Income tax rates above 50 percent?

One of the headline events at the NTA Annual Conference was a speech by Emmanuel Saez on fundamental tax reform. Emmanuel, a rightly multiply-prize-winning mainly empirical economist, is perhaps best known for his work showing how income distribution has changed over the last few decades, becoming vastly more concentrated at the very top. He has also done other excellent work - for example, in behavioral public finance with work showing how framing can affect consumer responses to substantively identical tax instruments.

But he has not previously been all that involved in thinking or at least writing about the tax base. Thus, it was with clear interest and anticipation that I sat in the big auditorium at the NTA main event session on Thursday afternoon, as Jim Poterba introduced him with the note that Emmanuel, unlike any of those other wimps out there who needed to collaborate with others in order to produce comprehensive tax reform plans (OK, Jim didn't actually quite put it that way), had personally designed his own plan. As a fresh voice in the field, what would he say?

The results were interesting in one sense, but frankly less so in another. Saez drew on his work about income distribution to show what the income tax system would need to do in order to reverse even just, say, the last 20 years of radically rising income concentration at the top. He accordingly endorsed a system that is so far removed from where the discussion is these days that I consider it big news meriting widespread attention. The debate ought to extend at least as far as he wants to go, whether one or not one agrees about going all the way there. But in other respects the talk's content was perhaps a bit less substantial.

In the optimal income tax (OIT) approach, the point to high tax rates on high-earners is to achieve redistributive benefits in excess of efficiency losses, or more precisely the optimize the tradeoff between the two. Typical OIT models have ended up (initially to the surprise of those designing them) to call for relatively flat rather than graduated rates, with the rate at the very, very top (in theory - not necessarily in practice) dropping to zero. Emmanuel instead calls for a very graduated system, with a top rate in excess of 50 percent. This is not because he rejects the theoretical basis for the OIT approach (utilitarianism or other welfarism), but due to his modifying some of the typical assumptions - e.g., by assuming a very low marginal utility of income at the top (given the impossibility of consuming so much wealth any time soon, non-monetary motivations for wanting to earn the highest amount, etcetera).

I find this plausible and think it absolutely should be more prominent in public debate. Perhaps we've moved too far in the direction of thinking low rates are necessarily better. Wherever one comes out in the end on this question, top rates above 50% ought to be within the publicly debated policy space (which currently they are not), and the fact that a prominent and leading tax economist endorses them is genuinely newsworthy.

The exact normative basis for his proposed rate structure was a bit less intellectually satisfying. Even if one is greatly concerned about the plutocratic turn that U.S. society has been taking (a key factor, I think, in our ongoing conversion into a third world banana republic), a specific inequality target, with the view that economic growth should be shared in a particular predetermined manner, is in tension with the OIT principle of optimizing the tradeoff between equity and efficiency. Technological changes that alter pre-tax wealth distribution - not to prejudge how important they've been compared to political economy factors and changing social norms - may indeed change where one wants to end up, by altering the costs and benefits. But one could defend targeting, say, 1989 wealth distribution as a simplifying political economy choice, even if it isn't how one should in principle go about finding the optimal tradeoff point.

So far, so good. More disappointing about the talk was that it's hard to really have much insight or depth in one's views about the myriad tax base issues that arise in this sort of exercise if one hasn't spent a lot of time thinking about them. So I thought the NTA crowd was really the wrong audience for this speech.

Take the income versus consumption tax choice, for example. Saez rejected the cash flow consumption tax because it has never been tried. No X-tax variant, possibly because he had never heard of it. And he appeared to think that, to make an income tax work, you just have to get rid of those darned loopholes. Not much understanding, so far as I could see, of the point that, once you have a realization-based income tax and need to collect corporate income at the entity level, you are really deep in the soup.

Not to be harsh here. I commend him for digging in. It's public-spirited of him, and unmistakably informed by the altruistic goal of performing a public service in lieu of burnishing his academic reputation (which depends more on doing cutting edge research, as indeed he has). He most affirmatively was not trying to position himself to enter public political life, as the contents would have been political suicide if that were his goal. Plus he is making an important contribution, as I noted above, by expanding the parameters of debate and "acceptable" opinion. But I'll admit to thinking during much of the speech: Why am I listening to how he'd resolve Issue X? Sure, I might resolve it the same way, or perhaps in a different way - although he generally got things right rather than wrong where there was such a rubric - but there's a whole lot more to think about here that one needs to have spent some time on to understand in real depth.

The plan's political prospects are not especially bright. For example, in addition to creating a top rate above 50%, he would repeal the home mortgage interest and state and local tax deductions. Leaving aside everything else, there's an unfortunate political paradox at work here that undermines efforts such as this. The richer and more powerful the top 0.1% is, on the one hand the greater the normative case for redistributing wealth way from them, but on the other hand the dimmer the political prospects for actually doing so. Getting wealthier makes them more powerful politically, not less. So they'd probably have to lose substantial ground for the political system to start treating them less favorably.

In that way, the political economy of redistributive politics has a kind of anti-insurance element to it. Winners are rewarded for becoming more powerful, while losers are punished for having become less so. Not an easy problem to try to address.