Dan Klein closes his critique of The Myth of the Rational Voter with recommendations for the next edition:

1. Soften the emphasis on beliefs and bias. Go more with bents.

2. Drop all “rational” talk. Just go with terms like folly, foolish, unenlightened.

3. Confess libertarian sensibilities. Confess that the package you are selling is broader than a technical analysis of the effectiveness of democracy.

4. Doing so will enable you to take the stress off of expertise and the like. Acknowledge high-strata statist bents.

5. Place the work within liberalism through the ages, including esp Hayek and Public Choice.

I don’t plan on following much of this advice. Here’s why, point-by-point.

1. I will never switch from “beliefs” to “bents.” Laymen and experts alike already understand the concept of belief. “Bents,” in contrast, is vague to laymen and confusing to academics.

2. I am not going to abandon “rational” talk in favor of “unenlightened,” especially when we closely examine what Dan means by “enlightened.” Here’s Dan:

“Enlightened” does not mean “good for you.” It does not mean “unbiased errors.” It means beliefs that we [libertarians] think would better the condition of humanity.

Frankly, this is extremely sloppy thinking. I can imagine why you might define enlightenment as “what would in fact better the condition of humanity.” But to define it in terms of what libertarians think would better the condition of humanity is absurd. What if what we think is wrong? Would it still make sense to call ourselves “enlightened”? If not, then Dan’s definition is wrong.

In any case, Dan forgets the plain fact that false beliefs can better the condition of humanity. Visualize a world where everyone believed that his head would explode if he tried to commit murder.

As far as I can tell, Dan’s objections stem from a larger “pragmatist” philosophical agenda which objects to all the concepts I regard as indispensable, starting with reality, truth, and rationality. I think this agenda is obviously wrong.

3. Should I more boldly confess to libertarian sensibilities, and try to sell the whole package? Judging from the reviews, many people thought my book was a bold confession of libertarian sensibilities. But in any case, trying to sell the whole package is simply bad marketing. I want to reach a broader audience, and the best way to do that is to focus on a narrower issue and find some common ground.

4. I don’t want to take the stress off expertise, because I genuinely accept the presumption that if experts and laymen disagree, the experts are probably right.

Dan’s best point is that I should focus more on “high-strata biases.” In a sense, my whole book is an attack on one high-strata bias: Social scientists’ belief in rational voter models. Still, I think Dan is right that high-strata biases exist, and merit further study. If these studies go somewhere, I’d be happy to add them to my next edition.

5. “Placing the work within liberalism through the ages” seems like more bad marketing to me. The whole point of argument is to find some common ground with your audience, then move forward. Dan’s approach fails to do that.

Should I have spent more time relating my work to Public Choice? To be honest, I think the connection between my work and this broader literature is hard to miss. I’ve got a whole chapter on my criticism of the field.

Should I have spent more time relating my work to Hayek? I don’t see why, because (a) Hayek (unlike Mises) didn’t influence my thinking on these issues; (b) he’s a terrible writer; and frankly, he’s grossly overrated, largely because his writing is so full of obsfucation.

To conclude: I think Dan Klein has been very successful precisely because he doesn’t follow the advice he gives to me. Reality, truth, and rationality are the guiding stars of his research – look at how carefully he crafts his surveys. And as his Econ Journal Watch shows, there are few economists alive who work harder than Dan to find common ground with intellectual opponents, and move forward from there.

The moral: Do as Dan does, not as he says.