Against hypothecation

One of the big reasons I don't like hypothecated taxes - taxes that are tied to particular projects - is that they lead to nonsense like this:

Would you support a soda tax? What if it was used to support universal pre-k? https://t.co/l36a7O9yVD pic.twitter.com/OhNvw3FKyE

— The Upshot (@UpshotNYT) April 4, 2016

Hey, everybody can play this game.

Would you support a tax on frappucinos if it supported orphans with cancer?

How about a tax on hipster beard oil if it supported recovery programmes for the New Zealand saddleback?

How about a daft tax on capital ownership at a high deemed assumed rate of return to support a basic income sufficiently low that there'd be massive political pressure to layer a welfare system on top of it?

Basic formula: find a product some groups would see as a bit of a frivolous luxury or as 'unnecessary' (and especially if those consuming the product can earn disapprobation from its consumption), bundle it with spending on some nice sounding thing that nobody could oppose, and tah-dah! You've ruined tax policy.

Here's a basic lesson in public finance. Set your bundle of taxes so that it raises money at lowest possible economic cost. Some taxes are actually efficiency increasing: true Pigovean taxes. Use those first. Once those are set at the right level, work your way up through other taxes until the marginal social cost of raising an additional dollar in taxes, including the deadweight costs of raising the money, matches the marginal social benefit of the next unit of public expenditure. Then stop.

If some tax you think is a good idea really is a good idea, work it into the tax base first. If the marginal social cost of raising tax revenue is then lower, you can either reduce total taxes by cutting back on the least efficient ones, or add in a bit more public expenditure on the next thing down the cost-benefit list, or a bit of both.

Think of it this way: if the tax made sense, why would it make sense only when bundled with some politically attractive spending programme?

There are a few exceptions. Road user charging through petrol charging is far more a fee for service than it is a hypothecated bundle. And some proposals levy taxes to directly tax the winners from a programme to compensate the losers: like my proposal for Canadian dairy.

But if somebody's selling you a tax-and-programme bundle, they're usually pulling a fast one. Look first, and carefully, to see if there's really good reason that the two things should be bundled together. If not, well, draw your conclusions.