Morality and Markets: A Comment on Predicting Crime

In their article, *Predicting Crime*, Professors Henderson, Wolfers, and Zitzewitz propose an intriguing and futuristic series of market-based models surrounding the broad topic of crime prevention. Harnessing widely dispersed knowledge among groups of people, including cops on the beat, criminologists, residents of neighborhoods, elected officials, snitches, and possibly even the criminals themselves, the authors posit that prediction markets will help to estimate crime statistics more accurately and therefore result in more efficient deployment of policing resources. Further, they hypothesize that posing particular policy alternatives—for example, the option of eliminating the death penalty—to a widely dispersed market will result in a more transparent and open decision-making process. In addition to making important contributions to questions of prediction market design, their article explores and amplifies a discussion already underway that seeks to identify productive and socially beneficial uses for prediction markets. Crime prevention would surely rank highly along any measure of important governmental functions, and the topic the authors have selected is therefore of particular significance.

Over the past decade, prediction markets have become both a more familiar and a more acceptable way to forecast a wide variety of future events. Prediction markets have recently proliferated because, among other reasons, technological barriers to entry are lower than they once were and numerous prediction market providers have entered the field. As a greater number of people learn about these markets and their potential for gathering and consolidating information, new uses and possibilities continue to be explored. Also, as a greater number of people learn that prediction markets can also be fun—certainly, winning provides a hedonic benefit—the lure to potential participants increases. Politically minded students have told me that they tracked prediction markets in the last election the same way that those in my age group check polls. The main difference is that while checking polls is largely a passive exercise, prediction markets, on the other hand, give their participants a stake in the outcome by allowing them to contribute information; thus, participants in prediction markets feel more engaged and involved as market participants.

As prediction markets are becoming more common, the legal regime surrounding “real” money markets has become chilly, due to the online gambling ban enacted in the United States. This legal development has meant that prediction markets have either been forced into using play money or moving offshore. Further, during the past two years, the worldwide financial crisis has made faith in the rationality of unregulated markets seem rather naive. Because of the financial crisis, many have questioned the assumptions underlying the workings of markets more generally (not just prediction markets). It is with this background that I turn from the more general subject of the growth and development in the areas of prediction markets to the matters raised in *Predicting Crime*.

As I agree with the authors’ goals and am convinced that adoption of prediction markets by policymakers would improve public safety and advance transparency in public policy, my major comment upon reading *Predicting Crime* is to offer some practical suggestions that would pave the way for public acceptance of prediction markets in the area of crime and criminal law. As currently formulated, there is some risk that the authors’ proposal will be difficult to implement because it will be viewed as controversial. First, allocating crime-fighting resources through a market mechanism that allows wide segments of the population to participate could allow those who are self-interested (including those who either have perpetrated crimes or could perpetrate crimes) to become part of the market. Second, proposing that matters as weighty as the death penalty be decided through a prediction market mechanism also seems destined for controversy. With that being said, I believe some of these (anticipated) problems could be averted with some minor—but directed—changes to the market design proposed by the authors.