Process of deduction

A couple of commenters and emailers have asked me to defend my assertion over at TPMCafe that:

At the time, Gore was offering tax deductions or credits for practically anything one might do, from getting born to entering a nursing home. These sorts of tiny lump-sum deductions are generally frowned on by economists; they distort activity, are costly to administer, and unlike marginal rate cuts, provide no positive incentives to increase work.

This is not actually particularly controversial, and the economics is (I think) kind of interesting and important, but at the same time, the explanation is not quite right for the TPMCafe format, so I'll try it here. For anyone who cares, it's below the fold, as it's rather lengthy and a little bit technical. Apologies for those who know all this and find it old hat; this is only for wannabe tax wonks.

One very useful way to think about labor markets is to model consumers as trading off between two things: leisure (L) and consumption (C). This is simple, to be sure, but it yields some powerful insights, so it's a useful framework to have in your package of analytical tools. There is a straight tradeoff between those two things; any consumer who wants to consume more (over their lifetime), must ultimately give up consumption in order to do so.

Now we introduce a second framework to explain how consumers make tradeoffs between rival goods when prices change. Any price change has two effects on consumption:

  • The Income Effect: An increase in the price of one of your goods effectively reduces your income; your income now buys less than it did before. Conversely, a fall in the price of one of your goods allows you to buy more with your income, raising your demand for the goods you consume.
  • The Substitution Effect: An increase in the price of one of your goods causes you to alter the mix of goods you can consume. Usually, this means that you substitute away from the (now) more expensive good, but in rare cases, known as Giffen goods, price increases can actually cause you to consume more of them.

Economics students often use a little grid to model how price changes play out. So let's say that wages increase. This raises income, but also raises the cost of leisure.


The effect is to unambiguously increase the demand for consumption. But it is not clear what effect it will have on demand for leisure, because we don't know whether the income effect or the substitution effect on leisure is bigger. That's an empirical question to be resolved (she said hopefully) by data analysis. Without an empirical estimate, we don't know whether an increase in wages will cause people to work more, because they want to make more money, or less, because they're suddenly richer and can therefore afford to buy more of that valuable luxury good, free time.

Okay, if you're not bleeding out your ears, or asleep, here's why this matters for tax policy. A change in the marginal income tax rate models exactly like a change in wages, only in the opposite direction. Which is to say, if I raise your taxes, you've effectively just taken a pay cut--as everyone who's ever taken a high-paying new job in New York City can attest. Supply-siders should note that this is one reason that the simple story about cutting taxes increasing peoples' incentives to work and save doesn't always operate as advertised. They might just decide that they no longer need to work as hard to maintain their standard of living.

But of course, there are other ways to cut taxes besides altering the marginal rate. We could also give everyone a flat deduction--$500 off every tax bill, say. And that models rather differently from a marginal tax cut. Unlike a marginal tax cut, a new tax deduction doesn't make leisure more expensive. So people consume a little more stuff, but some of them also decide to consume a little more leisure. Since [Economic Output = Labor Productivity X Labor Supply], GDP is pushed downwards.

The targeted tax cuts of which Clinton/Gore were so fond are bad for another reason: they distort peoples' choices. In general, if an activity isn't a good idea without a small tax deduction, it isn't a good idea with one, either. Finally, they add to the complexity of the tax code, and if there's anything that economists from left to right agree on, it's that the simpler the tax code is, the better it is for everyone.

That's why economists are not fun of the targeted tax deduction, or the deduction in general. If we want to give more money to the middle class, then we should reduce their marginal tax rates, not pepper the tax code with a thousand little deductions. Larry Summers knew this, and agreed with it; it's just politically, that was the only way the Clinton Administration (and prospective Gore Administration) thought it could accomplish various goals.