Are development experts becoming racists?
BY CHARLES KENNY | APRIL 30, 2012
Columnist John Derbyshire's recent effluvia on the subject of things your white kid should know about black people was met with suitable disdain and a rapid expulsion from the web pages of the National Review. Genetic determinism with regard to racial intelligence -- alongside the very idea that intelligence can be meaningfully ranked on a single linear scale of intrinsic worth -- has been firmly debunked by Stephen Jay Gould, among others.
Off to a good start there! Invoking the supreme authority of the late Stephen Jay Gould is a surefire way to persuade anybody familiar with the field of psychometrics that you know what you are talking about.
Sadly, Derbyshire-like prattishness on the intellectual inferiority of dark-skinned races and its impact on social and economic outcomes in the United States has a historied international equivalent. In fact, if anything, the academic consensus on why some countries are rich and others are poor is tacking closer to the shoals of genetic determinism than it has been since the days of high empire. Derbyshire's deserved disgrace is a needed reminder to throw brickbats at his partners in malodor who work in global development.
... Development economists over the past 50 years have eschewed genetic explanations for the wealth and poverty of nations, favoring factors from lack of investment to lack of health care and education to wrong policies to poor government institutions. But the mainstream is moving back in the direction of "deep causes" of development. These involve determinants such as the relative technological advance of regions some centuries (even millennia) ago or levels of ethnic diversity that have long historical roots. And Enrico Spolaore and Romain Wacziarg have gone even further back, arguing that "genetic distance" -- or the time since populations shared a common ancestor -- has a considerable role to play in the inequality of incomes worldwide. They estimate that variation in genetic distance may account for about 20 percent of the variation in income across countries.
Spolaore and Wacziarg take pains to avoid suggesting that one line of genetic inheritance is superior to another, preferring instead an interpretation that argues genetic distance is related to cultural differences -- and thus a more complex diffusion of ideas: "the results are consistent with the view that the diffusion of technology, institutions and norms of behavior conducive to higher incomes, is affected by differences in vertically transmitted characteristics associated with genealogical relatedness.… these differences may stem in substantial part from cultural (rather than purely genetic) transmission of characteristics across generations," they write.
But where Spolaore and Wacziarg are careful enough to step away from interpretations based on the superiority of certain allele types, more foolhardy scholars have been happy to jump in. Take the book by Richard Lynn and Tatu Vanhanen titled IQ and the Wealth of Nations. It suggests that the average IQ in Africa is around 70 points, compared with much higher averages in East Asia and the West. Based on their data, the authors suggest that higher average IQ scores are the cause of progress in measures of development, including income, literacy, life expectancy, and democratization. Lynn and Vanhanen even argue that IQ was correlated with incomes as far back as 1820 -- a neat trick given that the IQ test wasn't invented until a century later.
As that surprising finding might suggest, most of Lynn and Vanhanen's data is, in fact, made up. Of the 185 countries in their study, actual IQ estimates are available for only 81. The rest are "estimated" from neighboring countries.
Lynn and Vanhanen can do their analyses based either on "only" 81 countries or using estimates of neighboring countries. In either case, they get virtually identical correlation coefficients, suggesting the robustness of their approach.
But even where there is data, it would be a stretch to call it high quality. A test of only 50 children ages 13 to 16 in Colombia and another of only 48 children ages 10 to 14 in Equatorial Guinea, for example, make it into their "nationally representative" dataset.
The correlations look slightly stronger if you throw out Equatorial Guinea. Anyway, all this data has been updated by Rindermann.
Psychologist Jelte Wicherts at the University of Amsterdam and colleagues trawled through Lynn and Vanhanen's data on Africa. They found once again that few of the recorded tests even attempted to be nationally representative (looking at "Zulus in primary schools near Durban" for example), that the data set excluded a number of studies that pointed to higher average IQs, and that some studies included dated as far back as 1948 and involved as few as 17 people.
Wicherts and his colleagues also point out that there is considerable evidence the tests Lynn and Vanhanen use to make their case "lack validity in test-takers without formal schooling." It is, surely, hard to take a multiple-choice test when you don't know how to read. Not surprisingly, IQ test results in Africa are weakly aligned to other measures of intelligence that don't require written test-taking.
Right. As I pointed out in my VDARE.com review in 2002, Lynn and Vanhanen's finding of an average IQ of 70 in black Africa is strong evidence in favor of the nurture position that a better environment can raise IQs, because African Americans, who appear to be about 4/5th black, score 15 points higher. (Lynn subsequently adopted the logic of my critique.) So, Wicherts' finding that if you only count the IQ tests that he likes, on which black Africans average around 80 or a little higher, then that strengthens the hereditarian view. (This is much too subtle for Kenny to grasp, of course.)
On the other hand, there are reasons of predictive validity for including test scores where black Africans simply failed to grasp the point of using abstract logic to solve puzzles (typically, culture-free nonverbal ones, not "regatta" questions as Kenny implies). Long ago, Thomas Sowell recounted an anecdote where two 17-year-old African youths were asked a standard IQ test question. They wittily ridiculed the impracticality and absurdity of this highly abstract question, displaying quickness of mind in social cognition. On the other hand, as Sowell noted, if by the age of 17, your culture hasn't introduced you to abstract thought yet, you probably aren't going to pick it up very well as an adult, and you are probably not going to be highly productive in economic roles that demand that kind of nerdier thinking. Thus, the high correlation between low IQ scores in Africa and low per capita GDPs in Africa, even if some of low scores are due to lack of acculturation in modern thinking.
My guess is that the spread of cheap smartphones in Africa will stimulate the kind of black box logical thinking that IQ tests measure and which the modern economy rewards. As I pointed out in my review of James Flynn's 2007 book, the fascinating question is why IQ tests still possess so much predictive power more than a century after being invented.
Moreover, there are still some low-hanging fruits where 3rd World countries would benefit from public health programs that succeeded in the U.S. in the first half of the 20th Century in boosting IQ directly or or in boosting mental energy. Fortifying salt with iodine eliminated the medical syndrome cretinism. while fortifying wheat with iron also eliminated an IQ-sapping medical condition. The Rockefeller Foundation's war on hookwarm greatly benefited the physical and economic energy of Southerners by ridding them of a parasite.
Kiwanis International is the leading charity in salt iodization in poor countries. As you can see, these are not fashionable causes, but Bjorn Lomborg has long identified them as high bang for the buck development projects, as I pointed out in a 2004 VDARE.com essay.
I've been writing about the need for more micronutrient fortification to boost Third World IQ scores for over eight years, but practically nobody else will touch the subject because the topic of low average IQ scores in much of the Third World is off-limits.
Wicherts also points out international evidence that average IQs can rise dramatically over time -- by as much as 20 points in the Netherlands between 1952 and 1982, for example. In fact, Africa's current estimated "average IQ" is about the same as Britain's in 1948. The phenomenon of rising average IQ scores over time is known as the "Flynn effect," named after political scientist Jim Flynn, who popularized the result. It suggests that factors such as improved nutrition, health care, and schooling may all improve IQ test performance. Of course, Africa is currently behind richer regions on such factors, though it is rapidly catching up. Indeed, the Flynn effect may have added as much as 26 points to estimates of Kenyan IQ over a recent 14-year period. That's more than the gap between reported IQs in Africa and the United States estimated by Wicherts and colleagues based on samples from 1948 to 2006. In short, all of the evidence suggests lower levels of development cause lower test scores -- not the other way around.
But lower test scores also lead to lower development. For example, Singapore and Lagos are at the same latitude and altitude, but the high-IQ Chinese of Singapore have rid themselves of many IQ and energy sapping tropical maladies through well-conceived and well-executed public health programs. No doubt, the people of Lagos would benefit cognitively from better health, too, but it's hard to get the cycle started.
From Heiner Rindermann's new paper on whether IQ causes wealth or vice-versa:
Rindermann, H. (2012).
Intellectual classes, technological progress and economic development: The rise of cognitive capitalism.
Personality and Individual Differences, 53(2), 108-113.
Cognitive ability theory claims that peoples’ competences are decisive for economic wealth. For a large number of countries Lynn and Vanhanen (2002) have published data on mean intelligence levels and compared them to wealth and productivity indicators. The correlation between intelligence and wealth was supported by studies done by different authors using different countries and controls. Based on their pioneering research two research questions were developed: Does intelligence lead to wealth or does wealth lead to intelligence or are other determinants involved? If a nation’s intelligence increases wealth, how does intelligence achieve this? To answer them we need longitudinal studies and theoretical attempts, investigating cognitive ability effects at the levels of individuals, institutions and societies and examining factors which lie between intelligence and growth. Two studies, using a cross-lagged panel design or latent variables and measuring economic liberty, shares of intellectual classes and indicators of scientific-technological accomplishment, show that cognitive ability leads to higher wealth and that for this process the achievement of high ability groups is important, stimulating growth through scientific-technological progress and by influencing the quality of economic institutions. In modernity, wealth depends on cognitive resources enabling the evolution of cognitive capitalism.
Yes, it seems logical, as Kenny argues, that countries with high average IQs would suffer more from diminishing marginal returns. Yet, despite all the handwaving about the Flynn Effect, nobody has yet come up with much large-scale evidence for convergence.
One possibility is that the value of a strong back on the global market is in decline faster than the value of a high IQ.
Convergence is what everybody assumes will happen, but what actually seems to be happening is that East Asians have begun to pull away from the rest of the world. When I plotted Lynn's IQ data for the whole 20th Century in 2004, the main trend visible was rising East Asian scores relative to everybody else. The unreleased 2009 PISA scores from Chinese and Indian regions appear to show even poor, rural Chinese districts scoring in the same ballpark as European countries, while Indian states are scoring very badly, barely above SubSaharan levels. On the American SAT test, Asians (including, this time, South Asians) have been pulling away from everybody else over the last decade.
This is not to say that convergence won't happen at some point, but that there is remarkably little evidence for it so far.
There is a simple explanation for why the IQs of the offspring of colonists appear higher than those of the first descendants of the colonized. It's because the colonizers acted much as Thomas Carlyle's writing suggested they would -- as overlords with little or no interest in providing public services like a decent education or health care to a native population viewed with disdain. This left local populations malnourished, in poor health, and ill-educated -- if they were lucky enough to be in school at all.
The good news is that decolonization began a process of leveling the playing field, with rapidly climbing and converging indicators of health and education worldwide. Thanks to the Flynn effect, IQs are doubtless on a path of convergence as well, and the poisonous idiocy of genetic explanations for wealth and poverty will soon lose what little empirical support they might appear to have today.