win?
As can be seen from the table above, the bidders failed to guess each other’s values. But the greater problem was the public attention this auction drew. McMillan (1994) described the failure as follows: “In one extreme case, a firm that bid NZ$100,000 paid the second-highest bid of NZ$6. In another the high bid was NZ$7 million and the second bid was NZ$5,000.” This created a lot of public outrage. As a response, the government shifted to a more standard first price sealed bid format, which did not guarantee higher prices, but concealed the bidders’ values from a curious public.
How could this have been prevented? Paul Milgrom argues the government could have adopted the so-called Dutch flower auction design, alternatively known as a descending bid auction. Starting high, when the price goes down to a value lower than a bidder’s true value, they can win and buy as many slots as they want. There would have been little guessing involved: if a bidder wins, he buys as many licenses as he wants at the price he is willing to pay.
Works cited:
McNeal, Richard A. “The Brides of Babylon: Herodotus 1.196.” Historia: Zeitschrift Für Alte Geschichte, vol. 37, no. 1, 1988, pp. 54–71. JSTOR, http://www.jstor.org/stable/4436038. Accessed 31 Oct. 2022.
Milgrom, Paul. “Putting Auction Theory to Work – UCLA Economics.” UCLA Economics Department, 2003, http://www.econ.ucla.edu/riley/271/Milgrom-Putting%20Auction%20Theory%20to%20Work.pdf.
Talwalkar, Presh. “4 Crazy Moments in Auction Theory History.” Mind Your Decisions, 22 July 2014, https://mindyourdecisions.com/blog/2014/07/22/4-crazy-moments-in-auction-theory-history/.