The Leaky Bucket Fallacy

Robert Waldmann

It appears that the worst tax ever TM is the brainchild of the normally sane Olympia Snowe.
How could a sensible intelligent person come up with such a terrible idea ?
What is wrong with the leaky bucket metaphor for the economics of redistribution ?
What is the leaky bucket metaphor for the economics of redistribution ?
What is this worst tax ever anyway ?

Answers after the jump.

The worst tax ever TM is the proposal that large employers who don’t provide their employees with health insurance pay the subsidies that those employees receive. It is a terrible tax, because the subsidies are supposed to go to people who need the money and the proposal will give firms incentives to avoid hiring people who need money and instead hire people who don’t.

I think that Sen Snowe was confused, because she took the leaky bucket metaphor seriously.

The leaky bucket metaphor is very simple. It is generally agreed that most taxes and transfers create dead weight losses because they distort incentives. This means that the dollar benefit to the recipient is less than the dollar cost to the taxpayer. If the recipient needs money more than the taxpayer, the transfer can increase average welfare, but not as much as it would if there were no dead weight losses. This encourages the use of a metaphor that redistributing wealth is like carrying water in a leaky bucket.

If we tax the metaphor literally, we conclude that if we want to redistribute it is best to carry the leaky bucket a short distance than a long distance. The loss of water is proportional to the distance we carry it.

Returning to public economics we discover that this insight is worse than useless. There is no meaningful way in which it is more efficient to redistribute wealth a short distance than a long distance.

Simple models (public finance 101) suggest that income taxes are more efficient than excise taxes and that a broad based income tax is better than a narrow based one. Basically, the standard result is that if the problem is that some people are poor and their is an opportunity because some other people are rich and have more money than they need then take from the rich and give to the poor. This means that wealth will travel long distances (Say from Alaska to Mississippi) and that the net contributor will have nothing to do with the net recipient.

It very definitely says that the people bearing the burden should not just be the employer of the poor person. Thus there is a presumption from public finance 101 that the earned income tax credit is more efficient than the minimum wage. Similarly HR3200 (a payroll tax on large employers who don’t provide insurance) is more efficient than the Baucus Snowe plan. I think the appeal to Snowe is that the money wouldn’t go far, just from an employer to an employee. This pretty much maximises the social cost of the resulting dissincentive to hire.

The leaky bucket has played another role in the gang of six negotiations. Back when he was pretending to negotiate, Sen Grassley used a stupid pun and stupid visual aids to denounce the HR3200 provision that subsidies be funded by a tax on high incomes. He insisted that money for health insurance subsidies must come from within the health sector. Now there is a case for taxing especially expensive insurance, but this is not that case.

More generally, the employer mandate itself is a case of a mess made by the leaky bucket. A tax funded single payer plan would probably be more efficient. It is politically impossible for obvious reasons, but one very silly reason is that the money would appear in the federal budget. Mandated insurance is like taxing the employuer and buying insurance for the employee, but mandated insurance does not appear as taxing and spending in the budget. So what ? I mean it makes no difference so who cares ?

Most people care, but they shouldn’t.