Ronald Coase and Salvation from Anthropological Economics

Economics as a subject has never enjoyed healthier times — a universe of Freakonomics clones is appearing and the subject is galloping along in popularity as an undergraduate major. Yet, these are also the most worrisome times ever for the subject, because it is in danger of losing sight of the big mission — building conceptual models of the economy at large — that makes it so valuable to the rest of us. I’ll explain why it is a problem, and how the coming of The Chosen One, a descendant of Ronald Coase, can get economics back on track addressing the important problems of our century. Let’s start with a little family tree.

Economists Social Network

(portraits by Yurij Alexander)

Let’s recall the main storyline. Adam Smith got us thinking about the invisible hand, David Ricardo systematized the game (and along the way, said a few things about comparative advantage that became everybody’s mantra on outsourcing). Keynes got neoclassical economics to maturity and said we ought to attempt to control the economy. Friedman disagreed with him (“why” doesn’t actually matter for this article), but kept to the same level of abstraction and conceptual modeling. The minor characters, as far as the logic of the narrative goes, don’t really matter for the purposes of this piece (I don’t want libertarians jumping down my throat for apparently dismissing Von Mises — I am also setting aside my favorite, Schumpeter).

The key players for this story are Ronald Coase, since a Neo who will take forward the line of thinking he started is our only hope for salvation. The other key player is Steven Levitt, who came in from left field through an obscure line of descent involving behavioral economics (Kahnemann, Simon and bounded rationality, and so forth — see my piece, the Broken Brain books). He truly is the Rogue Economist the subtitle of Freakonomics labels him. And not in a good way.

Levitt and Freakonomics Reconsidered

My argument in this piece isn’t that Levitt is the bad guy. It is that he is distracting the study of economics by bringing in an unrelated agenda that belongs in psychology and sociology, and a down-in-the-weeds methodological bias that is basically noise as far as the truly interesting and important questions (to me at least) of economics are concerned. The method is anthropological, focused on narrative and thick description, with a glossy overcoat of (occasionally interesting) mathematics. The opening anecdote (about Levitt in an interview at Harvard) of Freakonomics, which the authors frame as evidence of Levitt’s genius, is actually a telling symptom of this malaise:

Disquietingly, one of the senior fellows said to Levitt: “I am having a hard time seeing the unifying theme of your work, could you explain it?” Levitt was stymied, he had no idea what his unifying theme was, or if he even had one. [… brief description of the people in the room attempting to find one]…until the philosopher Robert Nozick interrupted, “…why does he need to have a unifying theme? Maybe he’s going to be one of those people who’s so talented, he doesn’t need one.”

There you have it. Admiration for raw, undirected cleverness winning over a questioning of fundamental importance. I wish there’d been someone in the room like Fight Club’s Tyler Durden (played by Brad Pitt), who responds to a smart remark by Edward Norton’s narrator character with, “Clever. How’s that working out for you, being clever?”

Again, not that I think Levitt’s work is unimportant. It’s actually quite interesting — as psycho-sociology done with a mathematical-anthropological methodology. But it is irrelevant as far as progress in “real” economics goes. I strongly suspect the sex-appeal of the field created by Levitt (and the me-too’s that followed) will attract too many students to economics for the wrong reasons. With perhaps damaging consequences. But maybe I am being pessimistic — maybe some of the brighter kids will think for themselves and return to the fundamental questions of economics.

Coase and the Big Question of 21st Century Economics

Enough about Levitt. Let’s talk about Coase and what he did. Coase started a line of thinking in economics that viewed the firm as the fundamental building block of the economy. Nothing in either Keynsenian or Friedmannian economics really implies or underlines the role of the firm — those models might as well be about an anarchy of individuals transacting in pair-wise Von Neumann games.

Ronald Coase

(drawing by Yurij Alexander)

Coase was bold enough to ask, way back in the 30s, the important question, why does the firm exist at all, let alone in a variety of different sizes? He also asked similar questions about (real) markets and the role of Law. To an extent, he did to economics what (or so I am told, by knowledgeable people) John Rawls did to the utilitarian/empiricist philosophies of Bentham and John Stuart Mill — modeling the roles of institutions properly.

His answer, roughly, was in terms of transaction costs associated with information flows in market interactions, which most other economists ignored. These come in 3 varieties: search costs, negotiation costs and monitoring costs for a unit of work to be done in the economy, by a provider, for a buyer. The key conclusion is that the size of the firm is determined by the levels of these “overhead” costs. The lower those are, the smaller the optimal size of a firm for given conditions. Take them to zero, and you get a firm-less economy of free agents.

Why is this important? Simply because the biggest impact of the Internet on the nature of the economy is to lower these transaction costs, leading to the broad, sweeping conclusion that the average size of firms will likely shrink rapidly, both by way of shrinkage of existing enterprises as they reduce themselves to infrastructure hubs with surrounding ecosystems, and directly, by way of growth in SMBs and freelance work in the economy (though a few mega firms, larger than any in history, may still emerge, for other reasons, the average will shift downwards). Data supports this conclusion — enterprises are shrinking, SMBs are growing.

In everything I’ve read, this is the only coherent conjecture I’ve seen about the nature of the economic impact of the Internet.

And that’s why we need Neo. A modern-day intellectual descendant of Coase who makes transaction-cost economics as sexy as Levitt has made behavioral economics. There is a LOT of work to be done. We need good models of how firms shrink and grow. We need models of the impact of on-demand labor, sourced from online marketplaces like elance and odesk, on the cost structure of large enterprises. We need a fundamental theory of crowdsourcing. We need Ricardo’s comparative advantage ideas folded into a broader model of the costs of doing work in various organizational structures.

This agenda isn’t going to follow itself while Levitt’s fans are busy studying who steals bagels and how Chicago school teachers cheat and why drug dealers live with their mothers.


In the interests of full disclosure, my only formal training in economics is a sophomore course that used Samuelson’s text (the professor teaching the other section — we had a huge, 400-strong undergraduate cohort — used Friedman material). The rest of my understanding is at Wikipedia level (barring Coase, whose work I am currently studying in the original).

If I am getting something egregiously wrong, please correct me.

Get Ribbonfarm in your inbox

Get new post updates by email

New post updates are sent out once a week

About Venkatesh Rao

Venkat is the founder and editor-in-chief of ribbonfarm. Follow him on Twitter


  1. I’ve been doing some disorganized musings along these lines – haven’t really read Coase just yet.

    At work, I was struck by the realization that we seemed to do splendidly when we outsource packets of work – say, some GUI design – there’s a nice contract, work starts, work gets done, money gets paid.

    If I had to do the same thing internally, there’d be no end of hassles. First I’d have to request an internal group to do it, then they’d tell me they have no resources, no budget, no bench strength. Then I’d have to convince them to convince their management to get some headcount. Or I’d have to give them some headcount (but it’s only a 6 month project – what happens after that? I can’t afford to give up a whole headcount for that!)

    And what happens when the Grim Reaper stalks forth after a bad quarter to do some ritual bloodletting – how does this headcount get affected? What’s the optimum strength of the GUI design team?

    Given the way finance is handled here: beginning of the year, you think you’ll need an absolute minimum of X. You ask for 2X, because you know it’ll be cut down by half. So you actually get X/2 if you’re lucky. Then you scrimp and save and at the end of the last quarter, you figure you still have $100k left and indulge in an orgy of spending. Why? Otherwise it’ll disappear, and if you return the budget, you’ll get less next time.

    It’s richly ironical that big companies in a capitalist economy are run in exactly the same way as planned economies. And they inherit the same problems – the economic calculation problem, for one. There is no good currency to facilitate trade between groups. and money cannot be retained or saved but disappears at the end of the budget cycle, and headcount – which is retained – is too coarse-grained to facilitate numerous and frequent transactions.

    I guess my company is on the wrong side of Coase’s theorem, which fits in well with the thesis that large companies will – or should – shrink.

    As for Freakonomics – interesting stories, but I got really irritated with the cheerleading. I would have liked Pulp Fiction‘s Mr. Wolf to have been present, with his classic response to some fawning on the part of Jimmy and Jules: “Let’s not felicitate each other quite yet.” Well, he used a different word, but there are children present.

  2. Hmm… and I thought only my company worked that way; fighting over headcounts in a grim battle of attrition quarter by quarter :)

    But you are totally right about the irony. There are of course internal cost transfer arrangements in large companies, to maintain an actual economy among organizations, but there is still far too much friction compared to the external economy.

    The really weird thing is that Coase’s model is almost being turned on its head in some cases of pure information work. The external economy is actually vastly more transparent on the 3 costs (discovery, negotiation, monitoring). Dan Miller, who just wrote the book ‘No More Mondays’ coined a phrase ‘eaglepreneur’ for an entrepreneur who outsources everything (different from a free agent or self-employed…)


  3. It seems to me many economist lately are really focused on analyzing data. Nothing wrong with that but instead of really a focus on economic thought they are just using the significant ability to analyze huge stores of data. So then they just go off and analyze data and since they have a PhD in economics it is called economics. Much of what Levitt does I think it interesting but I am not really convinced it is economics.

  4. I am one of the Heterodox economists! I really appreciate the work done by Coase. In fact he contributed to the improvement of the fundamental principal of welfare economics which is based on market completeness assumption(complete contract) which seems to be unrealistic and utopian!
    Coase demonstrated that even when there is no complete information, if transactions are costless, still we can achieve a Pareto optimum.