Quirky, Loveable, and Brilliant

Many Austrian economists only know of Bryan Caplan through his "Why I am Not an Austrian Economist" and more recently his blog posts on why the Austrian theory of the business cycle does not explain our current economic situation.  In other words, they know of Caplan mainly as a critic and a harsh and persistent one.  But there is a side of Bryan which those who know him appreciate --- he is a big kid, a loving father, and one of the happiest and most comfortable with who he is people you could ever met.

I first was made aware of Bryan's existence when he was just starting U of California and had written a defense of anarcho-capitalism.  He has a great and infectous enthusiasm for ideas and debate.  I have disagreed often with Bryan --- about Austrianism and Kirnzer in particular, and about political economy and democratic efficiency in particular.  And the root of our disagreement in both cases is strickingly similar and related to the concept of 'ignorance'.  I am still not persuaded by his criticisms, and I have often accussed Bryan of not possessing a 'subtle understanding' of the issues.  (To which Bryan responded once "Pete, you have to understand I want to be the least subtle economist ever.  Subtlety is overrated.")  In Bryan's understanding of politics, we get the government we demand, and thus the one we deserve.  But what we deserve is far from ideal, and that is because when it comes to politics while the machinery works more or less efficiently, the context of democratic voting enables us to pursue irrational beliefs rather than the calculating prudence one might find in the context of the stock-market.

But Austrians, and other social scientists and the general public, dismiss Bryan's criticisms too quickly at their own peril.  A sample of his insights into the political process and the relevance to contemporary events can be seen in this op-ed from the Washington Post.

Comments

[Josh H]

Bryan's book is pretty persuasive. I'd be interested to see a critique of his op-ed column. Where do you think he's going wrong?

[G]

I do disagree with Caplan's assertion that voters are irrational. I don't think they are irrational at all - they are just operating on limited data. In light of Hayek's knowledge problem and the burden democracy places on the polity, I believe it is quite clearly the system that is irrational.

The system asks the impossible: That each individual vote with the specialized knowledge which is dispersed throughout society and is utterly unknowable to him. When given limited data and asked to make a choice, one has no choice but to base the choice on that data. The fact that voters do not make good choices when voting is not indicative of irrationality when they cannot vote well no matter how hard they may try.

[Bogdan Enache]

I strongly believe that Caplan has set up a straw man for himself and he keeps boxing him ever since he set it up.

Traditional public choice theory started by applying the idea that individuals have a rational (in the instrumental, humean/economic sens of the word) behaviour in the marketplace as well as in politics, law and so on. Now Caplan either wants to dismisses this hypothesis and subsequent public choice - and for that matter economic - analysis of politics all together or he wants to propose an ad-hoc definition of rationality in politics that is non longer instrumental or descriptive but is normative and takes the results of instrumental rational behaviour in the marketplace to be the normative rationality criterion for political behaviour.

In the first case, Caplan basically says that voter behaviour can no longer be understood through the lens of rational economic behaviour, i.e. fitting scarce means to maximize ends. They, the voters, are irrational. Note : he is not saying that rational behaviour in politics leads to different results than in the marketplace, like all public choice theorists before him or as I will argue Austrians would say; for him voters are just irrational. If this is so, economics no longer has anything to say about this, enter psychology.

But if Caplan wants in fact to set-up a normative theory of rationality in politics, then what does irrational means for him in this context ? Well, it appears that it simply means that voting leads to different results than market exchanges ! (just like "sub optimal" market performances like in a business cycle is elegantly and promptly relegated by him to psychology and will probably be labeled a case of market irrationality...some animal spirits o f some sort...). But how does this refutes the hypothesis of rational economic behaviour in voting in the first place ? It doesn't because he completely bypass it : traditional public choice started with the hypotheses of rational behaviour of voters as well as market agents and then explained a sub-optimal outcome through a different incentive structure, a different knowledge generating structure, and so on; but Caplan's argument goes the other way around, he basically says that the results of voting behaviour are different, i.e. material welfare decreasing, then exchanges in the(well functionning) marketplace, ergo the former are irrational while the later are rational.

Anyway, if, according to Caplan, voters are irrational (just like economic agents when there is a business crisis) and only psychology illumitates this sort of behaviours, why do economists waist their time to explain this phenomena and don't let the psychologists do the job properly ? Is this rational or irrational ?

[Steve Miller]

It's called rational irrationality, people. He is not, in the economic/procedural/Misesian sense, saying that voters are irrational. He's saying they hold "irrational" beliefs (i.e. biased beliefs) only in the conventional sense. But people's bias follows a rational pattern -- they tend to hold biased/irrational beliefs when the cost of holding those beliefs is low. It's a twist on rational ignorance is all. He argues that bias, not ignorance is the real problem. I've heard Pete argue that it's both, and I tend to agree. Or rather, Pete or I, when arguing this point with Bryan, have a different view of what "ignorance" is, or can be: 1) Knowing what you don't know, 2) Not knowing what you don't know ("sheer" ignorance), and 3) Where what you know isn't so.

Most economists, Bryan included, think only about category 1 when they hear the word "ignorance." Austrians think about 1, 2, and sometimes 3. Bryan doesn't acknowledge sheer ignorance, but he is all about category 3. He calls it bias, and argues that in the conventional sense that bias is "irrational," but in the economic sense it is actually rational to hold low-cost biased beliefs.

His claim is that people aren't merely ignorant. They are flat-out biased. And if you don't believe people hold biased beliefs about economics, to take one important category of belief, then he has a mountain of evidence to prove you wrong. Conventional notions of ignorance don't explain the general public's economic beliefs. You can argue that Austrian notions do, and in fact Bryan has a (prize-winning) paper with Ed Stringham that points to Mises and Bastiat; they made the Caplanian argument long before he did (without the empirical evidence to back it up).

Finally, to respond to Pete's post: I came to GMU in 2002 to study Austrian Economics. I knew Bryan almost entirely as an anti-Austrian, and I have to confess I was a bit disappointed when Pete assigned me to be Bryan's research/teaching assistant. How could I work with an anti-Austrian on Austrian Economics? Well, I was wrong to be disappointed. As it turns out the anti-Austrian stuff is a tiny facet of his work, whereas the bulk of his work is arguably Hayekian (I would say it's actually Misesian).

The side effect is that I now would fail pretty much any "Austrian Purity Test" you could come up with. It's worth it, though, because I'm so much better off for having worked with him. Everything Pete says about him is true, he is so genuinely content and happy that he makes you feel more content about your life. Yet he has spurred me to be more ambitious than I ever imagined. My impression is that he has thoroughly charmed his colleagues at GMU, and I know that other former students who've worked with him (Ed, Scott, Geoff, etc.) would agree with Pete's sentiments.

And to say Bryan is quirky is a huge understatement. He's also the nerdiest person I've ever known, and that's not at all an insult. I still make fun of him when he puts on sunscreen and insect repellant to take a 15-minute walk around GMU's campus. We won't even talk about the pillow he carries around.

[Brian Pitt]

I agree with Steve Miller about Bryan Caplan. While I do not know him personally, I certainly believe that the man makes Misean arguments (Btw, I assume that Miller is talking about the 2005 paper with Edward Stringham, which is excellent to say the least.)

I always find it strange that economists do not accept that the political world is NOT analogous to the market world. But, very shrewd economists like Bryan Caplan do. However, what Miller did not explicitly mention is that Caplan is not dismissing "rationality" in the economic sense of the word. Caplan is simply noting that the rationality assumption may need to be modified outside of market contextts. That is to say, in political contexts, people "trade off" their self-interest for their political biases/beliefs. And all of that leads to a theory of "rational irrationality."

[G]

Steve, I hear what you are saying. It just seems a bit harsh for Bryan to criticize voters over what, in my opinion, is a totally systemic flaw in democracy. I just can't see the system working "as advertised" at all, and it seems wrong to place the blame on anything but the system itself.

Although I've not read his book in entirety, so I may be making a straw-man here. I do think his idea is spot-on, and I can spot many examples of "rational irrationality" in people I know personally.

[Bogdan Enache]

If Caplan upholds the rational behaviour hyphotesis in analysing voting (and I know that he often argues by doing so) then what is with all the talk about irrationality from an economic/positive point of view ? In other words, what does "rational irrationality" really means ? Is it just a case of rational ignorance or does it actually want to denote informed, knowledgeable preference for bad economic policies ?

[Brian Pitt]

Bogdan,

I believe that rational irrationality simply "capsizes" basic economic theory by maintaining that otherwise rational actors tend toward irrationality (i.e., beliefs and biases) on political matters.

The best way to conceptualize this is viewing actors as trading off material wealth and political beliefs. Think of an "irrationality" demand curve; where more of one means less of the other. What you will find is the quantity of irrationality falling where its price in material wealth terms rises. Caplan and Stringham (2005) give a great example of this by noting that "unless one is pursuing a career in biology, rejecting evolution is unlikely to affect one's bank account."

[matthew mueller]

Here is yet another great post. I found Steven Miller's comments particularly instructive. However, I wonder if whether it is correct to conflate bias and irrational behavior, and then argue that bias can lead to predictable outcomes because the cost of holding such views is not costly.

Some thoughts. First, Douglass North has a great chapter on the role institutions play in enabling people to exercise their beliefs at little cost to themselves (Institutions, Institutional Change, and Economic Performance, chapter 10). So I think democratic processes and the institutional structure that supports it can say a great deal about how individual preferences can find expression at little cost.

However, the existence of bias, in my view, does not automatically translate into 'rational irrationality'. First of all, the very notion of 'bias' makes sense only if we can speak of it in reference to 'unbias' --- that is, the idea that the possession of some sort of objectivity or truth (unbias) can be attained. Understood in this way, bias would then seem to signify any movement (deviation) away from this 'centrist' position.

So, to relate this back to the discussion, I would ask in what way are voters bias? Against what "unbiased account" are they being compared? There are many ways to read, interpret, process and classify information, especially information as complex and multidimensional as economic and social phenomena. Therefore, it would seem that all views are positioned ideologically in some way. Thus it follows that there is clearly NO unpositioned or unbiased criterion by which one can judge the degree of bias.

To conclude. By throwing out the term "bias", do we not also extricate ourselves from the complicated dichotomy of rational and irrational behavior?

[G]

Matthew,

I don't believe we need to abandon the term "bias" at all. Just by looking at what Caplan believes causes the bias in voting, we can see how biases, while pervasive, come in many different gradients.

Caplan believes that bias in politics is related to the extremely low cost of holding an incorrect political opinion. Contrast that with the cost of holding an incorrect in opinion in physics, say concerning the law of gravity. Could anyone try to demonstrably deny the law of gravity, the same way voters deny laws of economics? I don't think so, not if they enjoy living. There is direct feedback which corrects some biases more quickly than others, at steeper prices than others. Or as Mises pointed out so many times, experimentation can be used to dispel biases involved with the natural sciences. The same is true of almost all sorts of engineering.

But among voters, the costs of bias is very low, and are likely delayed some time from the act of voting (making causal relationships "I voted for X and Y happened" harder to see). Even among economists, the costs of bias are also pretty low, in my opinion, although much higher than those of voters (how many economic biases are truly testable in some fashion?).

If you accept Caplan's cause of bias, the biases of voters can be shown to be less than the biases displayed by economists. The later have established certain causal relationships in the economy, while the former have not. The later can, given their tools (praxeology, empirical testing, or whatever) correct many biases present. Your average voter having his average political conversation has no such tools. And even if he did, he'd pay little penalty for his incorrectness. Blatantly wrong economists are shunned from economics (thus no longer adding their bias to Caplan's expert polls); a steep price. Blatantly wrong laymen really don't care if college professors believe they are simpletons, so laymen must be more biased.

My 'unbiased' reference would be one any demonstrated bias would cause an immediate and obviously causal penalty to be paid to the biased person. Ideally this penalty would be (as in the case of bias against the law of gravity) death. Then demonstrated bias would quickly erase itself.

[Steve Miller]

"So, to relate this back to the discussion, I would ask in what way are voters bias? Against what "unbiased account" are they being compared?"

Against experts in the field. For economic issues, he compares the public's beliefs to those of economists. This has raised a couple of important objections.

Typical objections:
1) How do we know economists aren't biased toward free markets, etc.? (left-wing objection)
2) Maybe the economic consensus today favors economic freedom, but what about the economists of 50-60 years ago, weren't they very interventionist, socialist even? (right-wing objection)

The answer to #1 is that we can test a number of claims about economists' bias. For example, one hypothesis is that economists are all right-wing ideologues. But the SAEE shows that the median economist is a moderate democrat, and Klein and Stern's data on social scientists in California shows essentially that economists may be more right-leaning than other social scientists, but are still less conservative/more liberal than the general public. So there's no evidence to support the claim that economists are, as a group, ideologically biased. Incidentally, this is why one of the most common criticisms of MotRV is, in a word, stupid. You'll read over and over again that Caplan's argument is that people who disagree with ultra-libertarian Caplan about economics are biased. It's simply false, his argument is that the public's systematic disagreement with the consensus of (overall, politically moderate) economists is evidence of bias. The other argument, aside from claiming economists are right-wing ideologues, is that economists are biased because it is in their self-interest to view markets favorably, etc. Economists do, in fact, have higher incomes and more job security than most people. But, even controlling for differences in income, job security, etc. the belief gap between economists and the public is quite strong. In fact, there are two controls that (together) explain a very large portion of the belief gap between economists and the public: IQ and education. More intelligent, more educated people are less likely to exhibit the four biases Caplan identifies: anti-market, make-work, anti-foreign, and pessimistic.

The second objection is more troubling to me, but Caplan's explanation is, however bad you or I might think the economic consensus was during the height of Keynesianism, it was far less biased than the views of the general public. I tend to agree. Let me put it this way: who has a greater understanding and appreciation of markets: a Keynesian economist or a random member of the public. Personally, my money is on the Keynesian. Once you look at enough public opinion on economics, I think you'll agree with me.

The key point is, even in what many Austrians would consider to be the darkest days of the economics profession, the economic consensus was much more pro-market than the public consensus.

There is another objection that Jeffrey Friedman has made, but since there are plans for an entire issue of Critical Review on MotRV, that debate can play out there.

[Steve Miller]

I guess objection #2 is more of a libertarian objection than a "right-wing" objection.

[matthew mueller]

Steve Miller,

I think you misunderstood the point I was trying to make in my response. Instead of quoting the passage of mine you selected, I would have chosen this one:

"Therefore, it would seem that all views are positioned ideologically in some way. Thus it follows that there is clearly NO unpositioned or unbiased criterion by which one can judge the degree of bias."

Bias is a given. It exists among the public, AND among economists. I don't see how we could make the world intelligible without the assistance of bias. Also, I think it is unfair to use the "bias" of economists to measure the degree of bias among the public. For example, you can use the bias of economists as a criterion to measure the degree of bias of all other opinions, or you could use the bias of the public as a criterion; it makes little difference simply because ALL views are positioned ideologically in some way. It would be very bold of you to say that economists possess the truth and are therefore bias free.

But I don't think you would do this because you admit in your post that consensus among economists is by no means absolute --- differences of opinion do exist. Additionally, could it not be that education and IQ, instead of eliminating bias, simply redirect it? A Ph.D., for example, is not a sign of being bias free. It is just a specious claim to having authority over a particular subject.

To emphasize again. What sense does it make to say that the public are biased? Against what unbiased account? "Experts in the field"? Again, expertise does not eliminate bias. I always feel uneasy when citing 'philosophers' (especially postmodern philosophers) in economic debates, but I think the work of Michel Foucault is very instructive on this point. In his Power/Knowledge he makes the point that "truth" is a way of establishing order --- "[t]ruth acts as a censor -- it draws the line". Truth therefore is an artificial construct employed by recognized experts in a particular field that is responsible for determining the procedures that must be followed in order to arrive at solutions that are acceptable by the experts. From this, Foucault concludes that "[t]ruth is linked ... with systems of power which produce and sustain it."

[Steve Miller]

I know what you're saying, but I'd like to see a hypothesis about *why* economists are biased that actually holds up under scrutiny. Saying that economists rely on (necessarily imperfect) heuristics X, Y, or Z, isn't really the same as saying they're biased.

[Peter Boettke]

Matthew,

Caplan rejects the sort of post-modernist epistemology that you are relying on. As you have identified, there is an underlying 'true' model of the world, and then there are 'biased' estimators of that 'true' model.

Now, what Caplan does show is that there is a great consensus among economists than what the public thinks, and that even in that case when it comes to economic policy, the public does not choose to follow the consensus opinion. Instead, the public continues to suffer from an everyone is there own economist syndrom (something Buchanan has expressed concern over as well).

If you want to get a good sense of Caplan and my colleague Robin Hanson are up to, Hanson's blog "Overcoming Bias" is a good start. Hanson is completely taken with the Bob Auman puzzle of how 2 'rational' individual could ever disagree.

There is a literature in philosophy that has pursued these sort of questions --- not every philosopher is a post-modernist, just some. I would conjecture that there is probably more consensus among economists on policy than there is among philosophers on epistemology.

[Steve Miller]

Matthew,

After reading Pete's post, maybe I should bring Jeff Friedman's argument into the mix, since it lies somewhere between you and Caplan. You are kind of taking the stance, I think, that people have to use mental shortcuts, "untrue" beliefs about reality, to possibly make sense of reality, and those untrue beliefs, helpful though they are, are biased (I'm obviously a bit out of my depth here). So everyone, basically, is biased somehow, because of how they deal with the world around them.

Jeff Friedman's claim is that the public isn't biased, and economists are, but not because of self-interest, ideology, etc. The basic argument, and I hope I'm not mischaracterizing it, is that the public is simply too ignorant to be biased (they don't even know what the policy terms they use mean, they don't know what the terms "liberal" or "conservative" even mean, etc.). Economists, on the other hand, are far too informed about economics *not* to be biased. We have spent too much time wearing the same pair of glasses, to borrow Pete's metaphor. A trained economist tends to see everything in a certain way, and cannot take the glasses off. Thus, they are biased.

Just something to think about.

[G]

Matthew,
"To emphasize again. What sense does it make to say that the public are biased? Against what unbiased account? "Experts in the field"? Again, expertise does not eliminate bias."

Experts obviously pay a much higher price for their bias than a layman, and so will tend to be less biased. The layman's "systems of power" (which could literally be politics in this case?) are more removed from feedback from the object of study which supports or denies their "truths". What do you find incorrect about that?

Sure its true that an economist doesn't pay a terribly high price for being wrong compared to say, a civil engineer constructing a skyscraper. But I do feel inclined to point out, as Pete sort of did, that epistemologists and philosophers in general pay an even lower price for mistakes.

I'm referring to bias in the way Matthew seemed to to use it (as views deviant from reality, willful or not), not the way Steve did (which seemed to be just willful deviance?).