Not wonky enough?

Matt Yglesias, in a generally excellent column about the sorts of federal income tax changes that the Administration and Republican Congressional leadership may seek in the next couple of months, rightly notes that tax reform, in common usage, means “reducing tax rates without massively reducing federal tax revenue because the losses from the rate cuts are offset by closing loopholes and eliminating or curbing deductions.”  He also rightly notes that the Republicans are unlikely actually to attempt this, since base-broadening sufficient to offset the revenue loss is no fun politically. He then asks why, on policy grounds, one might favor bona fide tax reform:

“What’s good about tax reform?

“The official wonky case for tax reform stems from a divergence in the way that a normal person thinks about taxes from the way that an economist [does] ....

“To a normal person, taxes are a necessary evil. The evil thing about them is that after you pay taxes you have less money than you had before. Most people like money, so they like the idea of getting a tax cut, and they don’t like the idea of getting a tax hike, primarily because they are focused on the impact of tax changes on their after-tax income. If Congress changed the law to cut my taxes by $500, I’d be pretty happy about that. And I wouldn’t care whether they did that by tweaking the individual exemption, tweaking the dependent exemption for my 2-year-old, making his preschool tuition tax-deductible, or cutting my marginal income tax rate. At the end of the day, the point would be to get the $500.

“But from a supply-side viewpoint, these are very different policies. Making the individual exemption a little bit bigger puts $500 in my pocket but it doesn’t give me any new incentives to work harder and earn more money. If anything, by making me a little more economically comfortable it reduces my incentive to work harder and earn more money. By contrast, cutting my marginal income tax rate doesn’t just put $500 in my pocket. It makes it more worth my while to try to go out and hustle up some paid speaking appearances or otherwise find ways to earn a bit more.

“Tax reform is good, on this view, because it’s a way to greatly improve incentives without costing the government much in the way of revenue. The 1986 tax reform bill, for example, eliminated enough loopholes to pay for a cut in the top marginal income tax rate from 50 percent all the way down to 33 percent — drastically increasing the incentive of rich people to go out and try to become even richer.”

In the above text, Yglesias offers a generally useful explanation. However, he may overstate (or, at least, slightly misstate) the main character of the economist’s case for tax reform.  To show this, let’s make two tweaks to his illustration.  First, let’s suppose that I’d get $500 via Congress’s newly making preschool tuition tax-deductible, rather than by its either (a) raising personal or dependent exemptions or (b) lowering marginal tax rates.  Second, let’s suppose that I would decide to work and earn more if I were sending my child to a costly preschool than if I were keeping the child at home.

Now, with the tax treatment of preschool tuition tax being an input to my labor supply choice, it lines up more the marginal tax rate change, and less with changing the individual and/or dependent exemption. So, if we already had preschool tuition tax-deductibility as a feature of current tax law, and we were thinking of repealing it to fund a lower marginal tax rate, the “tax reform” might end up, at a first approximation, not increasing my incentive to earn. E.g., suppose that, when I was thinking of working more so I could send my child to preschool, the combination of (a) an increased after-tax return from my work and (b) an increased after-tax cost to the preschool, left me in the same place as I was before.

To make this a more plausible revision of the hypothetical, suppose that both decisions are scalable. E.g., I can work just a little bit more or a lot more; I can send my child to a cheaper preschool or a more expensive one. The bottom line may still be a lot less change than might otherwise have been expected to my labor supply incentives. I work to earn money to spend on things I want, and the base-broadening makes it costlier than it was before to buy the things I happen to most want.

Why point this out? In the 1986 tax reform, at least some studies found afterwards that work incentives hadn’t really changed much, by reason of the base-broadening. E.g., suppose I previously hadn’t minded the 50% top rate because tax shelters would prevent me from actually having to pay it. Now the rate was lower but the tax shelters were gone, so I would actually have to pay the headline rate. This may have been a good policy change for other reasons that I’ll get to in a moment. But it could mean that my incentive to earn more $$ wouldn’t actually have increased.

Now let’s turn to contemporary thought experiments in which, say, state and local tax deductions, home mortgage interest deductions, and/or 401(k) deductions are repealed to fund lower rates.  (In the 401(k) example, even if the current deduction is replaced by a future exclusion that’s slated to apply when I withdraw the money, suppose that I’m either myopic or else don’t trust future Congresses to retain the exclusion.) If I’m paying about as much tax on balance as I was before, it’s plausible, depending on the details and on my level of understanding, that my view of my after-tax incentive to work will also remain about where it was before, given the changes to the tax treatment of some of my earnings’ possible uses.

With that in mind, let’s re-ask Yglesias’s question: “What’s good about tax reform?” The answer, if any, is that, even if we haven’t substantially changed my incentive to work and earn, we have done so with respect to how I spend my earnings.

In terms of the 1986 example, it’s presumably good that I’m no longer investing in tax shelters (assuming that their tax benefits were socially undesirable), since I will now seek opportunities that offer greater expected pretax profitability. But it’s my inter-asset choices that are now less tax-distorted than previously, rather than my incentive to work and earn.

Likewise, for the 2017 scenario, suppose we see no relevant social difference between my spending my $$ on (a) preschool tuition or (b) fancy vacations. Then eliminating the distortion to my choice that resulted from the hypothetical tuition deduction is a good thing here, albeit by express stipulation. But if we don’t think the deduction was bad, e.g., because we see positive externalities to people’s sending their kids to preschool instead of spending the $$ on own consumption, then so much the worse for the case for “tax reform” here.

I myself tend to like the inter-asset efficiency consequences of a revenue-neutral tax rate cut plus reduction to the home mortgage interest deduction. But that’s because I don’t like the deduction, or more precisely what I take to be its main effects. The state and local tax deduction example is potentially quite different. So is the 401(k) deduction repeal, although there we would need to specify a lot more about what people expect, how they make savings choices, how we should think about socially optimal savings choices, etcetera.

Returning to Yglesias’s main point about tax reform, we’ve now both (a) made the economist’s argument wonkier still – by relating it to distortions in asset or activity or consumption choice, rather than to labor supply as such, and (b) shown that it needs to be evaluated case-by-case, in terms of the arguments for a particular tax benefit that would be eliminated to pay for tax reform.

Of course, as I agreed with Yglesias at the start, the Republicans appear unlikely to attempt bona fide “tax reform” (with significant base-broadening) anyway. But just in case anyone should ever attempt it in the future, I hope it’s useful to clarify the nature of the main argument for it.