Tuesday, February 21, 2006

Hahn & Tetlock on Legalizing Prediction Markets

My current writing project discusses a variety of strategies for legalizing prediction markets. Very little scholarship on the topic exists. I wrote about it a few years ago, granted, but even I regard that effort as incomplete (hence my present project). I thus took great interest in a forthcoming paper offering a fresh perspective on legalizing prediction markets: Robert W. Hahn & Paul C. Tetlock, A New Approach for Regulating Information Markets, J. REG. ECON. (forthcoming). Thinking that some of you might share my interest, and confident that I would benefit from your comments, I here offer from my draft paper an excerpt, absent footnotes, wherein I review Hahn and Tetlock's commendable, if flawed, proposal:

- - -

In one of the few academic papers to address the legality of prediction markets, Robert W. Hahn and Paul C. Tetlock argue that the CFTC should have the power to regulate the field. The authors argue that extant law gives the CFTC sufficient authority to act on their suggestion. Should that prove infeasible, however, they would welcome new legislation clearly giving the CFTC jurisdiction over qualifying prediction markets. Hahn and Tetlock's proposal would at all events preempt state regulation of prediction markets and clarify their legality, subject to various regulatory conditions, under federal law. In that, the authors deserve credit for setting forth a relatively detailed and plausible account of how prediction markets might escape from the legal failure that now inhibits their development.

The law can fail in many ways, however. Although Hahn and Tetlock's regulatory scheme would certainly dispel much of the legal uncertainty surrounding prediction markets, it would do so only by imposing significant legal burdens on them. Perhaps, on net, we would come out ahead. We might rationally favor a few heavily regulated but clearly legal prediction markets to none at all. So, at least, Hahn and Tetlock appear to reason. That supposed policy bargain suffers from an excluded middle, however: The option, described here in various guises, of clarifying the legality of prediction markets and regulating them with little more than generic contract and anti-fraud laws.

Hahn and Tetlock probably never even considered that third, relatively light-handed approach to legalizing prediction markets. It would have been ruled out by the two questionable assumptions driving their legal analysis: First, that prediction markets would necessarily qualify as gambling services under state law, thus making federal preemption necessary; and, second, that winning preemption would require a fair amount of regulation by some federal agency. Hahn and Tetlock thereby both misdiagnose the legal disease afflicting prediction markets and prescribe an unnecessarily painful cure.

Hahn and Tetlock overestimate the likelihood that that courts would place prediction markets at the mercy of state gaming laws, placing it at near certainty. Granted, there remains some risk that an over-eager attorney general might attack prediction markets as illegal gambling conspiracies. A cool-headed analysis of gaming law and policy indicates, however, that courts would most likely protect prediction markets from such legal grandstanding. That largely obviates Hahn and Tetlock's primary justification for seeking the preemptive protection of CFTC regulation.

Nonetheless, we cannot fault Hahn and Tetlock for wanting to protect prediction markets from whatever uncertainties state law threatens. Nor can we fault them for seeking shelter under federal preemption. The problem arises when they assume the necessity and propriety of having the CFTC regulate prediction markets largely as the agency regulates existing futures markets. Granted, Hahn and Tetlock would have the CFTC hold off, at least presumptively, from burdening all prediction markets with the full weight of its regulatory authority. In particular, they advocate two exemptions that would give fairly free rein to small-stakes prediction markets: one exemption for those that sharply limit the size of investments and another for prediction markets that stay below certain volume or revenue caps. Of any prediction claim that falls outside of those limits, however, Hahn and Tetlock say, "[T]he CFTC should regulate it as a futures contract."

Should it, though? Probably not. Hahn and Tetlock err in giving the CFTC too broad a jurisdiction over prediction markets. They would define the ageny's authority over such markets based on an "economic purpose test." Their test would authorize the CFTC to regulate any prediction claim that either provides significant financial hedging opportunities or the price of which "could provide valuable information for improving economic decisions." The first criterion should cause no alarm; it basically echoes the current test for establishing CFTC jurisdiction. The second criterion, however, would extend CFTC jurisdiction far beyond its current limits, to dangerously uncertain bounds. What contract price does not provide "valuable information for improving economic decisions"? Hahn and Tetlock try to justify their second economic purpose test as "a logical extension" of the price discovery and dissemination functions that at least in part justify CFTC regulation. However beneficial, though, price discovery and dissemination cannot alone suffice to define the agency's jurisdiction. Under the CEA Act, as written and interpreted, those represent at most necessary but not sufficient conditions for CFTC authority.

To their credit, Hahn and Tetlock recognize that they may have misjudged the extant authority of the CFTC. "[W]e are not legal experts," they confess. Foreseeing that the law might block their proposed regulatory fix, Hahn and Tetlock suggest that, as a fallback strategy, "Congress could provide more explicit guidance on the type of market it wants the CFTC to regulate."

Having thereby dealt with the legal objections to their proposal for regulating prediction markets, Hahn and Tetlock turn to defending it on grounds they doubtless find more familiar: via cost/benefit calculations. Their economic analysis suffers from the same sorts of blind spots that plague their legal analysis, however. Hahn and Tetlock recognize only two costs to their proposed regulatory program: an increase in the CFTC's workload and a slight risk that legalized prediction markets might generate some of the same social costs allegedly associated with gambling. They propose increasing the CFTC's budget to cure—or at least shift to taxpayers—the first cost, and discount the latter cost as easily outweighed by the many benefits that would follow from legalizing CFTC-regulated prediction markets.

Hahn and Tetlock thus do not consider the largest cost that their proposal would impose: The opportunity cost of burdening prediction markets with unnecessary regulations. That lacuna in their economic analysis follows directly from a lacuna in their legal analysis. As noted above, the authors apparently assume that prediction markets face a choice between prohibition by state anti-gambling laws or regulation by federal agencies. In fact, the law admits several alternative strategies for legalizing prediction markets, strategies that chart a course between the rock of outright prohibition and the hard place of heavy-handed regulation.

- - -


Jason Ruspini said...

Regarding the second criterion, Richard Shilts, Director of Market Oversight at the CFTC said last week in New York that they were thinking about this very point — whether they have authority over non risk-sharing, price-discovery markets.

Some of the markets that the CFTC has authorized arguably do not have significant hedging opportunities. Does this mean that their jurisdiction has already expanded? Take a market tied to the non-farm payrolls release. The vast majority of risk-sharing here can already be accomplished with equity futures. Does anyone have financial exposure to that release per se? Sure, the assembly line worker at GM might, but by that logic Michael Jackson's zookeeper would have also had a legitimate hedging opportunity in the Tradesports market tied to his trial, not to mention Jackson himself.

Tom W. Bell said...

I'm willing to bet, Jason, that the parties offering the questionable contracts you cite willingly sought CFTC jurisdiction. They were almost certainly exchanges already regulated by the CFTC, used to dealing with the agency, cautious to stay in its favor, and eager to continue winning the CEA Act's preemption of state law. In other words, the CFTC didn't exercise its authority to anyone's chagrin.

Jason Ruspini said...

Actually, CME's non-farm payroll auctions are not CFTC-regulated, and only ECPs can trade. Something to keep an eye on though.. the CFTC possibly expanding its authority while some exchanges gain advantage.