Monday, 27 April 2009

Markets and asymmetric information

A common reason, we are told, that markets may fail is asymmetric information. This idea goes back, at least, to Akerlof's famous paper on adverse selection and his example of the used car market. However we know that markets can deal with this problem. But a commonly held belief is that for markets to work, formal enforcement of contracts is necessary to prevent failure when information asymmetries exist. If fact, as Trevon D. Logan and Manisha Shah point out,
[...] it is hard to imagine the contracts that would be entered into in a world void of formal enforcement. How could fraud be detected and punished? Would assets be misappropriated? Wouldn’t the unscrupulous actors in the economy drive out the good, leaving us with corruption, uncertain property rights, and low levels of economic progress?
and they add
The standard results in game theory and contract design say yes. Theoretically, formal enforcement per se is not necessary if one uses, for example, long-term relationships. Without such added structures, however, a rational principal would never trust the information received from an agent, as it could likely be cheap talk. Also, the fact that additional structures must be in place suggest that the absence of formal enforcement is less efficient than formal enforcement. Any discussion of the role of institutions, the rule of law, and contract enforcement is predicated on the existence of formal enforcement. In the institutional developments now taking place in Iraq and Afghanistan, the development of formal institutions is a pressing policy concern.
But as Logan and Shah argue in a column, Information and illegal market mechanisms, at the question of whether formal enforcement is necessary to overcome the problems of asymmetric information is an inherently empirical one. Logan and Shah explain,
If, using the definition of Robert Hall and Charles Jones (1999), “good” social infrastructure is the mechanism that gets the prices right, we would like to see if markets without formal enforcement (which would presumably be riddled with “bad” social institutions such as corruption, rent seeking, and theft) get the prices wrong more often than markets with formal enforcement, which are presumably the “good” social infrastructure. In other words, does the data support the theory?
But there is a problem,
Testing this proposition, however, requires that we have access to the prices in an illegal market, if only to guarantee that the underlying contract is illegal and therefore unenforceable. Illegal markets, by their very nature, tend to stay hidden, so finding one where the information that agents communicate varies only increases the difficulty. To be sure, it is the rare illegal market where prices are widely known and where information on contractual arrangements can be observed by outsiders. In a recent paper (Logan and Shah 2009), we exploit one such market to answer this important question – the online market for male sex work. What we find is fascinating, and at many turns supports the existing theory of the economics of information, but also cautions that formal enforcement may not get the prices right more often than markets without formal enforcement, especially in the digital age.
What about the market for male sex work? Logan and Shah write,
First, male escorts are independent owner-operators who advertise online. The ability of male escorts to price directly without intermediaries and the large number of escorts create a market setting similar to competitive market assumptions where we expect markets to function well. Since this market is an illegal market, however, there is potential for escorts to mislead clients and engage in fraud. In particular, an escort’s ability to post unreliable information should lead to adverse selection in the market. While escort claims are verifiable ex post, there are no formal institutional penalties for ex ante misrepresentation. Indeed, most of the scenarios one could think of (posting deceptive information, obtaining and maintaining a false reputation, etc.) would lead us to predict that information would have little value. We found, however, that the market for male sex work is well regulated, informally, by the principals. Clients police escorts in two ways – through posts to independent client-owned forums and through detailed reviews of escort services on the escort websites, which are linked to the respective escort’s specific advertisement. This well developed system allows clients to review the services of male escorts and alert other clients to deceptive male sex workers. In fact, clients have identified, and made known, their preferred signal of escort quality – face photos in an escort’s advertisement.
So while informal enforcement is ubiquitous, the question has to be asked, Is it enough to get the prices right? Logan and Shah say, yes.
We find that this illegal market values information just as much as legal markets where truth-in-advertising and subsequent contracts are rigorously enforced. Interestingly, the price premium we estimate for the total number of pictures in escort advertisements (about a 1.5% increase in the price for every picture) is quite similar to the premium for pictures estimated by Gregory Lewis (2009) for used automobiles on That information has a value in any market is quite intuitive, but that it has such a large value in an illegal market is unexpected. In particular, we find that the market does not respond to all types of information – the premium to information in this market is driven entirely by face pictures, which command a premium in excess of 3% per face picture.
Logan and Shah conclude,
Our investigation into the market for male sex work provides a case where the richness of the information environment overcomes some of the problems of asymmetric information. The illegality of the market and the near-impossibility of guaranteed truthful disclosure imply that the market should deteriorate into a “lemons” market or one where information has dubious value. Surprisingly, clients informally police the market, punish misrepresentation, and reward credible disclosure. This enables male escorts to credibly signal their quality, and prices in the market respond accordingly. The illegal market mechanism is strikingly similar to it legal counterpart.
So, yet again, we find that markets do work and can find ways of dealing with asymmetric information. This is another example of the failure of market failure.

1 comment:

Steve_Roberts said...

"it is hard to imagine the contracts that would be entered into in a world void of formal enforcement"

Err, no. A little history shows that contract law, and the associated possibility of enforcement through the courts, is a recent development. Somehow we managed thousands of years of trade, locally and internationally without it !