Previously by Ron Unz: China's Rise, America's Fall
At the end of April, Charles Kenny, a former World Bank economist specializing in international development, published a blistering attack in Foreign Policy entitled u201CDumb and Dumber,u201D with the accusatory subtitle u201CAre development experts becoming racists?u201D Kenny charged that a growing number of development economists were turning towards genetic and other intrinsic human traits as a central explanation of national economic progress, often elevating these above the investment and regulatory issues that have long been the focus of international agencies.
Although Kenny suggested that many of his targets had been circumspect in how they raised these highly controversial ideas, he singled out IQ and the Wealth of Nations, published in 2001 by Richard Lynn and Tatu Vanhanen, as a particularly extreme and hateful example of this trend. These authors explicitly argue that IQ scores for different populations are largely fixed and hereditary, and that these rather than economic or governmental structures tend to determine the long-term wealth of a given country.
Kenny claimed that such IQ theories were not merely racist and deeply offensive but had also long been debunked by scientific experts notably the prominent biologist Stephen Jay Gould in his 1980 book The Mismeasure of Man.
As Kenny soon discovered from the responses to his online article, he had seriously erred in quoting the authority of Gould, whose fraud on race and brain-size issues, presumably in service to his self-proclaimed Marxist beliefs, last year received further coverage in the New York Times. Science largely runs on the honor system, and once simple statements of fact in Gould's case, the physical volume of human skulls are found to be false, we cannot trust more complex claims made by the particular scholar.
Despite Kenny's obvious lack of familiarity with the technical questions he raised, these issues remain important ones to explore, given today's globalized world. After all, it is generally acknowledged that some people are smarter than other people, and this almost syllogistically raises the possibility that some peoples may be smarter than other peoples.
Most nations prefer material wealth to poverty, and it seems plausible that smarter people might be better at generating the productivity needed to achieve this goal. We should hardly be surprised that this possible factor behind economic advancement has attracted the interest of the development experts criticized by Kenny, and just as he alleges, IQ and the Wealth of Nations ranks as perhaps the most extreme academic example of this analysis.
Although u201Cintelligenceu201D may be difficult to define precisely, most people have accepted that IQ scores seem to constitute a rough and measurable proxy for this trait, so Lynn and Vanhanen have collected a vast number of national IQ scores from the last 50 or 60 years and compared these to income levels and economic growth rates. Since experts have discovered that nominal IQ scores over the last century or so have tended to rise at a seemingly constant rate the so-called u201CFlynn Effectu201D the authors adjusted their raw scores accordingly. Having done so, they found a strong correlation of around 0.50–0.75 between the Flynn-adjusted IQ of a nation's population and its real per capita GDP over the last few decades, seemingly indicating that smarter peoples tend to be wealthier and more successful.
From this statistical fact, Lynn and Vanhanen draw the conclusion that intelligence leads to economic success and since they argue that intelligence itself is largely innate and genetic that the relative development ranking of the long list of nations they analyze is unlikely to change much over time, nor will the economic standing of the various groups within ethnically mixed countries, including the United States.
Now this hypothesis might indeed be correct, but it is not necessarily warranted by the empirical data that Lynn and Vanhanen have gathered. After all, if high national IQ scores are correlated with economic success, perhaps the high IQs cause the success, but it seems just as possible that the success might be driving the high IQs, or that both might be due to some third factor. Correlation does not imply causality, let alone the particular direction of the causal arrow. A traditional liberal model positing that socio-economic factors strongly influence performance on academic ability tests would predict exactly the same distribution of international results found by Lynn and Vanhanen.
Fortunately, a careful examination of the wealth of empirical data they have gathered provides some important evidence on the relative plausibility of these conflicting hypotheses, allowing us to draw useful conclusions in this extremely taboo subject.
The Distribution of European Intelligence
Critics have often suggested, not without some plausibility, that when Western-designed IQ tests are applied to Third World peoples, the results may be distorted by hidden cultural bias. There is also the possible impact of malnutrition and other forms of extreme deprivation, or even practical difficulties in administering tests in desperately impoverished nations, as Kenny emphasized in his critique.
In order to minimize these extraneous factors, let us restrict our initial examination to the 60-odd IQ datapoints Lynn and Vanhanen obtained from European countries and their overseas offshoots over the last half-century. Obviously, some of these countries have at times been far poorer than others, but almost none have suffered the extreme poverty found in much of the Third World.
What we immediately notice is a long list of enormous variations in the tested IQs of genetically indistinguishable European peoples across temporal, geographical, and political lines, variations so large as to raise severe doubts about the strongly genetic-deterministic model of IQ favored by Lynn and Vanhanen and perhaps also quietly held by many others. (Unless otherwise indicated, all the IQ data that follow are drawn from their work and incorporate their Flynn adjustments.)
Consider, for example, the results from Germany obtained prior to its 1991 reunification. Lynn and Vanhanen present four separate IQ studies from the former West Germany, all quite sizable, which indicate mean IQs in the range 99–107, with the oldest 1970 sample providing the low end of that range. Meanwhile, a 1967 sample of East German children produced a score of just 90, while two later East German studies in 1978 and 1984 came in at 97–99, much closer to the West German numbers.
These results seem anomalous from the perspective of strong genetic determinism for IQ. To a very good approximation, East Germans and West Germans are genetically indistinguishable, and an IQ gap as wide as 17 points between the two groups seems inexplicable, while the recorded rise in East German scores of 7–9 points in just half a generation seems even more difficult to explain.
The dreary communist regime of East Germany was certainly far poorer than its western counterpart and its population may indeed have been u201Cculturally deprivedu201D in some sense, but East Germans hardly suffered from severe dietary deficiencies during the 1960s or late 1950s when the group of especially low-scoring children were born and raised. The huge apparent testing gap between the wealthy West and the dingy East raises serious questions about the strict genetic interpretation favored by Lynn and Vanhanen.
Next, consider Greece. Lynn and Vanhanen report two IQ sample results, a score of 88 in 1961 and a score of 95 in 1979. Obviously, a national rise of 7 full points in the Flynn-adjusted IQ of Greeks over just 18 years is an absurdity from the genetic perspective, especially since the earlier set represented children and the latter adults, so the two groups might even be the same individuals tested at different times. Both sample sizes are in the hundreds, not statistically insignificant, and while it is impossible to rule out other factors behind such a large discrepancy in a single country, it is interesting to note that Greek affluence had grown very rapidly during that same period, with the real per capita GDP rising by 170 percent.
Furthermore, although Greeks and Turks have a bitter history of ethnic and political conflict, modern studies have found them to be genetically almost indistinguishable, and a very large 1992 study of Turkish schoolchildren put their mean IQ at 90, lending plausibility to the low Greek figure. We also discover rather low IQ scores in all the reported samples of Greece's impoverished Balkan neighbors in the Eastern Bloc taken before the collapse of Communism. Croatians scored 90 in 1952, two separate tests of Bulgarians in 1979–1982 put their IQs at 91–94, and Romanians scored 94 in 1972. While the low scores of the Croatian children might be partly explained by malnutrition and other physical hardships experienced during the difficult years of World War II, such an excuse seems less plausible for other Balkan populations tested decades after the war, all of which seem to score in the same range.
Two samples of Poles from 1979 and 1989 provided widely divergent mean IQs of 106 and 92, with the low Polish figure of 92 coming from a huge sample of over 4000 children tested with u201CProgressive Matrices,u201D supposedly one of the most culturally-independent methods. On the other hand, more economically advanced Communist countries in Central Europe often had considerably higher scores, with the Slovaks testing at 96 in 1983, the Czechs scoring 96–98 in 1979–1983, and the Hungarians reaching 99 in 1979.
All of these Southern or Eastern European IQ scores follow the per capita GDP of their countries, a correspondence that supports either the IQ-makes-wealth hypothesis of Lynn and Vanhanen, or the contrary wealth-makes-IQ hypothesis of traditional liberals.
During this same period, the far richer non-Communist nations of Europe such as Austria, Britain, the Netherlands, Belgium, Norway, and West Germany all tended to score at or somewhat above 100. The wide IQ gaps between these European peoples and the previous group seem unlikely to have a heavily innate basis, given the considerable genetic and phenotypic similarity across these populations. For example, the borders of Austria and Croatia are just a couple of dozen miles apart, both are Catholic countries that spent centuries as part of the Austro-Hungarian Empire, and it is quite difficult to distinguish Austrians from Croatians either by appearance or by genetic testing. Yet the gap between their reported IQ scores 12 points is nearly as wide as that separating American blacks and whites.
It seems more plausible that most of the large and consistent IQ gaps between Western Europeans and their Balkan cousins are less a cause than a consequence of differences in development and affluence during the era in which these IQs were tested. For example, Austria had many times Croatia's per capita GDP during the period in question. One of the few European nations to exhibit a sharp decline in tested IQ, Poland whose score fell from 106 in 1979 to 92 in 1989 did so amid the economic turmoil of the 1980s, when its per capita GDP also substantially declined according to some measures, even while Western Europe was growing richer.
If these differences of perhaps 10 or even 15 IQ points between impoverished Balkan Europeans and wealthy Western ones reflected deeply hereditary rather than transitory environmental influences, they surely would have maintained themselves when these groups immigrated to the United States. But there is no evidence of this. As it happens, Americans of Greek and South Slav origins are considerably above most other American whites in both family income and educational level. Since the overwhelming majority of the latter trace their ancestry to Britain and other high IQ countries of Western Europe, this would seem a strange result if the Balkan peoples truly did suffer from an innate ability deficit approaching a full standard deviation.
Similar sharp differences occur in the case of Italian populations separated historically and geographically. Today, Italian-Americans are very close to the national white average in income and education, and the limited data we have seem to put their IQ close to this average as well. This would appear consistent with the IQ figures reported for Italy by Lynn and Vanhanen, which are based on large samples and come in at just above 100. However, there is a notoriously wide economic gap between northern Italy and the south, including Sicily. The overwhelming majority of Italian-Americans trace their ancestry to the latter, quite impoverished regions, and in 2010 Lynn reported new research indicating that the present-day IQ of Italians living in those areas was as low as 89, a figure that places them almost a full standard deviation below either their Northern Italian compatriots or their separated American cousins. Although Lynn attributed this large deficit in Southern Italian IQ to substantial North African or Near Eastern genetic admixture, poverty and cultural deprivation seem more likely explanations.
The Lynn/Vanhanen data on Jews also provide some suspicious IQ disparities. American Jews have among the highest tested IQs, with means being usually reported in the 110–115 range. Yet Lynn and Vanhanen report that Israeli Jews have strikingly low IQs by comparison. One large sample from 1989 put the figure at 90, while a far smaller sample from 1975 indicated an IQ of 97, with both results drawn from Israel's large Jewish majority rather than its small Arab minority. The IQ gaps with American Jews are enormous, perhaps as large as 25 points, and difficult to explain by genetic factors, since a majority of Israel's Jewish population in that period consisted of ethnic Askhenazi (European) Jews, just like those in America. The huge economic gulf between Israeli Jews, who then had less than half the average American per capita GDP, and American Jews, who were far above average in American income, would seem to be the most plausible explanation.
Similarly, a large 1990 test of South African whites placed their IQ at 94, considerably below that of the Dutch or English peoples from whom they derive, and again this may be connected to their lower level of national income and technological advancement.
Perhaps the strongest evidence supporting this cultural rather than genetic hypothesis comes from the northwestern corner of Europe, namely Celtic Ireland. When the early waves of Catholic Irish immigrants reached America near the middle of the 19th century, they were widely seen as particularly ignorant and uncouth and aroused much hostility from commentators of the era, some of whom suggested that they might be innately deficient in both character and intelligence. But they advanced economically at a reasonable pace, and within less than a century had become wealthier and better educated than the average white American, including those of u201Cold stocku201D ancestry. The evidence today is that the tested IQ of the typical Irish-American to the extent it can be distinguished is somewhat above the national white American average of around 100 and also above that of most German-Americans, who arrived around the same time.
Meanwhile, Ireland itself remained largely rural and economically backward and during the 1970s and 1980s still possessed a real per capita GDP less than half that of the United States. Perhaps we should not be too surprised to discover that Lynn and Vanhanen list the Irish IQ at just 93 based on two samples taken during the 1970s, a figure far below that of their Irish-American cousins.
Even this rather low Irish IQ figure is quite misleading, since it was derived by averaging two separately reported Irish samples. The earlier of these, taken in 1972, involved nearly 3,500 Irish schoolchildren and is one of the largest European samples found anywhere in Lynn/Vanhanen, while the other, taken in 1979, involved just 75 Irish adults and is one of the smallest. The mean IQ of the large group was 87, while that of the tiny group was 98, and the Lynn/Vanhanen figure was obtained by combining these results through straight, unweighted averaging, which seems a doubtful approach. Indeed, a sample of 75 adults is so small it perhaps should simply be excluded on statistical grounds, given the high likelihood that it was drawn from a single location and is therefore unrepresentative of its nation as a whole.
So we are left with strong evidence that in the early 1970s, the Irish IQ averaged 87, the lowest figure anywhere in Europe and a full standard deviation below than that of Irish-Americans, a value which would seem to place a substantial fraction of Ireland's population on the edge of clinical mental retardation.
Lynn seems to have accepted this conclusion. The current issue of the academic journal Personality and Individual Differences is organized as a tribute to Lynn and contains a lengthy interview in which he describes the turning points of his career, beginning with his appointment as a research professor in Dublin. His official responsibility was to investigate the social and economic problems of Ireland, and he soon concluded that the nation's backwardness was largely due to the low IQ of its people, with the only obvious solution being a strong eugenics program, presumably including sterilization of a substantial fraction of the population. But given the dominant influence of conservative Catholicism in Ireland, he doubted the government would consider such suggestions, which would probably just get him u201Caccused of being a Nazi,u201D so he u201Cchickened outu201D and chose to suppress his findings. A few years later, he relocated to Protestant-run Ulster, where he felt his racial ideas might find a more receptive audience, and he eventually became interested in whether the poverty of other countries might be due to the same low IQ causal factor which he believed explained Ireland's problems. This led him to the research that culminated in the publication of IQ and the Wealth of Nations.
But Lynn's late-1960s views regarding the mostly genetic cause of low Irish IQ seem unwarranted. Ireland was then overwhelmingly rural and poor, with a low per capita GDP, while Irish Americans tended to be an urban population and a reasonably affluent one, and this sharp difference in external material conditions seems the most logical explanation for the wide disparity in IQ results. In further support of this environmental hypothesis, we should note that it has been estimated that nearly one-third of Australia's population is wholly or substantially Irish in ancestry, with the balance mostly British, while the IQ results Lynn and Vanhanen report for Australia are all very close to the British average of 100.
The gathering of social science data, including national IQs, is fraught with difficulty, notably due to sampling problems, and two or three anomalous results might be explained away for those reasons. But the large number of examples cited above in which genetically indistinguishable European-ancestry populations show enormous variations in tested IQ seems to indicate a much broader difficulty. Not only are the results too numerous to be ascribed to chance error, but they follow a consistent pattern of their own, with European-ancestry groups living in affluent, well-developed countries almost invariably having IQ scores of around 100 or above, while their close kinsmen in much poorer regions have far lower scores. Indeed, in several of these cases, the countries and peoples are identical, being merely separated by a generation or less of local economic development.
To a small extent, Lynn and Vanhanen acknowledge the possible importance of non-genetic factors, and they devote a few pages to a discussion of the impact of health, nutrition, and education on IQ scores. But they never provide any clear estimate for the magnitude of these influences and claim that a number of twin or adoption studies have determined that IQ is 80 percent or more heritable. Their text seems to assume that genetics is the overwhelmingly dominant factor behind the national IQ disparities which they catalogue.
All IQ data was drawn from Lynn/Vanhanen. The per capita GDP figures are obtained from the World Bank and adjusted for Purchasing Power Parity (PPP 2005$) if available; otherwise being marked with an asterisk. Much of this economic data, especially for non-convertible East Bloc currencies before 1989, is somewhat uncertain and should be used only for rough comparative purposes.
Questioning the u201CStrong IQ Hypothesisu201D
The central thesis of Lynn and Vanhanen's work might be called the u201CStrong IQ Hypothesis,u201D namely that IQ accurately reflects intelligence, that IQ is overwhelmingly determined by genetics, and that IQ is subject to little or no significant cultural or economic influence after we adjust for the universal Flynn Effect. Since the IQ disparities discussed above seem to provide a powerful challenge to this theory, their validity has sometimes been disputed on the grounds that the populations being compared might actually be more dissimilar than we realize due to the impact of selective migration.
For example, one might speculate that the smarter Irish immigrated to America, while their dimmer relatives remained at home, and the same was also true for the smarter Southern Italians, Greeks, or other Balkan Europeans. Similarly, perhaps the smarter European Jews crossed the oceans to New York Harbor in the years before World War I, while their dimmer relatives stayed behind and later moved to Israel after World War II.
These explanations seem quite unlikely. The intra-ethnic IQ gaps being discussed are absolutely enormous often approaching a full standard deviation or more and that would imply a similarly enormous gap between the portions of the population that stayed and those that emigrated, with no contemporaneous source seeming to provide any indication of this. Indeed, during the period when these immigrant flows were occurring, most American observers emphasized the remarkable backwardness of the new arrivals and often speculated that they were intrinsically defective and might constitute a permanent burden to society. If anything, it was sometimes suggested that they were less intelligent than their stay-at-home co-ethnics and had come to America because they were unable to compete at home, hence their description as the so-called u201Cwretched refuse from a teeming shore.u201D
The limited ethnic IQ data we have from that period support this impression. In his 1978 book American Ethnic Groups, Thomas Sowell included a chapter that summarized the 1920s data on the average IQ scores of various Eastern and Southern European immigrant groups and showed that these were generally quite low, with Slovaks at 85.6, Greeks at 83, Poles at 85, Spaniards at 78, and Italians ranging between 78 and 85 in different studies. A separate analysis of the aptitude scores of World War I draftees published in 1923 came to similar conclusions. These published IQ studies by prominent academics led to widespread belief that the more recent European immigrant groups were much less intelligent than earlier ones and might drag down the national average, a belief that may have contributed to passage of the highly restrictive 1924 Immigration Act.
Even if we ignore all contemporaneous evidence and argue that 19th century European immigrants to America and elsewhere somehow constituted the IQ elite of their originating countries, the theory of selective migration still remains implausible. It has long been established on both theoretical and empirical grounds that IQ scores generally follow a mean-reversion pattern, in which the children of outlying individuals tend to regress toward the typical levels of their larger population or ethnic group. So even if we hypothesize that the Irish, South Italians, Jews, and Greeks who immigrated to America constituted the smartest small slice of their generation rather than, as seems more likely, often the poorer and most miserable roughly half their relative IQ advantage would have dissipated after a single generation. Thus, the apparent one standard deviation gap between American Irish and Ireland Irish a few decades ago would have required an initial gap of something closer to two standard deviations at the time the immigration occurred, a difference so large as to be totally implausible.
Furthermore, the most recent 2009 PISA international student academic tests sponsored by the Organisation for Economic Co-operation and Development provide us with results that raise further doubts about the correctness of the Lynn/Vanhanen IQ scores from a wide range of European countries. For example, although Croatia and Austria are geographically quite close, Croatians had IQs 12 points lower when their country was desperately poor just after World War II; yet today their overall PISA scores are not enormously lower, and are actually higher in reading, even though Croatia's average income is still lower by a factor of two. During the early 1970s, a huge national sample had placed the Ireland IQ at 87, the lowest in all of Europe, but today Ireland's PISA scores are about average for the continent and roughly the same as those for France and Britain, while Irish per capita incomes have pulled a little ahead.