Evan Soltas
Mar 18, 2015

The Rent Hypothesis

Monopoly
"Rent" is three things: an idea in economic theory, a payment for housing, and a musical. This post is, maybe unfortunately, about the first of those things.

Rent, as an economic idea, refers to some gain that, loosely, you can think of as unearned. It is, more rigorously, a payment for a resource in excess of its opportunity cost, one that instead reflects market power.

For instance, profits that a monopolist earns in excess of the profit that would exist in a competitive market are rents. So are excess profits from the production of a patented invention, or the extra compensation that an executive might get because he's buddy-buddy with the board of directors.

Rents are not always bad. Protections for intellectual property make sense not in spite of the rents they generate, but rather because of them. A patent, after all, is a temporary monopoly. These monopoly rents give incentives for innovation and allow innovators to cover the costs of research and development. (Warning: This logic can be taken too far.)

Usually, however, there is something less than endearing about rents. That's because, without some countervailing market failure, as in the case of innovation, giving away unearned gains is bad for incentives and bad for the allocation of resources. An economy where the cost of Internet access (to choose a not-accidental example) is determined by how much can be squeezed out of you and me, rather than how much it actually costs to provide it to us, is inefficient.

There has been, for the last few years, a "big idea" floating around the economics conversation that these rents are growing -- that unearned gains are eating up an larger share of income. Let's call it "the rent hypothesis." It's an appealing idea from a certain perspective. It seems to explain a lot.

Why are corporate profits so high? Shouldn't they be competed away? Why does finance make so much money? Can it really cost that much to do what financial intermediaries do? Why is executive compensation so huge? Is it really possible that, as Bloomberg put it, Larry Ellison is "still a bargain" to Oracle at $100 million a year? And why, if profits are so high, is new-business formation and business investment so low? Shouldn't those profits attract new entrants and encourage existing businesses to expand?

That last point comes from Paul Krugman, who has done some of the brainstorming about rents here and here:

So what’s really different about America in the 21st century? The most significant answer, I’d suggest, is the growing importance of monopoly rents: profits that don’t represent returns on investment, but instead reflect the value of market dominance. Sometimes that dominance seems deserved, sometimes not; but, either way, the growing importance of rents is producing a new disconnect between profits and production.
You can also find Josh Bivens and Larry Mishel discussing rents here. To Bivens and Mishel, executives would still do what they do even if they were paid less -- Larry Ellison is not about to sail off in his America's Cup yacht -- which means that executives are, in fact, being paid above their opportunity cost. (The logic here is, if their pay falls below opportunity cost, they would go do that other, next-best project which determines the opportunity cost.)

They provide evidence that executives are uniquely well-paid in comparison to other top earners, which they interpret as a sign that their pay is about their position rather than their skills or value. Similarly, the rise in financial-sector profits and incomes seems to match up well with the moment those industries were deregulated. And in a recent paper, Thomas Piketty, Emmanuel Saez, and Stefanie Stantcheva show that executives are paid more for the same performance in low-tax countries than in high-tax countries. To them, that's evidence that low taxes on the highest incomes encourage rent-seeking behavior. (Click the pictures to enlarge.)


Yet all is not well for the rent hypothesis.

There's a big problem: We can't observe rents. Corporate profits that are rents don't look different than profits that are earned in free competition. It's not as if the dollar bills are marked "rent."

What we're left with is trying to find evidence that suggests some income is a rent. To return to our earlier example, the facts of high corporate profits, low cost of capital, and low business investment seems to require some non-competitive explanation. High profits and a low cost of capital shouts "business opportunity" -- so why isn't someone taking it?

Trying to spot rents is, in this sense, a bit like trying to spot a black hole. (Or what I imagine it would be like. I am not an astronomer.) You doesn't see a black hole in itself so much as you notice an absence, a void -- and the motion of stars towards it. Similarly, the idea is to spot rents by what is missing, by the presence of a contradiction where only rents can fill the gap.

This is a fun game to play, but it ends up being pretty unconvincing. It may seem like the facts fit together, as in Krugman's account. But there are lots of other ways to get to the same result that don't involve rents. And, to be sure, most of the analyses of rents admit this. Bivens and Mishel: "The evidence on rent-shifting behavior should be viewed not as conclusive, but as highly suggestive."

If what I've given so far is a largely favorable telling of the rent hypothesis, then, let's consider a more skeptical take. The skeptic says, "This 'rents' framework amounts to a way for these economists to denounce economic phenomena that they are ideologically predisposed to dislike. Corporate profits? It's rent! CEO pay? Rent as well! See how easy this is when economists can just point at things that feel unjust or unbalanced, but never really show that they are uncompetitive? Everything they've said so far amounts to a more intellectual version of 'doesn't this look strange?' That's not a real argument."

I should emphasize that the view of "the skeptic" is not what I actually think. Some people think it -- I'm channeling John Cochrane here. I think, though, it is a critique that, if harsh, has some truth to it. Most analyses of rents, particularly on the topic of corporate profits, do an inadequate job proving that the phenomena they describe are the results of uncompetitive behavior. I am not convinced. Nor do I think you should be.

"I think we would have more intelligent economic debates if all participants were asked to state what could happen that would cause them to modify their views," Larry Summers once told Ezra Klein in an interview last year. It is a great comment. What, then, might a more convincing analysis show in support of the rent hypothesis? Here are four ideas.

  1. It should explore the relationship between industry profitability, firm entry, and investment. If the emergence of large profits in an industry once, but no longer, led to a response of an increase in new-business formation or an increase in business investment, I would be much moved in favor of the rent hypothesis.
  2. It should explore the determinants of profitability at the firm level. If the distribution of profits among firms has shifted in ways that are consistent with the rent hypothesis, I would find that evidence to be strong. For example, if the age of a firm, after accounting for firm size, was more strongly associated with profits than it had been, that would strike me as important. The idea is that the apparent return to incumbency would have risen. Relatedly, if patent count, or some measure of intellectual-property holdings, has become a more powerful predictor of profits, that would also be strong evidence.
  3. It should account for the recent increase in aggregate corporate profits in ways that would be consistent with the rent hypothesis. For example, has the increase been mostly driven by large firms? By firms in low-entry industries? By firms in industries were size or age or entry has changed most significantly in recent years? By firms in industries where government regulation has been relatively high or low, or where it has increased or decreased?
  4. It should explore the evolution of competitiveness. How has the number of potential sellers of major product categories evolved at the local level, and is there a measurable effect on profitability? For example, if the average number of telecommunications providers competing in a zip code (or county or state) have declined and total profits in the telecommunications industry have risen, and that pattern was repeated across other industries, that to me would be compelling.
The common thread among these ideas is that they test the rent hypothesis systematically. It must be forced to make predictions about the economic world that either pan out or do not. I'm writing these out in hope that someone thinks the ideas are interesting and worth giving a try, and I might myself.

The rent hypothesis has a lot to say about our economic world. If it's right, it provides a unified explanation of many important shifts. It gets to why corporate profits are so high, why the financial sector has expanded so much, why executive pay has risen, and why entrepreneurship and investment are struggling. Time to move beyond speculation and toward analyses that have a real chance of telling us new facts about how our economy is changing.

Update (2/23/16): Jason Furman and Peter Orszag's recent work on the changing distribution of firm profitability -- in particular, the persistence of "super-normal" returns -- is an excellent study on (2). And see the excellent paper by José Azar, Martin Schmalz, and Isabel Tecu on the anticompetitive effects of common ownership. I saw Orszag and Schmalz present their research last week.