Politicians have been floating the idea of permitting a casino in Chicago since at least the Jane Byrne years—in 1979, before I was born ("it would be the greatest thing that could happen to Chicago’s Loop," Ed Burke told the Trib). It comes up periodically when the city is in dire financial straits, so you’ll forgive me if I assumed the latest proposal would go nowhere. Not so, apparently: a bill that would permit a casino in the city and expand gambling opportunities elsewhere has passed the Senate, with the lobbying assistance of Mayor Emanuel, and awaits Pat Quinn’s desk.

I’m particularly interested in the intersection of gambling and public policy because presents important questions about politics in a pure, nearly abstract form: What’s the role of government in protecting its citizens from themselves and each other, versus allowing us personal and economic freedoms? Where does a regressive tax end and a progressive one begin? What’s the difference between one economic luxury and another? How do we combine politics, ethics, and data to answer these questions?

For instance: gambling is almost literally everywhere in Chicago. You can do it in almost any convenience or grocery store by buying a lottery ticket. Practically every block in Chicago offers an opportunity to blow money on a game of chance, as is the case throughout the country. As a result, way more people play the lottery than go to casinos. According to a 1999 National Opinion Research Center study from the University of Chicago, 63 percent of the population had played the lottery in the last year, and 26 percent gambled in a casino.

So that’s more of a scourge, right? Well, according to a 2002 report from the Journal of Gambling Studies (referenced here: PDF) that tracks with some other numbers I’ve seen, lottery players won or lost an average of $11 on their last play; casino players won or lost an average of $143.

With casinos, you’re talking about more volatile, and hence more risky, amounts of money. With lottery tickets, the figures are considerably more modest. It’s much, much easier to go deep in a hole at a casino than it is at a 7-11.

But lotteries lower the bar to gambling, making it quick, cheap, simple, and practically omnipresent, more like picking up the milk than going out for a special occasion. So one would assume that lotteries are more regressive than casino gambling. And there’s research to suggest that.

Emily Oster, a gifted young economist at the University of Chicago’s Booth School, looked at the question of whether state lotteries represent regressive taxation (PDF). Reviewing the literature, she found that it almost uniformly suggests that lotteries are regressive. But her research uncovered something very interesting: looking at the demographics of Powerball players in Connecticut, Oster found that the median income of lottery players increased as the jackpot grew. Oster speculates, with great caution, that a lottery would become progressive at a jackpot of $806 million.

Oster suggests a couple reasons for this. First, that the lottery is more "fun" as the pot grows relative to your income. Second, "if income tracks education, the rich may have a better understanding of the odds"—in other words, even though the odds are still pretty absurd, it makes more sense to play them when the jackpot reaches astronomical levels. Either way, in my personal experience watching my middle-class family play the lottery, I’ve found Oster’s findings to match my anecdata.

Granted, this is a bit of a tangent, but Oster’s policy conclusion is relevant:

It is worth noting, with caution, that there are some policy prescriptions that may come out of the empirical result in this paper. First, concentrating lotteries so there are fewer games, with longer odds and higher jackpots, could allay some fears about regressivity. In fact, there has been a trend in this direction even without the knowledge of the jackpot-regressivity relationship as lottery organizers have noticed that people appear to respond to the jackpot size much more than the expected value of a ticket (Forest, Simmons and Chesters 2002). Second, consistent with the evidence in Price and Novak (1999,2000), out of sample predictions here suggest that instant lottery games are substantially more regressive than higher-stakes lotto games. This suggests that, if regressivity is perceived to be a problem, interventions should focus more on the instant games and less on large jackpot lottery products.

It’s useful in part because it sums up what I think is the conventional wisdom when it comes to splitting the gambling baby. On one hand, most people (strict libertarians aside) are uncomfortable with gambling as a regressive tax on the poor, desperate, and/or stupid.

On the other, a lot of people are uncomfortable restricting the pursuit of happiness, and find the restrictions strangely enforced. For example, why is it illegal to play poker, which is difficult and complex, and thus theoretically valuable and entertaining, but legal to scratch a card for money? Who are we protecting, and what are we protecting them from, anyway?

So Oster’s suggestion, as an economist, is to change the structure of the game so that it encourages more participation from those who can afford it, and less from those who can’t. It’s a compromise: the freedom to gamble is still permitted, the revenue stream is assured, and the weight of the tax shifts up the income scale.

Now, if one were to apply this to casinos, the conclusion is obvious: more cards, fewer slot machines. Slot machines are practically identical to lottery tickets, a cheap fix that requires no engagement and a minimal financial risk per play. Card games, on the other hand, demand a minimal understanding of odds-making and require a greater risk from the player. And as you might guess, the demographics track that:

The demographic analysis of those gamblers showed that a typical Massachusetts visitor to Mohegan Sun or Foxwoods is in or around their 40s. A typical slot machine player is a woman age 40 or older, with an annual household income of $25,000 to $75,000, while a typical table games (poker, blackjack, etc.) player is a male age 40 or younger, with an annual household income of $75,000 or more.

A study commissioned by Harrah’s (referenced here: PDF) found that the median household income of casino players was higher than the national median income: $56,663 to $48,997, though I’m guessing that would change if casinos became more prevalent and casino gambling became less of a destination affair.

To broadly generalize: lottery games have the potential to less harm to more people, many of whom cannot afford it. Card games have the potential to do more harm to fewer people, who are more likely to be able afford it. (The NORC study asked gamblers how much they were ahead or behind in the past year. Obviously self-reporting one’s gambling success should be taken with a massive grain of salt, but the breakdown was $36k/$30k for casinos, $1420/$6099 for the lottery. I kind of doubt that casino gamblers are coming out ahead—that’s not really how it works—but the stakes are obviously higher.)

Eric Zorn, for one, can’t decide whether he’s for or against the expansion of gambling in Chicago and Illinois. I don’t think he’s alone, since I have a lot of the same conflicting opinions:

* The social liberal in me doesn’t like casinos, but thinks the lottery is probably worse.

* The libertarian in me thinks that casinos should be legal, and thinks it’s especially dumb that the lotto is legal in Chicago but casinos aren’t.

* The gambler in me thinks card games have entertainment and intellectual value, while the lottery is dumb.

* And the wonk in me thinks if you’re going to allow more opportunities for people to gamble, it should be weighted towards games of skill and not games of chance. (The bill in question would permit slot machines in airports and racetracks. Having flown many times through Las Vegas, I should forewarn you that airport slot machines are impossibly irritating.)

In short, all gambling is not the same, even if it gets lumped together in legislation. For example, you could make a plausible argument that the lottery is worse than casinos, since it’s more likely to attract the poor, and the player has no chance to swing the odds in his or her favor. But you could also make a plausible argument that casinos are worse, because more money is typically at stake for the individual.

The debate, in other words, is subtly complex, and the data only gets you part of the way there. At the end, you’re left with some of the most difficult, fundamental questions of self-governance, not to mention a state that’s in debt, desperate for cash, and apparently more willing to gamble on those questions than it has been in decades.

(And that’s not even getting started on the relationship between casinos, crime, problem gambling, unemployment, and so forth, which I hope to touch on tomorrow.)

Photograph: Cordey (CC by 2.0)