« Are stock and bond markets contradicting each other? | Main | Assorted links »
How to sell a dollar for more than a dollar
In this time of cutbacks and furloughs professors need a quick source of extra cash. David Zetland explains how to sell a dollar for more than a dollar (and teach some game theory and something about political lobbying at the same time).
Posted by Alex Tabarrok on November 2, 2009 at 10:37 AM | Permalink
Comments
I do this in class, but I use a $20 bill. That seems to work better.
Posted by: Jason Brennan at Nov 2, 2009 10:46:38 AM
"bribe one more congressman"? i thought all lobbying was informational!
Posted by: anon at Nov 2, 2009 11:45:33 AM
The Prisoner's Dilemma is a better analogy for political lobbying or arms races, where the optimal move is to help yourself which indirectly harms others.
In this game, the optimal strategy is not to bid or bid exactly a dollar. We don't think that all lobbying is bad for the lobbyists themselves, which would have to be the case for this to be a good analogy.
Posted by: Jason at Nov 2, 2009 12:27:29 PM
For this to be an accurate model for lobbying, the benefits from successful lobbying would need to be about the same order or magnitude as the funds available for lobbying. In reality, the potential benefits from lobbying must often far exceed the typical money available for lobbying. The farm lobby spends a lot, but not tens of billions per year. A better model would be bidding for a 100 dollar bill when you know that there are only a handful of potential bidders and you have a rough but not exact idea of how much each has to spend (and much, much less than 100 in any event).
Posted by: jdm at Nov 2, 2009 12:54:16 PM
Brilliant!
Posted by: anon at Nov 2, 2009 12:55:05 PM
Lesson from the Video: The Republicans sold themselves too cheap for committing at the beginning of the game not to support healthcare reform. They would have been better to be in play, rather than be committed.
Lesson No. 2: Washington is a big isometric exercise: one arm pushes against the other, and both arms get stronger, and nothing moves.
Posted by: Bill at Nov 2, 2009 2:20:48 PM
Bill: we should be so lucky!
Actually, I think that this model works very well. Of course there are some major differences.
In practice, there are often two bidders: someone wants a bid, and someone does not. This means that the bids are cumulative, which this exercise misses. (A bids 1,3,5 while F bids 2,4,6. Total bidding is 11, not 21.)
Note that this example can get even more interesting in this case, as the bids might well run .7, .8, .9, 1.0, 1.1, 1.2.... Where each bidder is only adding .2 to their bid to gain 1.
Second, despite the fact that these are modern times, politicians tend to be very much bought rather than rented. Just from the title of a bill, I would expect that we can predict the results for individual congressmen to 80% or better. Since the bids are more for a term of control of a seat, rather than an individual bill, the total amount of money to be had is in fact flexible.
Posted by: Right Wing-nut at Nov 2, 2009 4:27:19 PM
Bidding $0.25 is not a "sunk cost" because a dollar is worth a dollar. A "sunk cost" would be the cost invested in an outcome much larger than the one-off auction, such as the entertainment value of playing such a silly game. So if the opportunity cost is, say, a $2 video on iTunes then perhaps you could bid up to that and not "lose."
Posted by: aohl at Nov 2, 2009 5:44:33 PM
This is a great, classic example for students, but I thought the instructor here did a very awkward job.
Posted by: Paul N at Nov 2, 2009 8:14:31 PM
I always use as an example the argument for escalation during the Vietnam War that if we do not escalate all those boys who died before you would have died in vain, never asking if we would win if we did escalate and at what additional cost. Argument is usually accompanied by the phrase "cut and run" just for good luck.
Posted by: Bill at Nov 2, 2009 8:25:48 PM
I always loved this game theory example, and came across it first through dinosaur comics a few weeks ago. I would recommend http://www.qwantz.com/index.php?comic=1564 if you want a more in depth analysis of the game.
Posted by: Zach Piso at Nov 2, 2009 9:24:22 PM
The first bidder should bid $.99.
Posted by: floccina at Nov 2, 2009 10:32:56 PM
@floccina,
No, the first bidder should bid a dollar. If you bid .99, I might get some spite benefit, and bid a dollar, but I would be uncertain that you would not want to do the same to me. Safer not to bid or to bid true value as first bidder as dominant strategy.
Posted by: Bill at Nov 2, 2009 10:43:51 PM
Who else remembers "pure distilled evil" site swoopo:
http://www.marginalrevolution.com/marginalrevolution/2008/12/profitable-unti.html
Posted by: Steko at Nov 3, 2009 2:19:52 AM
I do this in class, but start with 5 or 10 cent increments. it really highlights the difference between the Marginal Cost, and the Cost already incurred (sunk cost as noted). Marginal Cost being the bid, and the Marginal Benefit (the chance to win the dollar, and if there are only a couple people bidding - as is most often the case- the chance to win is pretty good, making the MB = chance * $1 > MC)
Rational to keep bidding. Starting to Bid is unwise though, unless there is significant enjoyment from the thrill of the bid (i sold one a few years ago for $8).
Posted by: Brian Held at Nov 3, 2009 11:15:27 AM
I do this in class, but start with 5 or 10 cent increments. it really highlights the difference between the Marginal Cost, and the Cost already incurred (sunk cost as noted). Marginal Cost being the bid, and the Marginal Benefit (the chance to win the dollar, and if there are only a couple people bidding - as is most often the case- the chance to win is pretty good, making the MB = chance * $1 > MC)
Rational to keep bidding. Starting to Bid is unwise though, unless there is significant enjoyment from the thrill of the bid (i sold one a few years ago for $8).
Posted by: Brian Held at Nov 3, 2009 11:15:47 AM
I found this to be extremely interesting, and I completely agree that no one wins in this scenario except for the person putting the dollar up for bid. I mean how fast did the bidders perceptions change from trying to make a profit to trying to minimize losses. It truly makes me wonder just how far will lobbyists will go to try and get their version of a law passed. Because it seems as though the lobbyists spending will exponentially increase until one finally backs down or until they bankrupt themselves. It's no wonder many politicians these days seem to be out of touch with the people when lobbyists are getting into competitions like this over their favor.
Posted by: tdk at Nov 3, 2009 1:39:45 PM
It is a very intriguing video for me, especially since we have just studied Prisoner's Dilemma in class. This auction reminds me of it in a way. More or less this is about marginal cost, marginal benefit and sunk costs. Economic analysis suggests that one should not take actions for which marginal cost exceeds marginal benefit. It makes sense to make a bid up to a dollar to generate a profit. But as soon as price reaches a dollar and then exceeds it, bidder needs to establish the maximum "sunk cost" level. It also depends what kind of benefit we are talking about. If the price is 1.98$ and one is willing to pay 1.99$, it means he certainly hopes for a better return than 0.99$. On the same side if he gives up at the current price he looses 1.98$ on the spot.
Question: Assuming someone paid more than a dollar for a dollar, how long of a time that person is allocating to actually enjoy that marginal benefit?
Posted by: Aliaksandr Kikoin at Nov 4, 2009 1:51:24 AM
