Tax policy colloquium: Gale and Thorpe, Rethinking the Corporate Income Tax: The Role of Rent Sharing, part 2

 My prior post set the stage for responding to the Gale-Thorpe paper's finding that, under plausible empirical surmises, firms' sharing of their excess returns with particular employees may do little if anything, to reduce one's estimate of the progressivity of corporate income taxation. Here I turn to some of the main issues that the paper's analysis raises.

1) Are highly profitable US companies' excess returns rents, or merely quasi-rents, and does it matter?

In economic theory, rents derived from ownership or control over a limited asset or resource, attained without any expenditure or effort by the resource-holder, and that exceed any opportunity cost of reaping them, can efficiently be taxed. Moreover, the incidence of the tax can't be shifted by the resource-holder to anyone else.

But there are two main problems with calling excess returns rents in practice. First, given the role of luck, observed ex post high returns don't prove the existence of a high ex ante expected return. One would not, for example, try to measure the return on investment (ROI) from New York State lottery tickets by looking just at the winners' ROI.

This leads to the quasi-rents issue that the paper indeed notes. For example, even if successful patents earn a lot, what about all the failed efforts to create valuable ones? Might taxing the resulting excess returns create efficiency and incidence issues after all?

Second, I'd argue, as did my old friend David Bradford, that excess returns are definitionally returns to labor in a certain sense. This need not mean strenuous effort, but at least the exercise of choice.

Suppose I find $10,000 lying on the sidewalk. Not being an economist (as per the old joke), I believe that it is really there, and hence I pick it up. Clearly, I have been lucky. But in addition, I walked by the spot, looked down, and stooped to clutch in my excited (?) fingers. In that sense, I think it's useful to say that this was a return to my labor (as well as to luck).

This does not mean that I "deserve" the $10,000 as the fruit of my effort, earned by the sweat of my brow, etcetera. As I'll discuss further below, there's a huge ideological component, at least in US ideology,* to saying that, because something is labor income, it therefore is personally deserved. 

*Well, not just US ideology, if you think, say, of Locke's Second Treatise.

The labor / choice component of deriving excess returns, like the ex ante vs. ex post issue in measuring excess returns, raises the possibility - excluded by definition under the concept of pure economic rents - that companies' excess returns actually can't be taxed completely efficiently and without the possibility of incidence-shifting.

But this seemingly big theoretical point may actually make less of a practical difference in evaluating how to tax companies' ex post excess returns than one might initially have thought. Here are several distinct points that may all lie in the direction of saying (to overstate it a hair): So what?

a) The margins that are being taxed here - e.g., pertaining to seeking to hit an ex post home run via risky investment with a high upside, along with supplying labor to the same end, may not be very elastic. Hence, the behavioral responses that result in inefficiency and incidence-shifting may be fairly modest.

b) A given country that is pondering how to tax excess returns has an externality reason for doing so. The deterrence of future highly profitable investment may have spillover effects that are global, not just local, whereas one gets the revenue. Now, perhaps one should in general be cautious about urging countries to exploit positive spillovers in this way. But, given highly profitable companies' immense global economic and political power, I suspect that the end result may not be too bad. Indeed, they may have sufficient clout for one to suspect that on balance they will be undertaxed, not overtaxed.

c) The downside of reduced risky productive effort may take a while to manifest, causing it to be borne more by future than present generations. Suppose that one thinks that  our generational policy is not in fact unduly tilted towards us relative to them (as, for example, Neil Buchanan argues here). This might then significantly mitigate one's weighing of the future harm relative to the current benefit.

2) How should we define progressivity?

Here I raise a question, not about the results of applying a given metric (e.g., how fast do tax burdens need to rise with "income" - or whatever - in order for it to count as "progressive," but rather how we should define the metric itself.

a) ECI vs. other metrics - Expanded cash income, or ECI, is the metric used by the Brookings models that underlie the analysis in the Gale-Thorpe paper. As I noted in the prior post, ECI is an imperfect standard for economic income. But economic income is itself a rightly controversial metric. Indeed, it lies oddly and perhaps uneasily in the middle, between the metrics suggested by a snapshot approach on the one side of the spectrum, and a lifetime (or even multigenerational) approach at the other side.

From a snapshot perspective, the right standard (or at least one closer to being right) is wealth. This might perhaps be expanded to include expected future earnings - at least for high-wage people who can borrow against the expectations, like working, and anticipate earning a lot more than their support needs. If one favored wealth as one's perspective, economic income arguably would undervalue the benefits enjoyed by wealth-holders, by reason of including only the current period's return to the wealth.

From a lifetime perspective, the proper metric is lifetime income + transfers received (such as through bequests). This is the classic consumption tax baseline, reflecting the view that normal returns to saving merely represent how one has chosen between present and future consumption. This perspective is used for example, in generational accounting studies as well as recent work on intra-generational accounting. From this standpoint, economic income is mistaken in including the normal return, and also confuses lifecycle effects with permanent effects. (E.g., if I have saved adequately for retirement, then in the year when I retire my income declines, due to the loss of the current year labor income component, but in fact I am no less well-off than previously).

b) Labor vs. capital - There has been a lot of talk in recent years about the tax system's unduly favoring capital relative to labor. But one might also characterize the same set of concerns, in very different language, as involving high-earners (whether from self-earned lifetime income or those from bequests) relative to low-earners.

Because of our imprisonment by misleading rhetoric, it makes a great deal of sense for people on the left, who favor a more progressive system, to say that we are unduly favoring capital relative to labor. And it makes equal sense for people on the right, who favor a less progressive system, to say it's all labor, with capital (properly defined in terms of the risk-free return that it can earn) being just a thing rather than basis for applying the vertical metric.

Once again, in American (but not just American) ideology there's all this stuff honoring labor income as deserved, earned by the sweat of one's brow, etcetera. It's a version of the "myth of ownership" that Murphy and Nagel attacked in the book bearing that name.

But high-wage needn't mean that you deserve it. You are still lucky, may be rewarded for doing bad things, may be reaping the advantages of privilege, etcetera.

To me, high-wage is a descriptive term that is entirely neutral in terms of its relationship to desert. But given how people view it, I would anticipate continued debates about whether it's actually about labor versus capital. (And I also don't want to wholly discount here the relevance of "capital" as such - a term that I may be writing about after my somewhat arduous current semester is done.) But one should firmly in one's mind distinguish between the rhetorical issues and the substantive ones.

3) Other topics

At this point in the annual cycle I am keeping too many balls in the air to devote time and space here to the many other interesting issues raised by the Gale-Thorpe paper. But here is a list of some of them:

--Rent-sharing is a very interesting topic, with multiple implications for how to think about taxation, labor markets, efficiency, etcetera.

--How should the shareholder-level tax figure in one's analysis of corporate income taxation and xcess returns?

--To what extent does the existing US federal income tax fall on excess returns as compared to normal returns? How should these two components be defined and measured?  How different would tax incidence be, say, under a VAT or DBCFT?

--Since rent-sharing incurs in part through self-dealing by corporate managers, what are the links between the issues raised here or in the paper, and issues of corporate governance?