This document is available here as an Adobe Acrobat PDF.
A version of this article appeared in the Wall Street Journal on May 11, 2007.
May 2007
Imagine the president had a crystal ball to predict more accurately the impact of broader prescription coverage on the Medicare budget, the effect of more frequent audits on tax compliance, or even the consequences of a political settlement in Iraq on oil prices. Now, stop imagining. Such crystal balls are within our grasp. But they cannot be used without running a gauntlet of federal and state regulation.
These crystal balls are called prediction markets or information markets, and they help forecasters, for example, by allowing traders to vote with their money on the future unemployment rate or the winner of the next presidential race. If you visit the Iowa Electronic Markets, you can take a financial position on the Democrats' chances of winning the White House in 2008. As this is being written, a contract purchased for $6.15 would yield $10 if a Democrat wins, allowing analysts to infer that the “market” believes the Dems have a 61.5 percent chance of taking the election.
Information Markets and Businesses
Some businesses use internal information markets to predict outcomes of specialized interest. Hewlett-Packard asks employees to predict revenues and operating profits, believing the exercise can add useful information to conventional forecasting methods. HP is even pondering the sale of a commercial version of its BRAIN (Behaviorally Robust Aggregation of Information in Networks). Eli Lilly has used these markets to help predict which newly developed chemical entities will become successful drugs. Google has used the approach to forecast product launch dates.
These markets often predict more accurately than experts. Why? They draw on the knowledge of people who might otherwise be ignored. Their anonymity frees participants from pressures to agree with opinion leaders. They also create straightforward profit incentives that encourage participants to search for better information.
Many academics across the political spectrum are excited that prediction markets could improve decision-making in a whole host of arenas. Yet regulatory restrictions imposed by federal and state anti-gambling laws make these markets risky to operate. The Iowa Electronics Markets--deemed by regulators to be a teaching device--represent the only public forum in the United States in which the technique can be used with great flexibility, low cost, and little fear of government intervention.
Other publicly accessible markets for unconventional futures contracts, such as HedgeStreet, are not a direct substitute for a research-oriented prediction market. Compliance with regulation designed for high-volume commercial futures markets is expensive; and the sorts of futures contracts available are limited to those with a clear economic function--hence, for-profit U.S. exchanges do not offer contracts on elections or noneconomic policy decisions. That is a key reason the leading commercial prediction market, Intrade, is based offshore.
The Future of Information Markets
So what can be done? A consensus plan endorsed by more than twenty leading researchers--including Nobel laureates in economics Kenneth Arrow, Daniel Kahneman, Thomas Schelling, and Vernon Smith--and published by the AEI-Brookings Joint Center suggests the creation of a safe harbor for small-stakes, not-for-profit prediction markets to encourage experimentation. One could, for example, introduce exemptions for research-focused markets in which the size of individual investments does not exceed $2,000 per participant. The Commodity Futures Trading Commission (CFTC) could provide this safe harbor in the form of a "no-action" letter. Alternatively, the commission could create formal guidelines that make it cheaper and easier to start these markets.
Researchers should also have broad latitude to experiment with the design and subject matter of prediction markets. The CFTC may want to restrict these experiments to markets that can potentially improve economic decisions or mitigate financial risks, but otherwise it should adopt a laissez-faire approach. Prediction market research could shed light on how to increase the depth of the markets and make them less susceptible to manipulation. It could also address politically contentious questions, such as how to prevent criminals from benefiting from the use of these markets--say, by allowing terrorists to buy contracts that pay off when an attack is actually made.
Ultimately, Congress has a role to play here, too. At the very least it should fund CFTC regulation of prediction markets. And if the social benefits of these markets prove to be as significant as we expect, the CFTC should exempt them from prosecution under federal anti-gambling laws and preempt state laws that similarly inhibit their operation.
Prediction markets have become more than fodder for television news features on what those zany Internet folks will think of next. They are coming of age as serious tools for information gathering and analysis--tools with great potential for improving the efficiency of government and the productivity of industry. To help achieve that potential, Washington needs to nurture their development and keep them from becoming collateral damage in the endless war over who can gamble and where.
Robert W. Hahn is a resident scholar at AEI and executive director of the AEI-Brookings Joint Center for Regulatory Studies. Paul C. Tetlock is an assistant professor of finance at the University of Texas at Austin, McCombs School of Business. They are coeditors of Information Markets: A New Way of Making Decisions (AEI-Brookings Joint Center, 2006).
This document is available here as an Adobe Acrobat PDF.
