What's Wrong With the Endowment Effect?

Josh Wright —  6 October 2009

Gordon Smith asks the question in response to a 16 part post (with slides and pictures!) from John Carney offering up the explanation that the behavioral economists have overclaimed and that “the Endowment Effect may really be a response to the counterparty risk faced by early humans.”  Larry Ribstein chimes in with support for Carney and a general word of caution about behavioral finance.

Gordon doesn’t like John’s explanation because “the Endowment Effect appears in the lab even when counter-party risk is absent” and replacing irrational preferences with a just so story about evolutionary psychology involves “making assertions about human nature that are not susceptible to falsification.”  That’s all fair enough.

But what about the answer to the question?  What’s wrong with the endowment effect?

Ok, bait taken.

One possible answer is behavioral literature repeatedly has shown that some cognitive biases that are robust in the lab aren’t so robust in actual markets, where the repeat play and the profit motive at work.  So perhaps doing this in the laboratory with undergraduates performing tasks will show less than we think.

But that’s not the biggest problem.  The real problem with the endowment effect in particular is that, as of yet, EVEN the experimental findings are not robust.   And certainly not sufficient to stand as the intellectual foundation of an new paternalistic regulatory regime.  In other words, somebody has already taken on the behaviorists for finding irrationality where it doesn’t exist. Zeiler & Plott (or here), in my view, provide burden-shifting quality evidence that the endowment effect observed in the literature is better explained by experimental procedures than preferences.  Proponents of regulation based on the endowment effect, in my view, need not agree with my interpretation of these findings but they ought to respond to them if they want to be taken seriously.  Unfortunately, as I discuss here, out of the 255 articles in JLR discussing the endowment effect (210 also discuss regulation, btw), only 16 cite either Zeiler and Plott article.  I find that ratio discouraging for the discipline of behavioral law and economics generally.

It strikes me that all the discussion about plausible alternative stories for explaining the endowment effect are very interesting and important (including both rational and irrational accounts), but they miss the incredibly important point that we’ve currently got folks in areas ranging from finance to consumer protection to antitrust creating regulatory agencies, and proposing regulations that will surely have significant consequences (intended and otherwise) on the basis of a phenomenon for which we do not have compelling evidence.

And that is the problem with the endowment effect.

3 responses to What's Wrong With the Endowment Effect?

  1. 

    Just to add to something you mentioned, one of my professors, a generally heterodox friendly one, told me a couple of years ago that the problem with the behavioralists is that they put people in non-market situations with bias from experiment structure and use that to claim irrationalities and therefore policy interventions, yet when experimental economists do market experiments or computational economists use adaptive agent simulations analyze these allegedly irrational actors in a setting including markets and proper institutions, the irrationalities are frequently shown to be irrelevant to the overall performance and efficiency of the market. Basically, the behavioralists often ignore markets and institutions (besides for any experiment structure bias) and fail to show that their results are relevant in that context but are using their results to justify interventions into the real world which does have markets and institutions.

  2. 

    I think it is up to 26 cites now, and I think that the ratio (no mentions/mentions) is changing pretty quick. Maybe alot of that effect you observed was the result of lag.

  3. 

    Great post. My thought in reading the Carney blog and Gordon’s response was that Carney made an odd sort of error. He took the results of the endowment effect studies seriously but criticized their characterization–which he replaced with another characterization that Gordon (rightly) found not to change the basic result that people behave irrationally from the point of view of our standard economic models. Even if Carney were right that the behavior is evolutionarily adaptive to life on the Savannah 10,000 years ago, that tells you little about its relevance to today’s models of human behavior. So Gordon is left wondering what his re-characterization proves. Far more important is the point you make here: Whatever its cause, whatever its characterization, there needs to be sound evidence of the fact of the endowment effect before it matters for policy.