24 July, 2012

Race, IQ and wealth: A preliminary reply to Unz

In The American conservative, Ron Unz has written what I think is the only plausible attack on the Lynn & Vanhanen thesis that national prosperity is tied to average national IQ. As he himself remarks, it is amusing that the Leftists who oppose the thesis have never even attempted a dissection of the evidence for it. A conservative had to do the job for them.

The person best equipped to reply to Unz is of course Richard Lynn, and I imagine he will do so in due course. In the meanwhile, however, I hope I may be able to add a few notes that will be helpful towards developing the debate.

I don't intend to "fisk" what Unz has written but I cannot help noting my amusement at this statement: "Greeks and Turks have a bitter history of ethnic and political conflict, [but] modern studies have found them to be genetically almost indistinguishable". He would be crucified if he said that in Greece!

So for a start, let me say that I adopt a middle way. I think some IQ differences are important but don't think we can take all the Lynn statistics at face value and generalize from them. Why?

1). As even Lynn would admit, the sampling underlying most of them is poor to nonexistent. They are simply what is available and may therefore not be representative.

2). As a libertarian conservative, I see the prevalence of free markets as the key to prosperity so cannot see an overwhelming influence for IQ. Lynn might argue that smart people create market-dominated societies but there are too many instances of that not happening for that the be a good reply. Nobody disputes the high IQ of the Chinese and the Russians, for instance, but both are still tyrannies, although now more market-oriented than they were.

3). The correlation between IQ and prosperity is supposedly .80, which is extraordinarily high. For comparison, the contribution to IQ from genetics is usually held to be around .60. I am inclined to think that should set an upper bound for the contribution of IQ to prosperity. A very clear reason why the Lynn correlation is so high, however, is that it is an "ecological" correlation: A correlation between sets of pooled data. Such correlations are often very high and, as such, tend to exaggerate the underlying relationship.

4). The IQ figure that seems to matter in national development is not the mean (average) but rather the "smart fraction" -- i.e. the average IQ of the top 5%. This may not always track the national mean -- particularly in societies of mixed ethnicity.

An outstanding example of that is Israel. The brilliant contribution that Israel makes to knowledge generally (science, technology and even culture) cannot be denied yet Israel has on the latest Israeli army figures an average IQ of only about 100 -- which is typical of Western European societies that have nothing like the per head intellectual productivity of Israel. How come? It is because Israel is ethnically mixed (to put it mildly). The largest slice of the Jewish population came from Muslim lands, where IQs tend to be low, but there is also a substantial Ashkenazi (Western) element in the population that is generally brighter and which, more importantly, creates a smart fraction that is very smart indeed. That very bright smart fraction at the top of Israeli society is what accounts for Israel's outstanding intellectual accomplishments and the high level of Israeli civilization generally.

So it is only insofar as the smart fraction tracks the average IQ that average IQ is important to an understanding of economic development. It probably does do so in ethnically homogeneous populations but such populations are becoming increasingly rare.

5). And that leads me on to my next point: I admit to being a latecomer to the study of history but it has nonetheless been the focus of most of my studies for the last 20 years (See here, here here and here) and one thing that no student of European history can deny is the vast population movements and interminglings that have occurred in Europe for at least the last 2,000 years. European populations have been put through such a blender that it is rather a wonder that there are still different languages in Europe. That being so, I would not expect great differences between the innate abilities of different European nations. The considerable economic differences observable are therefore attributable to politics: How market-oriented have their political systems been?

So I would argue that Unz was studying precisely the wrong population subset in looking at predominantly European populations. If West Germany had not had the Adenauer/Erhardt combination immediately after the war might not Germany now be among the poorer countries of Europe? The rather dismal East German experience would certainly suggest so. So I am arguing that it is politics, not IQ, that accounts for differences in prosperity among European nations.

So what IQ differences do I take seriously? One that I am still doubtful about is Hispanics. I long ago challenged the validity of the figures that suggested a lower Hispanic IQ. It seemed to me that the great civilizational achievements of the pre-conquistador Meso- and South-Americans cast such figures into doubt. But this study of poor educational progress among Hispanic children is hard to get around. So it is rather interesting that Unz has pulled out some recent GSS figures that show Hispanics as having a much higher average IQ than was thought. But how many Hispanics participate in the GSS survey? Maybe only the brighter ones.

Another doubt concerns the poor assimilation of Hispanic youth. The children of just about any immigrant group that has come to America have usually assimilated completely to the host society. Your surname may be Krikorian, which means that your parents came from Armenia, but you yourself will be so assimilated that you may even head an American think-tank. But such a complete assimilation seems to be missing from the children of Hispanics. Anyone famililiar with Hispanic gangs will not be surprised to hear that the children of Hispanics born in the USA are highly crime-prone, though not as crime-prone as blacks. Crime is strongly associated with intelligence (negatively) so this would on face value suggest a low average IQ in the group concerned. In this case, however, the very fact that most are the children of people who broke the law may account for their criminality.

In summary, I think the old Scottish verdict of "not proven" is best applied to all claims about Hispanic IQ. The indicators are conflicting.

One IQ difference that I have no doubt about, however, is the most incendiary and "incorrect" one of all: The black/white difference. It is at once multiply replicated and multiply validated. To put that another way, it always emerges in any examination of it and blacks behave exactly as you would expect a low IQ group to behave, with appallingly low levels of occupational, economic and educational achievement and appallingly high levels of criminality. They are so bad at getting what they want by legitimate means that they very frequently resort to crime to get some semblance of what they want. And that applies not only to African-Americans but also to Africans in Africa, Africans in Britain and Africans everywhere.

And American academics and educators have run themselves ragged trying to get black educational achievement up to white levels. They have tried everything conceivable for many years without success. IQ and educational achievement are highly correlated but among blacks the IQ required for a high level of educational achievement is usually just not there.



Go to John Ray's Main academic menu
Go to Menu of longer writings
Go to John Ray's basic home page
Go to John Ray's pictorial Home Page
Go to Selected pictures from John Ray's blogs