Better than a Bank Bailout: A Federal Lottery

Better than a Bank Bailout: A Federal Lottery


The growing outrage over the Fed’s rescue of too big to fail banks was stoked by news from Bloomberg that banks made 13 billion dollars in profit from low interest discount window loans. However, as many people have pointed out, the outrage isn’t that the banks were rescued – we’d all be worse off if they had been allowed to fail. The outrage is that similar help was not offered to households struggling with the recession, in particular the lack of job creation initiatives to help workers looking for jobs in such a lousy economy. The banks were bailed out even though they played a large role in creating the problems in the economy, but the typical household – a household that did nothing to cause the crisis – was all but ignored by policymakers.

It didn’t have to be this way. It would have been possible to help both banks and households and in the process avoid much of the anger over the inequities in the bailout process. Here’s how. Suppose that when the crisis hit, we had held a national lottery. The lottery would be based on Social Security numbers, and all adults with a Social Security number would be eligible (tax return data could be used to limit the lottery to one adult per household). All households would have an equal chance of winning, and if your household’s number comes up one of several things will happen. First, if you have a mortgage, then the government will pay off one third of it up to a maximum amount. If you owe less than the maximum, you get the balance in cash.

Second, if you have student loans, the government will pay them in full if your number is drawn (up to a predetermined maximum). And again, if your loan is less than the maximum you get the balance in cash. Finally, if you are a renter and do not have student loans (i.e. everyone not in the first two categories), then you get cash if your number is drawn. The assumption is that these would be mostly households with high propensities to consume and hence the cash would be mostly spent, but the winning households would be free to spend it as they wish.

Having the government pay off one third of a
large number of mortgages would result is a
substantial cash infusion to banks.

If the program was for $250 billion, and if the mortgage or student loan benefit was capped at $75,000, then there would be six million winners. And if the total payout was bigger, or the payout per person smaller, that number would be even larger.  Importantly, all households would have an equal chance at having their number drawn, and hence anger over who gets the bailout would diminish. Since the benefit would go to households first, not banks, that should reduce anger as well.

But banks would still benefit from this program. Having the government pay off one third of a large number of mortgages would result is a substantial cash infusion to banks. It would be as though the banks had the ability to instantly call in a substantial fraction of their loans to provide reserves against losses. Thus, if the size of the program is sufficient, then those reserves ought to be more than enough to counter losses from bad mortgage loans. The mortgage and student loan payments would be made directly to banks on behalf of households, and they could be required to hold a substantial fraction of the payment as reserves against losses. Thus, banks would be bailed out, but households would be the main beneficiaries of the program.

The program could be adjusted in various ways to target particular populations. For example, the lottery could be limited to households with less than $75,000 in income (or some threshold), or if you are unemployed then the payment could be doubled, etc. However, one of the virtues of an equal chance lottery is that it avoids moral hazard problems among households. No matter how you change your behavior, it doesn’t affect your chances of winning so moral hazard of this type is eliminated. But if the payment is higher for some categories, e.g., when the lottery winner is unemployed, has an underwater home, etc., then there is an incentive to have yourself classified in this way and moral hazard questions resurface. In addition, if payments are higher for those who are in the most trouble you then give the appearance of rewarding those who have somehow managed to create the biggest problems for themselves. I don’t think the moral hazard problems are severe, so adjusting the payments would not be a big problem, but the perception of unfairness that is interfering with our ability to address economic problems would resurface.

From an economic perspective, bailouts that directly target the underlying problem are best. But when the best bailout from an efficiency perspective is to hand money to people who played a large role in creating the problems, or are perceived to be undeserving of help, then questions of equity come to the forefront and interfere with our ability to address economic problems. The best overall policy is one that gets the job done as efficiently as possible while still being politically viable and a national lottery is attractive for that reason.

When the crisis hit, we would have had difficulty holding a national fiscal policy lottery even if we had wanted to. The time it would have taken to plan and execute such a program would have been prohibitive. But with planning we can do better, and hopefully avoid the need to Occupy Wall Street the next time this happens.