I can see I’m not nearly as libertarian as Scott Sumner is. In a recent post he makes the interesting and non-obvious observation that as presently operated state lotteries are a very regressive form of tax:
Today I’ll focus on another policy shift that would reduce inequality—sharply cutting the tax on lottery tickets, which now absorbs roughly 40% of all money gambled in state lotteries. Here are some points worth considering:
This floored me: Americans in the 43 states where lotteries are legal spent $70 billion on lotto games in 2014.
Seventy billion? I thought. No, that’s impossible. That’s more than $230 for every man, woman, and child in those states–or $300 for each adult.
But it’s true: According to the North American Association of State and Provincial Lotteries, lotteries took in $70.1 billion in sales in the 2014 fiscal year. That’s more than Americans in all 50 states spent on sports tickets, books, video games, movie tickets, and recorded music sales. . . .
Lotteries set aside about 40 percent of their ticket sales as state revenue that often goes to schools. Then, winners of more than $600 are subject to 45 percent windfall taxes on their good fortune. “The house” is winning, even when it’s losing.
But it’s the poor who are really losing. The poorest third of households buy half of all lotto tickets, according to a Duke University study in the 1980s, in part because lotteries are advertised most aggressively in poorer neighborhoods.
Basically, lotteries are a very inefficient transfer from low income earners to other low income earners. I would think that we would want to encourage transfers from higher income earners to lower income earners but apparently not.
I’m not sure that analysis is complete. I think you’d need to factor in the income effects of the state spending the lottery revenues support. Since the state spending goes overwhelmingly to support the income of people in the professional class who earn between $60,000 and $300,000, I think that lotteries are probably an even more regressive form of taxation.
But that’s true about gambling more generally. In places where gambling is legalized most of the money is spent by people in the lower third of income earner who, except in a very small number of places, live locally. The proceeds go either to national gambling operators (like Donald Trump) who don’t live where the gambling is going on or to the state where it has the same economic effects as state-sponsored lotteries. There’s a transfer from the localities in which gambling takes place to other localities.
I also want to remark on his aversion to banning lotteries:
Some misguided people want to help the poor by banning state lotteries. But that would make things even worse. Buyers would have to rely on the black market, or drive all the way to New Hampshire or Rhode Island to buy their tickets. The solution is simple, cut the state’s take from 40% to 10%, and make up the lost revenue through higher sales and property taxes, where the burden is shared more fairly across all of society.
I’m not so sure about his point about sales and property taxes is correct but I think his observation about black markets is even worse. I believe it’s based on the frequent untrue assertion usually phrased as “Prohibition didn’t work”. That is only true if you think that the purpose of Prohibition was to eliminate the consumption of alcohol. It wasn’t. The Volstead Act had provisions that allowed people to make beer and wine for their own use. Its purpose was to reduce alcohol abuse and it was quite effective in doing that. The rate of alcohol abuse declined sharply when Prohibition was enacted and did not recover to its previous high rate.
While I agree with him that it’s obvious that ending state-sponsored lotteries wouldn’t eliminate lotteries, I’m skeptical that it wouldn’t reduce the rate at which people spend money on lottery tickets. He hasn’t proved it. He’s just claimed it.