Bryan Caplan  

Baffled by "Analytical Egalitarianism"

PRINT
Envy and Resentment... California Energy Tax Proposal...

There's a peculiar doctrine coming out of George Mason in recent years. It's called "analytical egalitarianism," and has been energetically promoted by my brilliant colleague David Levy, his co-author Sandra Peart, and quite a few others. (See here and here for the main book-length treatments; see here here, and here for related Econlib columns). I've read quite a bit of this work, and argued with David in many a seminar. As best as I can tell, analytical egalitarianism amounts to the following:

Proposition 1. All people are equally talented and have the same preferences.

Proposition 2. If people do have unequal talents or different preferences, these are ENTIRELY the result of environmental differences.

Proposition 3. If there are any innate/genetic differences between people, it's exceedingly dangerous for the world if anyone says so.

I'd like to be more charitable, but I can't. The arguments Levy, Peart, and others marshall for the preceding three propositions come down to:

Evidence for Prop. 1. Historical analyses of allegedly innate differences that eventually faded away (like Irish underachievement), and some instances where experts were wrong, and non-experts were right.

Evidence for Prop. 2. Appeal to the authority of Adam Smith and J.S. Mill, such as:

The difference between the most dissimilar characters, between a philosopher and a common street porter, for example, seems to arise not so much from nature, as from habit, custom, and education. (Smith, Wealth of Nations)
Of all vulgar modes of escaping from the consideration of the effect of social and moral influences on the human mind, the most vulgar is that of attributing the diversities of conduct and character to inherent natural differences. (Mill, Principles of Political Economy)

Evidence for Prop. 3. Detailed discussions of the life and works of odious literary figures like Carlyle and Ruskin, as well as various eugenicists who range from boorish to murderous.

I find this evidence underwhelming. Point by point:

Rebuttal to Prop. 1. The fact that some allegedly innate differences went away hardly shows that there are no innate differences. There are plenty of allegedly innate differences that failed to go away, too.

Similarly, the fact that non-experts have sometimes beaten the experts hardly shows that everyone is equally talented and knowledgeable. No one's infallible; you've got to look at average performance (or, if Levy prefers, median performance).

Rebuttal to Prop. 2. Frankly, who cares what Adam Smith and J.S. Mill thought about heredity vs. environment? How can you prefer the speculation of two economists - however brilliant - to the hard facts of modern twin and adoption studies? There is now an overwhelming consensus that genetics explains a great deal about human variation. Smith and Mill were just wrong - though in fairness to Smith, the above quote says only that nature is less important than other factors, not that it doesn't matter.

Rebuttal to Prop. 3. Yes, there have been some loathsome people who have "reasoned" from the fact of innate human inequality to the conclusions of murder and slavery. But these malevolent conclusions do not remotely follow from the premises, as I argued here.

Even more tellingly, there have been plenty of loathsome people who have "reasoned" from the fact of innate human equality to the conclusions of murder and slavery. Check out the Black Book of Communism or my Museum of Communism for details.

The bottom line is that atrocities have been committed in the names of nature and nurture alike. Extreme environmentalism is no protection against man's inhumanity to man. In fact, to defend extreme environmentalism in the face of mountains of evidence from twin and adoption studies implicitly grants the crazy premise that IF non-environmental differences existed, murder and slavery would OK.

Lest I be misunderstood, let me say that many proponents of Analytical Egalitarianism, most notably Levy himself, are extremely smart and knowledgeable. Their very existence is evidence against Proposition 1, because they've got way more than the average endowment of ability. Just like the Austrians, I'm going after analytical egalitarians because they're wasting their talent and energy defending the undefendable.


Comments and Sharing





TRACKBACKS (3 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/587
The author at The Filter^ in a related article titled Analytical Egalitarianism II writes:
    I've harped on about this before, but was interested in Bryan Caplan's critqiue of the Levy-Peart project: I'm going after analytical egalitarians because they're wasting their talent and energy defending the undefendable.Caplan views analytical egalit... [Tracked on October 27, 2006 4:41 AM]
COMMENTS (14 to date)
Steve Sailer writes:

Well said.

Of course, "analytical egalitarianism" isn't unique to Levy and Peart. It's the Conventional Wisdom, overwhelmingly. If you explicitly dissent from it, citing emprical examples of differences in talent between groups, your career will be punished.

Tim Lundeen writes:

Also in Adam Smith's defense, the mechanisms for sorting by ability were very much less developed in his day. He would have seen a lot more variation in ability in all the roles in society than we do mow. So Adam Smith might well have seen people in more menial, less educated roles who were quite bright, people who just didn't have the opportuny to do better for themselves. Today, it is easy to earn a living that makes good use of one's abilities, and variation in ability by profession is much more narrow.

Scott Clark writes:

The way I viewed analytical egalitarianism when i was in Levy's class was not that everyone is the same as everyone else. But when analyzing and making policy decisions, it best to act as if everyone were equal.

The problems arise when the english think they are inherently better than the irish and make policy that is imposed on them.

Certainly there are genetic differences between people, but just because there are differences doesn't mean there should be different sets of rules for different people.

At least that's what I got out of the concept. If Dr. Levy is pushing anything more sweeping than that, I must have missed it.

Ragerz writes:

Interesting commentary.

Why I am not particularly impressed with "analytical egalitarianism" I am not sure that the claim in your previous article that "one of the main causes - if not the main cause - of economic, cultural, and other forms of success is genetic" is backed up by "massive empirical support" is correct.

First of all, success is not a single-dimensional thing and is a very difficult thing to measure. Indeed, I would even venture to suggest that "success" cannot even be defined without reference to normative judgments. Thus, empirical evidence is ill-equiped to resolve the question. Empirical evidence resolves very specific questions that one might have about particular attributes. For example, person X has a better memory (and potential to remember) than person Y. While person Y has better analytic geometry problem solving skills (and a better potential in this area) than person X. Surely, empirical evidence can answer the question of who has the better memory, or who can solve geometry problems faster. However, what it can't do is tell you how to aggregate these results. If X is better in attributes A, D, and F by certain varying magnitudes and person Y is better and B, C, and E by certain varying magnitudes, we have no non-arbitrary means of giving weight to and aggregating these attributes. Is attribute A twice as important as C, or is it three times as important...? Thus, the idea that a particular person is "superior" to another is extremely problematic. (Unless X is superior to Y in all possible attributes. This would be a very rare case. And when it does occur, we might also ask are we measuring ALL possible attributes?)

Second, the idea that "cultural success" in particular is genetic is problematic. Different civilizations have been technologically ahead or behind with respect to various accomplishments and inventions at different times. If "cultural success" were genetic, you would expect that only one society and culture, the genetically superior one, would be superior, throughout all times, within a given dimension of success.

Third, no one denies the ability of an unfavorable (or less than maximally favorable) environment in limiting one's potential. To the extent that environment has limited a person Y's potential, we cannot say that person X's superior performance with respect to a particular attribute is genetic. Further, it is nearly certain that most everyone's potential is limited by environment factors to some degree. The fact that you were not taught certain skills from a very young age (and no one is taught every skill imaginable) limits your potential vis a vis that skill.

Fourth, anyone in the real world knows that economic success to a large degree is a matter of relationships with others. Why the ability to build relationships with others is a skill that might be influenced by genetics, there are certain relationships which happen due to random chance or that one is born into. To the extent that economic success is dependent on relationships and those relationships are a product of chance (and aren't most relationships a product of chance to some degree?) and/or birth, one cannot say that economic success is genetic.

Fifth, while one's ability to be disciplined (to align ones actions with one's ideals) may be partially genetic, there is no doubt that discipline can be taught to most people. (i.e. discipline is partially an environmental factor. Those who have the ability to disregard their inclinations will have the ability to be more successful, whether you define success hedonistically (willing to forego short-term pleasure for greater long-term pleasure) or in nobler terms (able to align one's actions with one's ideals).

Sixth, there is no set non-arbitrary formula for weighing environment variation versus genetic explanation with respect to a particular attribute. In any study involving identical twins, many environment factors will be the same and many will be different. There is no set formula for determining how to weigh factors that were both the same. For example, if both children receive adequate nutrition, how much did that common factor lead this particular set of children to have common experiences, versus another set of children who also have adequate nutrition. That is, it would not be suprising if inadequate nutrition effects children differently due to genetic factors. Thus, while you could isolate the weight one should given inadequate nutrition for a particular set of twins (if you were unethical enough to put your subjects in such an environment) you could not assume that the effect would be the same for other sets. Only by repeating this particular experiment many times (with everything else held constant) would one be able to understand the likely distribution of impacts of inadequate nutrition. Essentially, what this comes down to is that there is no principled way to assign weights to environmental deficiencies that would be unethical to apply in an experiment, but which are common in real life. It would not be THAT suprising if identical twins end up more similar when there are not any major environmental deficiencies, but would end up quite different if one were subject to major environmental challenges that would be unethical to purposely introduce. Our only data with respect to such deficiencies occur by happenstance, and it is inadequate. As the following Wikipeda article puts it: "With virtually all psychological traits however, there is an intermediate mix of nature and nurture, and opinions about the relative importance of each will often vary widely."

Overall, what we have is overwhelming empirical data that genetics matters, not overwhelming empirical data that environment does not. There is not yet a principled way to assign weights to nature versus nurture with respect to particular attributes. Though in principle, this could be done, to the extent that such attributes can be measured objectively. (Of course, the choosing of what attributes to measure is not exactly a matter of science. What level of generality should we measure an attribute? Are all important attributes subject to measurement? For example, how do you measure how manipulative someone is? At best, for things like how manipulative a person is, you are stuck with qualitative data. Finally, and most importantly, we are talking here about the problem of measuring particular discrete attributes. But there is and can never be a non-arbitrary means of giving weight to and aggregating these attributes. But of course, when you judge a whole person, that is precisely what one must do. One cannot say that Adam Smith was wrong when he said: "The difference between the most dissimilar characters, between a philosopher and a common street porter, for example, seems to arise not so much from nature, as from habit, custom, and education." Why? Because of this aggregation problem, among many other problems.

Of course, there are some easy cases. We know that a street porter that is mentally retarded because of genetics could never be a philosopher, regardless of environment. However, it would take a deep and abiding and irrational faith in genetic determinism to conclude that there does not exist even one street porter that could have been a philosopher, if only environmental factors had been different.

Finally, it should be mentioned that the outcome of this debate does not determine our normative values. Genetics and environment are both within and beyond our control to some degree. That a particular inequality arises due to genetics neither implies that society should reinforce, nor correct, nor remain neutral with respect to that inequality. In the general case, genetic inequalities do not normatively justify social or relational inequalities. There is no doubt that inequalities arising from genetics do occur. But knowing that this is so, does not tell us how we ought to organize society. (Though, of course, in particular cases, genetic information can inform our decision-making according to our normative standards. But genetic information does not tell us what our normative standards ought to be.)

K writes:

I think 'vulgar' had a different meaning to Mills than to readers today.

The three premises seem utter tripe. (I don't know much about tripe but you wouldn't like the words I could have substituted.)

The program as stated looks like a jump back to when Marx, et al, began pondering why and where the industrial age would lead mankind.

Surely Levy and the others are not doing that.

Omer K writes:
Lest I be misunderstood, let me say that many proponents of Analytical Egalitarianism, most notably Levy himself, are extremely smart and knowledgeable.

Not so smart or knowledgeable if the old environment is king is what they are spouting...

Unless you define smart as pandering to the powers that be? Call me cynical, but I call that spineless, not smart.

Tom West writes:

In the general case, genetic inequalities do not normatively justify social or relational inequalities.

I suspect that it is the belief of the Levy and others that widespread belief in genetic inequalitoes *will* inexhorably push society towards social inequalities. Think of the Steve Sailors et al, who would restrict immigration from certain racial groups. Given the patina of scientific credibility, there's a significant chance that such odious practices could gain significant traction.

Let's take a slightly less a slightly less odious example. There's a claim that women's IQ has a smaller standard deviation than men's. If one lends that serious credence, then it makes sense to transfer funding from programs supporting high ability women (because there are so few) to enhanced general education (which will benefit the vast majority) or even to programs supporting high ability men (who are far more numerous). For example, we don't have educational programs aimed at those with 180+ IQ because there are so few. Money for such programs would be used for general enhanced classes for those with 130+ IQ, who occur in significant numbers.

Failure to do so, once backed up with "fact", becomes a case of outright favoritism towards a very small number of students (as opposed to current high-ability programs which are for a small, but educationally significant number of students) in the face of irrefutable logic. You can start making credible cases that those who still want classes for females in the high end are misusing educational funds.

To argue against this is essentially to argue that your religion of equality should supercede science and logic.

I'll leave the next step (racial differences) as an exercise for the reader...

So in conclusion, I suspect the authors believe that acceptance of the science of inequality of ability will lead to social inequality (a view I hold as well), so they fight the trend where it is currently most vulnerable, the science itself, and the reputation of those who pursued that science in the past.

An unintelligent design for science

From a biological and medical perspective these Analytical Egalitarianism propositions are simply absurd.

With these matters you have to decide whether you are doing science, or not. As soon as you start worrying about the possible socio-political effects of your conclusions are you have stopped doing science.

The description of Analytical Egalitarianism sounds like they are arguing back from what they consider desirable policy to generate what they consider acceptable science. But this procedure has nothing to do with science - being, indeed, analogous to the procedure used to generate 'Intelligent Design' as a supposed alternative to Natural Selection.

aaron writes:

What's with the double post?

Randy writes:

It seems to me that a truly analytical egalitarianism has only one possible proposition; While we understand that no two people are actually equal by any objective measure, we believe there is value in treating everyone as equal anyway.

meep writes:

Um, how much scientific research does one need to how ridiculous this is? I'm a 5'4" tall woman. I was never going to be an NBA player, no matter how milk I guzzled. Likewise, my husband was never going to give birth to three kids, like I did, no matter how I would have preferred for him to have done it.

Then people say "Of course, we didn't mean =physical= abilities, which are obviously different from cognitive abilities."

Then I have to wonder why cognitive abilities are supposed to be free from genetic differences, where they are evident in qualities such as height and strength. Are they trying to place cognitive abilities in the range of the non-material, seating "mind" in the soul?

eric writes:

I like Charles Murray's thought experiment in The Bell Curve: If you could choose to bequeath your child $1 million (or $10 million) or 15 more points in IQ, which would you choose? I would choose 15 IQ points. And as as former skeptic James Heckman found out IQ is relatively immutable.

Silviu writes:

I'm with Scott Clark on this one. Levy’s point has nothing to do with nature vs. nurture.

The "analytical egalitarianism" simply points out that a dictator's preferences do NOT count MORE when it comes to policy than those of a regular person, regardless of IQ. It is a direct response to Plato's Republic: there is no reason to hand policy over to "philosopher kings", however “smart” and good at geometry they may be. At the end of the day my low/high IQ does not give me the right to ban wrestling and make skiing a national, heavily subsidized sport. My preferences are no “better” or “worse” in some absolute sense than those of the next person: de gustibus non est disputandum.

Historically, the discussion was of course very often falsely linked to that about IQs. This is misleading, and comes from confusing ends and means. IQ helps you choose better means perhaps, but provides no guidance as to what the goals themselves SHOULD be. The smartest person in the world has little right to tell you which ice-cream flavor is “best”.

paul writes:

I have four children. Levy is wrong.

Comments for this entry have been closed
Return to top