What Can Tax Returns Tell Us About the Incomes of the Very Rich?

Brian Galle
Whatever Source Derived
2 min readAug 1, 2017

Last week’s meeting of the National Bureau of Economic Research saw, among other highlights, a major new paper on the sources of economic inequality. Drawing on never-seen-by-public administrative tax data, the authors (Smith, Yagan, Zidar, & Zwick, or “SYZZ,” which I’m pretty sure was the title of a Black Sabbath album) report on how top pass-through entity owners make their money.

The story is startling, for those who can recite Piketty from memory. SYZZ find that the vast bulk of income earned through pass-throughs is “active” business income. They find further that incomes at S Corporations drops sharply after the principal dies, suggesting that human capital, not wealth accumulation, is driving these firms’ success. These findings, they say straightforwardly, are inconsistent with Piketty’s theory that wealth accumulation — “capital” — is the major source of income for the wealthiest.

I don’t believe it — not yet, maybe not ever. To emphasize, these are tax returns. They can tell us only what their filers were willing to tell the government.

There are, therefore, at least two major problems with using these data to conclude anything about how top earners earn. The first is selection bias. Suppose it is the case that it is easier to conceal from the government income from capital accumulation than it is to hide active business income. Then the tax data will be systematically skewed towards active business sources.

That seems like a real possibility, doesn’t it? Active businesses have trading partners and employees who are also usually filing tax returns. They’ve got invoices sent and received electronically, making for a ready paper trail. More than that, they have a real footprint in the world — a car dealership or a dental office. Those are difficult to squeeze into a mailbox in the Cayman Islands. To be sure, it is possible to overstate expenses and so on, but when it comes to making revenue vaporize entirely, nothing compares to a Swiss banker.

Here’s the second problem. If you’re a tax lawyer, you probably already, er, realized it. Our tax system only taxes investment assets when they’re sold or otherwise disposed of. With some exceptions, it only taxes foreign earnings when those earnings are distributed to a U.S. taxpayer. So using the asset as security for a loan doesn’t show up on the tax return. Leaving the firm’s profits in the firm, and allowing the stock value to increase in value, doesn’t show up on the tax return. Licensing your name overseas, and leaving the resulting revenues in non-U.S. companies, often doesn’t show up on the tax return. Remember how our President probably didn’t pay any income tax for, like, 12 years?

Look, I found this an amazing paper that taught me many things I didn’t know before. I’m glad the authors wrote it. I just don’t think it means what they think it means. Am I wrong?

--

--

Brian Galle
Whatever Source Derived

Full-time academic (tax, nonprofits, behavioral economics, and whatnot) @GeorgetownLaw. Occasional lawyer. Also could be arguing in my spare time.