Let’s say you auction off a jar of pennies, and people make bids based on fairly random guesses. The deal is actually rigged in your favour, because if people are guessing randomly we would expect an equal number to overestimate the jar as underestimate. But by the nature of the auction the winning bid will be made by the person who overestimated by the most. In other words if several people are bidding for a resource – and it has the same value for them all – the winning bid will be the one that has the highest error term.
This is a nice illustration of the winner’s curse (h/t David Skarbek)
This phenomenon has a plethora of real world applications:
Why is it that cities that host the Olympics spend so long paying them off? Well, if we believe that the Olympics confers roughly the same economic benefits on any host city, the challenge is to bid an amount that is marginally less than this expected benefit. But if all potential locations do this the winning city will not be the one that can expect the largest windfall, but the one that overestimated the benefits by the most.
If a number of oil companies are bidding for the rights to excavate a particular geographical area, they will base their bid on their estimate of the net present value of the oil. Some will overestimate the benefits and some will underestimate. If you have the winning bid, instead of celebrating, you should ask yourself why no one else was willing to pay as much!
Mergers and acquisitions tend to produce lower than expected profits, and again the winner’s curse offers an explanation. Appealing to notions of “synergies” is often a cover for the fact that a company overestimated the value of the company they’ve bought, and we shouldn’t be surprised – by the very nature of the auction we expect to see buyer’s remorse.
This suggests that institutions are what bridge individual decision-making / cognition with outcomes. We need to focus on institutions to determine whether psychological explanations are enough.
For example, imagine that instead of auctioning off the jar of pennies we simply ask people to guess the value of the jar. In this case we would once again expect some people to overestimate and some to underestimate, but the average guess will typically be close to the actual value. This is an example of the wisdom of crowds.
There are plenty of situations where we wouldn’t want to rely on the wisdom of crowds (imagine a plane getting into difficulties and the pilot taking a vote on what to do about it), but the amazing insight remains: under the right institutional setting, the aggregated information of the crowd can beat expert judgement. It is institutions that determine whether human interaction delivers profitable or socially costly activity.