Placing Bets

Wagering, it appears, is a routine feature of life on the Court. From betting on how much snow has fallen on the Court steps to participating in a monthly penny-ante poker game, a Washington Post reporter has observed that it “seems that they'll bet on just about anything up at the U.S. Supreme Court” (Mulligan 1987; see also Kennedy 2003; Nusbaum 2003; Taylor 1988). Among the justices, Rehnquist is reputed to be the Court's most enthusiastic gambler. Indeed, one participant in his monthly poker sessions has noted that “The Chief Justice really keeps everybody moving fast. If people start telling jokes and talking politics and things like that, he says, ‘Come on, let's move things along and play poker and stop the folderol’” (Taylor 1988).²

Although the Chief organized the 1992 election pool, Justices Byron White, Antonin Scalia, and David Souter all recused themselves. Why? White, an erstwhile All-American and professional athlete, may have been schooled in the necessity of avoiding associations with gamblers;³

The six members of the Court who did participate in the pool each anted up $1 on whether Bill Clinton or George H. W. Bush would win each of the 50 states and the District of Columbia—a commitment of $51 per justice. The rules were simple. Every justice who correctly predicted the winner in a state would share equally in the $6 pot for that state. Thus, if all six justices got the winner of a particular state right, each would receive $1. If the bench was divided on a state's outcome, the money wagered would be split among those correctly forecasting the result. For example, as shown in Figure 1, Justice Thomas and the Chief incorrectly predicted that President Bush would carry Kentucky. As a result, the four justices who correctly predicted that Clinton would win Kentucky each earned $1.50 on that outcome.

Who Won?

Justice O'Connor was the big winner in the pool; as the Chief put it in the cover memo that accompanied the results, “The net result is that Sandra has won $18.30, Harry has won $1.70, John and I have lost $6.30, Tony has lost $2.30, and Clarence has lost $5.10.” That affirmed the widely held impression that O'Connor, the reputed swing voter on the Court, has sharper political instincts than any of her fellow justices.

To help us understand this outcome, we used the scientific method⁵

1. A justice's general attentiveness to the news, the idea being that accurate forecasting depends on solid information. The key to testing this hypothesis is Justice Thomas, who “reads no morning newspaper; he canceled his subscription to the Washington Post, and no longer reads the Washington Times either. He rarely watches TV news, depending largely on his assistants and friends for verbal news briefings” (Fisher 1995). According to Thomas's close friend Paul Weyrich, “When I mention to him things in the news, he just evidences no knowledge of it. He's just totally unaware of news events” (Fisher 1995). To be sure, Thomas does listen to Rush Limbaugh—“one of his few mass media sources” (Fisher 1995)—but his disinclination to keep abreast of the news prompted us to create a dummy variable (THOMAS) distinguishing his forecasts from those of the other pool participants.⁶

   That Thomas would even participate in gambling seems surprising, given his general aversion to risk-taking. For example, although he delights in cruising I-66 in his black Corvette IROC, he never exceeds the speed limit; in the words of a friend, “He drives like a turtle” (Fisher 1995).

2. The justice's first-hand familiarity with the political dynamics of a state. To test this hypothesis, we created a dummy variable (HOME STATE) indicating for each observation whether the justice in question had grown up, attended school, or practiced law in the state (Epstein, Segal, Spaeth, and Walker 1996, Tables 4–2, 4–4, 4–6).
3. The justice's ideological extremism, the idea being that a moderate would be less reluctant than an extremist to project the candidate he or she did not favor as the winner of a state. To test this hypothesis, we determined the relative frequency with which a justice had voted on the liberal side in cases that came to the Court before the 1992 term (Spaeth 1999), and used the absolute difference between that number and 50 to gauge his or her ideological extremity (EXTREME).
4. The political competitiveness of the state. The more dominant either the Democrats or the Republicans had been in a state, the easier it should have been for a justice to forecast the 1992 outcome for that state. To create a SAFE STATE variable, we subtracted 50 from the winning candidate's vote share in the state in the 1988 presidential election.

Of the justices' 306 predictions,⁷

Should We Care that Justices Gamble?

Evidence of gambling on the Court raises questions about the propriety and broader significance of the justices' behavior. To be sure, friendly bets are commonplace among politicians.⁸

Although it is tempting to excuse gambling on the U.S. Supreme Court as a harmless diversion from the grim, life-and-death business that the justices are called upon to consider, the issue is not nearly that clear-cut. The business consulting firm of Challenger, Gray, & Christmas has estimated that office pools cost U.S. employers billions of dollars per year in lost worker productivity (Dybis 2004). With that in mind, Figure 2 charts the number of signed opinions that the Court issued during the last seven presidential election cycles. The solid bars show opinions during the three years preceding the election, while the striped bars show opinions during the presidential election year, when the justices (with the obvious exception of Clarence Thomas) were preoccupied with cramming for the office pool. Clearly, the Court's productivity sags during election years. Skeptics may dismiss this as a coincidence, but closer examination of Figure 2 suggests otherwise. For one thing, the only time there was no productivity dropoff was 1996—an exception that helps prove the rule, because forecasting the 1996 outcome couldn't have been very time-consuming, given Bob Dole's stillborn candidacy. Moreover, the largest productivity decline occurred in 2000— hardly surprising, given the obvious difficulty of accurately forecasting such a tight race. Finally, note the sharp decline in productivity since Rehnquist's elevation to Chief Justice, consistent with his proclivity for getting his colleagues involved in gambling operations.

Court Workload

Of course, some might say that having this Supreme Court decide fewer cases would not be such a bad thing. Still, there are other reasons for concern. For example, having the Court rule on cases that involve gambling (e.g., Chickasaw Nation v. United States 2001) is like assigning a fox to guard the henhouse. Likewise, if the Court were ever called on to pass judgment in a case that could decide a presidential election—a remote, science-fiction scenario,⁹

Finally, the outcome of the pool may influence the justices' sociability. During an election night party in 2000, Justice O'Connor apparently became upset when CBS anchor Dan Rather called Florida for Vice President Gore. She exclaimed, “This is terrible!” and then proceeded, “with an air of obvious disgust,” to walk off to get a plate of food (Thomas and Isikoff 2000). Speculation abounded at the time about why O'Connor was so distraught,¹⁰

Should the justices cease and desist from gambling? As mere empiricists, we are ill-equipped to address this normative question. Perhaps William J. Bennett, an acclaimed arbiter of America's moral values, could be persuaded to address this question in the next edition of his Book of Virtues (1993)— though the fact that Mr. Bennett regularly lays his royalties on the line in a poker game with Justice Scalia and Chief Justice Rehnquist (Kennedy 2003) may lead some cynics to question his moral authority on this issue.

Should Political Scientists Write Articles Like This?

Finally, should we, as political scientists, really be devoting ourselves to the collection, analysis, and distribution of information that some might disparage as gossip best suited for the “Drudge Report?”

We consider our research an important public service, for we are providing information that citizens want and need. Consider the following: Woodward and Armstrong's (1979) The Brethren, a tellall, behind the scenes, gossip-filled page-turner about the Court, was a #1 national bestseller. Maltzman, Spriggs, and Wahlbeck's (2000) Crafting Law on the Supreme Court, a theoretically sophisticated and impeccably researched social scientific analysis of decisionmaking on the Court, is currently languishing at #1,152,180 on the Amazon. com sales list. QED.