Tyler Cowen has written an unusual book. From the title, one expects a book that addresses the current economic crisis and prescribes a remedy for it. Instead, Cowen concentrates on the traits and virtues of autistic people. If Cowen wants to write on this topic, why does he lead the reader to think he will discuss something else?
But Cowen's principal topic is not as irrelevant to the economy as one might first imagine. He contends that autistics often possess the ability to classify things in unusual and illuminating ways. They are especially good at seeing details ignored by others; indeed, Cowen points out, suggestions abound that many great thinkers have displayed autistic traits:
Charles Darwin, Gregor Mendel, Thomas Edison, Nikola Tesla, Albert Einstein, Isaac Newton, Samuel Johnson, Vincent van Gogh, Thomas Jefferson, Bertrand Russell, Jonathan Swift, Alan Turing, Paul Dirac, Glenn Gould, Steven Spielberg, and Bill Gates, among many others, are all on the rather lengthy list of famous figures who have been identified as possibly autistic or Asperger's [a related condition]. (pp. 25–6; A roughly similar, but not identical, list appears on pp.166–7)
Cowen applies some of the wisdom he gleans from the autistics to contemporary economic problems; and it is on one particular aspect of his treatment that I wish to concentrate. As we shall see, in doing so, I run the risk of being suspected of autism, at least by Cowen's definition; I too am concentrating on a few details.
Cowen makes a strong case for viewing autistics with sympathy, and his views on autism are the product of long study. But in his characterization of autism he seems to me to have taken a dubious turn. Cowen suggests that the personality problems often associated with autism do not form part of its essence. Autistics need not be isolated and withdrawn, able to communicate with others only with great difficulty. Indeed, some autistics may deal with others so successfully that their autism goes undetected.
Perhaps the personality traits most people associate with autism stem from a biased sample.
Medical professionals control the familiar definitions of autism and they meet those people or parents who come to them for help. It's no surprise that these people and their doctors are focused on life problems. At the same time, many of the autistics with relatively high social status don't want to affiliate with the concept, or, more frequently, they are genuinely unaware that they might qualify as autistic in some manner. (p. 23)
But this argument raises a problem. Once one drops personality disorders from the definition of autism, what is left? Evidently it is the ability to classify and the illuminating concentration on detail. Certainly these are highly desirable characteristics; and if they are the essence of autism, then Cowen is right that we have much to learn from those in this condition. But, taken this way, anything distinctive about autism has dropped out of the picture. All that Cowen would be saying is that to classify things in a creative way is a good trait.
An analogy will clarify my objection. David Shapiro has argued in his excellent Neurotic Styles that paranoids can often see details that "normal" people miss. It would not be a good idea to conclude from this fact that paranoia should be defined to encompass only excellent perception of details, and to omit the behavior generally viewed as disordered.
There is an obvious response to this objection. Cowen might say that it is not any instance of creative classification or attention to detail that marks someone as autistic. Rather, there is a distinctive intellectual style that characterizes autistics. To show this, he needs first to delimit a sample of autistics; but in his discussion, he appears to do so by looking at autistics in the ordinary-language sense.
Cowen's autistics do possess the odd personality traits in question; so it is not clear on what basis he deems these personality traits inessential to autism. He might respond that the distinctive intellectual type associated with autistics in the ordinary-language sense may exist without the personality disorders; and if it does, people with this intellectual style still count as autistic. But, though readers must judge for themselves, I cannot see that he has succeeded in specifying a distinct manner of intellection.
Cowen thinks we can learn from autistics, but he does not think we should place unlimited reliance on unusual systems of classification, however creative they may be. Here he takes a lesson from Buddhism, which stresses the fleeting character of our concepts. "Most fundamentally, Buddhist philosophy is suspicious of complex forms of mental ordering and that is where Buddhism parts company with autism" (p. 94).
In the course of his discussion of autism, Cowen makes a contribution that readers of The Mises Review will find of great value. He has given us a penetrating and subtle criticism of the use of behavioral economics to support government intervention in the economy.
Some behavioral economists, e.g., Richard Thaler and Cass Sunstein, argue in the following way: People often act irrationally. They let emotion overcome reason or commit straightforward mistakes in logic. Such "choices" do not reflect such people's true preferences. The state may justifiably act to nudge people toward what they really want. Such "libertarian paternalism" does not interfere with freedom in any objectionable way.
Cowen raises two criticisms of this argument. First, the "mistake" may not be an error but rather reflect someone's idiosyncratic classification. If the state "nudges" people out of a choice based on such a classification, it will add error rather than remove it. Further, even if someone has in fact made a mistake, the error may be the inevitable product of some scheme of classification that on the whole works well.
Behavioral economists sometimes write of human beings as subject to "framing effects," meaning that the presentation of the alternatives influences our choices.… Usually the presumption is that framing effects are to be avoided. To be sure, many framing effects are irrational but framing effects help put the guts into our lives. We spend time and energy framing things in the right way so that we can enjoy them more or learn more from them (p. 6).
Once more, heavy-handed interference will disrupt things, not make them better.
Cowen also suggests that a similar mechanism serves to block a related criticism of the free market. John Kenneth Galbraith, echoed by many others, claimed that consumers on the free market are not at all sovereign, as Mises contended. Quite the contrary, mass advertising campaigns ensure that consumers will choose what businessmen decide to produce. Cowen notes that consumers can shield themselves from the effects of such advertising by availing themselves of specialized markets.
In this connection, the Internet enables people to find such markets much more efficiently than was previously possible. "The competitive pressures from free fun on the web affect marketing prospects for virtually all goods and services, again because there is competition. If you're trying to addict me to drinking expensive bottles of red wine, such a habit now has some especially cheap competition, again as can be found on the web" (p. 142). Cowen's discussion of the advertising argument is one of the best since Hayek's classic, "The Non Sequitur of the 'Dependence Effect'."
Cowen invokes Hayek to argue that, instead of using the state to "correct" people's behavior, thus enabling them to realize their true preferences, it is far better to confine law to a strict system of general rules:
Most of all Hayek is skeptical about the ability of human beings to plan all outcomes in advance by using their reason. Hayek argued that a rich and largely unplanned order can blossom when society is governed by a relatively small set of abstract rules.… You don't have to share Hayek's conservative and libertarian version of this blend to find this an appealing vision. (p. 201)
Cowen, in his typically understated way, has in these instances ably defended a view consistent with libertarianism. I suspect that he has not deviated from his Austrian and libertarian roots as much as is sometimes imagined, though he is by no means a libertarian of the strict observance. But though his arguments just canvassed have great value, I sense in them a danger. If pushed too far, his line of thought could lead to an undue subjectivism, in which people's perceptions and classifications, rather than what actually occurs, would be the sole issues of importance.
I do not say that Cowen has succumbed to this danger, but his discussion of Robert Nozick's "experience machine" brings out what I have in mind. Nozick imagines a machine that will enable someone plugged into it to have any experiences he likes. If you were able to use the machine, would you do so? Nozick suggests that people would often refuse. Does this not show, against at least some versions of hedonism, that there is more to value than experiencing things "from the inside"?
Cowen responds that in some circumstances, people would choose to enter the experience machine. No doubt he is right, but to stress this fact is to ignore the thrust of Nozick's argument. Nozick's point is not that we would never choose to enter the machine but that we sometimes wouldn't, which suffices to show that value cannot be reduced to subjective experience.
But this is not the key indication of what I take to be Cowen's unduly subjectivist turn. This emerges rather from his suggestion that Nozick's argument rests on a questionable appeal to an "authentic" world, bereft of human ideas. Cowen suggests that we have no access to a world-in-itself; our perception is inevitably structured by our concepts and theories.
I don't think the so-called real world is very "authentic" at all. No one who refuses to plug into the machine is in fact choosing or defending pure authenticity. (p. 144)
But Nozick's argument does not depend on a direct realist theory of perception. Quite the contrary, it need only appeal to our ordinary notion of the actual world. So long as there is a distinction between doing something and imagining it, and we think this distinction important, Nozick's argument works. Cowen, by neglecting this point, is in danger of having constructed a defense of individuality that severs us from the world.
Nevertheless, he has given us a book of great value, and in it his remarkable range of reading is much in evidence.
I noted a few mistakes: the birth of the concept homo ludens did not come "a few decades ago, when my social science colleagues investigated our game-playing nature" (p. 13) — indeed Johan Huizinga's book entitled Homo Ludens was published so long ago as 1938; Harold Innis was an older colleague of Marshall McLuhan, and his notion of "the bias of communication" came before McLuhan's "the medium is the message," not later (p. 65); finally, though Armen Alchian wrote much less than some economists, e.g., Samuelson and Arrow, it is not accurate to say that he "published a relatively small number of articles in his career" (p. 49) — Liberty Fund has issued two thick volumes of his papers.
 Basic Books, 1999 . See chapter 3, " Paranoid Style," pp. 54ff.
 Cowen cites a characteristically wide number of publications on Buddhism, but I'm surprised he makes no mention of the most elaborate discussion of Buddhism by a noted economist, Serge-Christophe Kolm, Le bonheur-liberté: bouddhisme profond et modernité, Paris, 1982.
 See my review of Thaler and Sunstein, Nudge, in The Mises Review, Summer 2008.