By keeping average taxes the same, while reducing marginal tax rates, it is possible to encourage people to earn and report more income.

This is from Alan Reynolds, “Marginal Tax Rates,” in the first edition of The Concise Encyclopedia of Economics, and it relates to the discussion yesterday of my post on the Romney tax-cut plan.

When the government cuts a marginal tax rate, and that’s all it does, that cut has two effects in opposite directions: a substitution effect and an income effect.

The substitution effect is to make leisure more expensive. If you’re facing a 35% marginal tax rate (MTR), for example, and the rate is cut by 20% to 28%, your “price” of leisure rises by (72 – 65)/65, or 10.8%. When the price of a normal good rises (and leisure is a normal good), you buy less of it. So people work harder.

The income effect is to make you buy more leisure. The cut in marginal tax rates increases your real income and therefore you demand more leisure. That is, you work less.

Empirically, the substitution effect tends to outweigh the income effect slightly for men and strongly for married women. This means that cuts in MTRs alone will tend to increase income somewhat for men and a fair amount for married women. That was the effect of Reagan’s cuts in MTRs in the early 1980s.

But what happens when the government doesn’t just cut MTRs but also broadens the base by limiting deductions and exemptions? Let’s say the government does so to make static after-tax income (before the effect of the higher incentive to make money) the same as before. Or another way to say it is that the government cuts MTRs but keeps average tax rates constant (in a static sense.) Then there is no income effect. That means that the only effect is the substitution. That’s how you can get a big boost in taxable income. Of course, that also means that the average tax rate in equilibrium will be lower (because of the added income), but the average high-bracket person will pay more income tax (because the static amount of income tax was the same and the person is paying more tax on the additional income.)

Of course, as I understand it, it’s unlikely that the Romney tax cut would get rid of so many deductions that people would pay the same average tax rate (in a static sense.) Still, if the tax cut substantially reduced deductions and exemptions but didn’t go all the way to making income taxes paid (in a static sense) constant, there would still be an income effect. But it would be attenuated compared to the situation where the only thing that changed was the MTR. So the odds are still that the substitution effect, even for men, exceeds, and probably strongly exceeds, the income effect. Result: more income to be taxed.

There is one fly in the ointment. Because Romney wants to limit deductions and exemptions more for higher-income people than for lower-income people, he would need to phase out those deductions and exemptions with income. That is, past some income level, as people’s income increased, their deductions would decrease. Over the income range in which this happens, there is an added implicit MRT. So, for example, if he allowed $25,000 in deductions for people with income up to $100,000 and $0 in deductions for people with income of $200,000 or more, and if he phased it out linearly, as is commonly done, then someone in the income range from $100K to $200K would find himself losing 25 cents in deductions for every additional $1 in income. If the tax rate in that region of income is, say, 20%, then the person pays an additional 5 cents in tax for every additional $ of income. That’s like an additional 5 percentage points added to the MTR in that region of income. And, of course, that causes an added disincentive to make income.