The Nirvana Fallacy is Not the “Fiction” Fallacy

Adam Mossoff —  28 August 2012

In a response to my essay, The Trespass Fallacy in Patent Law, in which I explain why patent scholars like Michael Meurer, James Bessen, T.J. Chiang and others are committing the nirvana fallacy in their critiques of the patent system, my colleague, T.J. Chiang writes at PrawfsBlawg:

The Nirvana fallacy, at least as I understand it, is to compare an imperfect existing arrangement (such as the existing patent system) to a hypothetical idealized system. But the people comparing the patent system to real property—and I count myself among them—are not comparing it to an idealized fictional system, whether conceptualized as land boundaries or as estate boundaries. We are saying that, based on our everyday experiences, the real property system seems to work reasonably well because we don’t feel too uncertain about our real property rights and don’t get into too many disputes with our neighbors. This is admittedly a loose intuition, but it is not an idealization in the sense of using a fictional baseline. It is the same as saying that the patent system seems to work reasonably well because we see a lot of new technology in our everyday experience.

I would like to make two quick points in response to T.J.’s attempt at wiggling out from serving as one of the examples I identify in my essay as a patent scholar who uses trespass doctrine in a way that reflects the nirvana fallacy.

First, what T.J. describes as what he is doing — comparing an actual institutional system to a “loose intuition” about another institutional system — is exactly what Harold Demsetz identified as the nirvana fallacy (when he roughly coined the term in 1969).  When economists or legal scholars commit the nirvana fallacy, they always justify their idealized counterfactual standard by appeal to some intuition or gestalt sense of the world; in fact, Demsetz’s example of the nirvana fallacy is when economists have a loose intuition that regulation always works perfectly to fix market failures.  These economists do this for the simple reason that they’re social scientists, and so they have to make their critiques seem practical.

It’s like the infamous statement by Pauline Kael in 1972 (quoting from memory): “I can’t believe Nixon won, because I don’t know anyone who voted for him.” Similarly, what patent scholars like T.J. are doing is saying: “I can’t believe that trespass isn’t clear and efficient, because I don’t know anyone who has been involved in a trespass lawsuit or I don’t hear of any serious trespass lawsuits.”  Economists or legal scholars always have some anecdotal evidence — either personal experiences or merely an impressionistic intuition about other people — to offer as support for their counterfactual by which they’re evaluating (and criticizing) the actual facts of the world. The question is whether such an idealized counterfactual is a valid empirical metric or not; of course, it is not.  To do this is exactly what Demsetz criticized as the nirvana fallacy.

Ultimately, no social scientist or legal scholar ever commits the “nirvana fallacy” as T.J. has defined it in his blog posting, and this leads to my second point.  The best way to test T.J.’s definition is to ask: Does anyone know a single lawyer, legal scholar or economist who has committed the “nirvana fallacy” as defined by T.J.?  What economist or lawyer appeals to a completely imaginary “fictional baseline” as the standard for evaluating a real-world institution?

The answer to this question is obvious.  In fact, when I posited this exact question to T.J. in an exchange we had before he made his blog posting, he could not answer it.  The reason why he couldn’t answer it is because no one says in legal scholarship or in economic scholarship: “I have a completely made-up, imaginary ‘fictionalized’ world to which I’m going to compare to a real-world institution or legal doctrine.”  This is certainly is not the meaning of the nirvana fallacy, and I’m fairly sure Demsetz would be surprised to learn that he identified a fallacy that according to T.J. has never been committed by a single economist or legal scholar. Ever.

In sum, what T.J. describes in his blog posting — using a “loose intuition” of an institution an empirical standard for critiquing the operation of another institution — is the nirvana fallacy. Philosophers may posit completely imaginary and fictionalized baselines — it’s what they call “other worlds” — but that is not what social scientists and legal scholars do.  Demsetz was not talking about philosophers when he identified the nirvana fallacy.  Rather, he was talking about exactly what T.J. admits he does in his blog posting (and which he has done in his scholarship).

3 responses to The Nirvana Fallacy is Not the “Fiction” Fallacy


    Adam, two quick points:

    First, I contest that “I could not answer” your question. In fact, I thought I answered it. Here is what I said:

    “In one sense you could say that anybody (including me) making the comparison between patents and real property is committing a Nirvana fallacy, because they are implicitly comparing (a) the actual patent system of today, and (b) a hypothetical nonexistent patent system that in their idealized imagination operates like the real property system. I would agree that is a Nirvana fallacy. But the implication of that logic would be to stop making comparisons between patents and real property altogether, at any level of abstraction. That, to my understanding, is not your argument.”

    Second, lets take Demsetz’s example of the nirvana fallacy when an economist implicitly posits that regulation always works perfectly to fix market failures. I agree that is a Nirvana fallacy, because I disagree that anybody has a “loose intuition” this is true as an empirical matter. Please name a single person who would defend the proposition, as an empirical assertion about the real world, that regulation always works perfectly to fix market failures. Nobody does (at least I have a loose intuition that nobody does). There is an important difference between a loose and objectively provable or falsifiable intuition about the real world and an idealized fiction, which I think you are missing.


    Great paper Adam. I entirely agree it’s important to be clear what exactly is being placed on both sides of the analogy. I’d like to take one facet of your response to T.J. above in a slightly different direction, however:

    “Similarly, what patent scholars like T.J. are doing is saying: ‘I can’t believe that trespass isn’t clear and efficient, because I don’t know anyone who has been involved in a trespass lawsuit or I don’t hear of any serious trespass lawsuits.'” I agree you can’t infer trespass doctrine’s *clarity* from anecdotal impressions of the frequency of lawsuits. But it might still be efficient, in the sense of using a near-enough-to-optimal amount of judicial and party resources to resolve the number of disputes that do arise. Along those lines, I think a better response to your paper might be that it’s not *as important* for property law to be clear about the more ethereal aspects of property ownership such as easements, restrictive covenants, future interests, or adverse possession because those issues hardly ever come up in ordinary life — typically only around sales, which are infrequent events. The practical effect of property law in the experience of most people is to give legal effect to visible markers of ownership, such as lawns, fences, houses, fields, etc. — in other words, to help define what is and is not infringing on the right to exclude from physical boundaries — and it performs that background task adequately well. The legal specification of the physical boundaries of a piece property, and the designation of current ownership of property, tend to correspond well enough with the visible markers, so that ordinary individuals are able to tell without hiring lawyers or engaging in litigation where they can walk, where they can erect signs, where they can plant crops, etc. Patents lack visible markers, however, and the claim language in a patent (so goes the argument) is not an analogous substitute, such that the basic ordinary task of a patent for your average business of setting out what they can and can’t do without hiring a lawyer or engaging in litigation is not achieved.

    This argument is of course based on intuitions on how frequently these issues arise for your average person, and perhaps my intuitions about how frequently basic questions of property ownership arise are off, or perhaps easements, future interests etc. are much more frequently an issue for average people than I am familiar with. MERS might be an example where interests in real property not only do not correspond with visible markers, but are also very difficult to determine from public records, so difficult in fact that the interest-holders have a hard time proving the validity of their interests — but MERS itself, of course, has been accused of being dysfunctional. Or perhaps, despite all the news coverage, your average company in an innovating market does not need to worry about patent infringement all that often, about as often as real property holders need to worry about easements. (Lots of companies *do* worry about patents, of course, but then people worry about shark attacks too, so that doesn’t by itself tell us how frequently the event occurs.)

    Laura Bennett Peterson 28 August 2012 at 10:30 pm

    Isn’t perfect competition the paradigmatic Nirvana fallacy at least as an idealized counterfactual standard? (I’d have to reread Demsetz to see just how he defined it.)