The Intelligence of Nations
Richard Lynn and David Becker
London: Ulster Institute for Social Research, 2019
British psychologist Richard Lynn and Finnish political scientist Tatu Vanhanen published their first study of national IQ differences, Intelligence and the Wealth of Nations, in 2002. The book found that a nation’s average IQ correlated at .62 with its per capita income, meaning that IQ explained 38% of variation in wealth (0.38 = 0.62 squared). In addition to the expected opposition from egalitarians, some critics’ skepticism was aroused by shortcomings in the book’s empirical data: the authors had direct IQ measurements for only 81 countries; values for the other 104 were mere estimates based on neighboring countries.
Plenty of international IQ data has accumulated since then, and the authors updated their results first in 2006 (IQ and Global Inequality), and again in 2012 (Intelligence: A Unifying Construct for the Social Sciences, hereafter Lynn & Vanhanen 2012; reviewed in TOQ 12:2, Summer 2012). In this last work, they provided direct data for 161 countries; the correlation between IQ and income rose to .71. Some of the earlier skeptics were won over.
Tatu Vanhanen died in 2015, but progress in the measurement of national IQ has continued. Heiner Rindermann of Chemnitz Technical University in Germany made an important contribution with his book Cognitive Capitalism: Human Capital and the Wellbeing of Nations (2018). Prof. Rindermann found a correlation of .82 between national IQ and per capita income, meaning that intelligence explains no less than two thirds of the variation between nations. He found positive correlations between IQ and many desirable social variables, boldly concluding that “national well-being mainly depends on the cognitive ability level of a society.” Rindermann also demonstrated that the cognitive ability of the most intelligent five percent of a population generally has a greater positive effect on national achievements and other desirable outcomes that the average intelligence of the overall population.
In this new book, Richard Lynn’s first since the loss of his Finnish coauthor, he teams up with David Becker, a younger colleague of Prof. Rindermann in Chemnitz. The authors explain:
The main difference between this study and the previously published studies is with regard to the level of detail provided. Previous studies have been criticised for drawing upon unrepresentative, small or incomparable samples with regard to particular nations. In updating these studies, we endeavour to obviate this problem, so that; this information is as clear as possible to other researchers. The central aim of this revision is to standardize each individual step of the process through which we have reached our estimations and made our calculations. This has the advantage of complete transparency, such that other researchers can refollow our steps should they wish to do.
This aim necessitates a highly technical presentation less adapted to the needs of the general reader than Lynn’s and Vanhanen’s books.
Also new to this study is use of the “National IQ Dataset,” a working file first uploaded to the internet in August, 2018 and continually updated to provide the best current data on IQ around the world. The most recent version can be viewed at http://viewoniq.org/.
Stricter criteria have been established for inclusion in this dataset than were used by Lynn and Vanhanen. Oddly, this means that the new national IQ scores may sometimes be less accurate than those reported in Lynn & Vanhanen 2012, since some of the data they used did not make the new cut and better data has not yet become available. This especially effects the IQ estimates for low-scoring countries.
Thus, while the global pattern of IQ-distribution in the new book is similar to Lynn & Vanhanen 2012 with a mean deviation of just 2.44 points, no fewer than 14 countries now score under 60 and three score below 50. By comparison, the lowest national IQ reported in Lynn & Vanhanen 2012 was 60.1. The extreme case is Nepal, estimated at 78 in Lynn & Vanhanen 2012, but down by 35 points to 43 in the new study! This is clearly implausible, since on a priori grounds we would expect Nepal to score similarly to neighboring India (76.24 according to Lynn and Becker).
Part of the explanation is that
IQ tests are mostly designed to reflect the normal range of a population with an IQ similar to Britain. Many developing countries have mean scores close to or even within the percentiles where tests lose their reliability.
Raven’s Progressive Matrices are considered the best test for measuring low IQs, but they have their own peculiar disadvantage: the formulas that convert Raven’s raw scores to IQ scores, explain Lynn and Becker, cause a “lengthening of the variance-observed ranges of the Raven’s Matrices scales”—the estimates of the variation in IQ within a given country become greater.
For now, the authors propose a bottom cut-off of 60 for the scale of national IQ. When measuring correlations with national IQ, Lynn and Becker give figures both for their IQ list “as is” and with the proposed cut-off; the figures are usually only slightly different. A future version of the National IQ Dataset will provide a stronger focus on the problem of low IQs, with a flexible and customizable lower IQ-limit—as well as incorporating new and better data. BENIN http://viewoniq.org/
Data for Djibouti, Benin, and Sudan compared to British norms. From viewoniq.org.
The emphasis of the present study, however, is on rigor and transparency of method, so let us turn to the authors’ procedures.
Data samples used for the National IQ Dataset vary according to the number of persons tested, their ages and educational levels, whether they are representative of their nation as a whole or skew rural or urban, high or low socioeconomic status, and so forth. Lynn and Becker have devised an approximate index of sample quality that takes all these matters into account; it is called a “sample rating,” and ranges from 0 to 1. For well-studied advanced nations, sample ratings may be .9 or higher, while for some third world countries researchers might have to make do with sample ratings of .4 or lower.
Intelligence tests in current use include Sanford-Binet, Raven’s Progressive Matrices, the Cattell Culture Fair Test, several varieties of the Wechsler Scales, and a few others. To some samples, only parts of a test battery may be administered. The authors devise a “testing rating” analogous to the sample rating to give an approximate idea of the quality of the test or tests administered for each sample. They even have a “method rating” to assess how rigorously the raw test scores were converted to IQs.
The so-called Flynn effect—the gradual rise in IQ scores over much of the twentieth century—can produce “an overestimation of IQs if norms employed are older than the year of measurement, or an underestimation of scores from a measurement in an earlier year were converted to IQs by using more recent norms”’ this “needs to be compensated for in order to make IQ scores from different times comparable.” To complicate matters, the Flynn effect is not constant across nations and periods of time. Researchers Pietschnig and Voracek (2015) have assembled data on these variances for 31 countries. They found, e.g., that IQ scores in Germany improved at a rate of 0.60 of a point per year between 1956 and 2008, while those in Britain improved at a rate of only 0.11 per year over a somewhat longer period (1932–2008). With the help of these data and some third-degree polynomial formulas, Lynn and Becker have devised an original way of correcting for the Flynn effect.
It is common to norm IQs to a British average score of 100, but certain tests may be normed on the basis of other countries, viz., Canada, China, Germany, Spain, Finland, France, Israel, the Netherlands, or the USA. Accordingly, the authors provide formulas for rendering the results of differently normed tests comparable.
For six ethnically diverse countries—Brazil, Kazakhstan, Mexico, Serbia, the USA and South Africa—data was available for specific racial sub-groups; for use in calculating national averages, of course, this data had to be weighted according to the percentage of the group in the total population.
In addition to direct IQ testing, valuable information about intelligence can be derived from international school assessment studies such as the Programme for International Student Assessment (PISA), the Trends in International Mathematics and Science Study (TIMSS), and the Progress in International Reading Literacy Study (PIRLS). The authors converted results from these tests into what they call School Assessment Study IQs, or SAS-IQs, and list the results on pp. 167–172.
The final national IQs are calculated from a combination of the SAS-IQ scores (where available) and the weighted results of direct intelligence testing. Results are listed in the penultimate column of Table 16 on pp. 173–177 and mapped in Figure 3 on p. 179.
Although the technical apparatus of Lynn and Becker’s new study will in places be formidable for the general reader, it will make it far more difficult for critics to claim, as they did for earlier books by Lynn and Vanhanen, that the results are “highly deficient” or “technically inadequate… and meaningless.”
The third of Lynn and Becker’s chapters is devoted to cataloging the results of hundreds of studies measuring correlations between national IQ and various other social measures. They organize this information under 25 headings that include: educational attainment, per capita income, economic growth, cognitive achievement, political institutions, happiness and life satisfaction, religious belief, health, nutrition, crime, fertility, and climate. Here we can offer only a few highlights:
- National IQ correlates at .97 with educational attainment as measured by the PISA and TIMSS tests administered at age 15. The correlation with educational expenditure, on the other hand, is a negligible .19.
- Researchers have found correlations between IQ and per capita income ranging from .24 to .81 in recent years, with most estimates falling in the .60s and .70s. Use of a logarithmic scale for income increases the correlation. One researcher has estimated that a gain of 10 IQ points roughly doubles a nation’s per capita income.
- This implies a correlation between IQ and economic growth at some time in the past. Often, no significant correlation can be found over short periods of time, but they do show up in long-term studies, e.g., for the period 1950–2001 (.75), and 1500–2005 (.79).
- National IQ correlates .87 with the number of published academic papers per capita, .83 with innovation as assessed by the World Intellectual Property Organization .59 with the average numbers of books in a home, .55 with the number of patents, and .39 with Nobel Prizes in the sciences.
- Robust correlations of .44 to .56 have been found with democracy, between .49 and .61 with economic freedom, .60 and .64 with the rule of law, and .63 with property rights. National IQ correlates negatively with corruption at between -.54 and -.63.
- Correlations with crime are consistently negative, but study results vary widely: e.g., between -.25 and -.64 for homicide.
- Correlations with life expectancy range mostly between .74 and .85 across thirteen studies, plus one outlier at .51. Correlations with infant and child mortality across nine studies range from -.65 to -.79, plus one outlier of -.34. Correlations with malnutrition and nutritional deficiencies range between -.47 and -.77.
- Correlations with religiosity and belief in God are negative, ranging from -.48 to -.76.
- Correlations can even be measured for such apparent intangibles as happiness and life satisfaction. In their 2006 study, Lynn and Vanhanen found only a negligible correlation of .03 between the IQ and happiness of nations, but this may have been a fluke. Their 2012 study found a correlation of .64, and two other studies have found .60. Correlations of between .54 and .70 have been found with life satisfaction.
- Intelligence is associated with cold environments even today, with a nation’s IQ correlating negatively with its average temperature. But the most impressive inverse correlation is with a country’s estimated temperature during the Würm glaciation between 30,000 and 11,700 years ago: -.87. This supports the theory that higher intelligence evolved in the European and Northeast Asian peoples because of its survival benefits in colder latitudes during that most recent ice age.
- Correlations with fertility are negative, at between -.69 and -.80. This means, of course, that human intelligence is currently in decline.
Lynn and Becker’s final chapter considers possible ways of raising IQ. Intelligence can (apart from genetic intervention) be improved in three basic ways: through nutrition, health and education.
- Improvements to nutrition are probably responsible for the rise in phenotypic IQ over the twentieth Century known as the Flynn effect, but there may be little room for further such improvement in developed nations. The use of nutritional supplements could be of measurable benefit in poorer countries, however.
- Diseases still common in the developing world and the practice of cousin marriage have a harmful effect on phenotypic intelligence. But even in wealthier nations further health improvements are possible if women can be convinced to abstain from smoking and drinking during pregnancy.
- Educational improvements on the Caribbean island of Dominica are said to have raised intelligence there at a rate of 5.14 IQ points a decade over a period of 35 years. With education, as with nutrition and health, the greatest improvements are likely to be possible in developing countries. It is reported, e.g., that in Libya children aged 6 years have an average IQ of 98 (relative to British norms), but that the figure declines to 74.25 among 11-year-olds; similar results have been found for Syria and the United Arab Emirates. The reason is likely to be that education in the Middle East consists largely of memorization, without developing pupils’ reasoning abilities.
- Genotypic intelligence can only be increased through positive or negative eugenic intervention. Positive eugenics refers to incentivizing larger family size among the more intelligent. This can be done through tax policy, using income as a proxy for intelligence. Negative eugenics refers to the discouragement of procreation among the less intelligent, e.g., by abolishing child allowances for those dependent on public support. Immigration policy could also be designed to attract the highly intelligent and keep out the unintelligent.
At least, one would think immigration policy might be reformed along such lines. But a strange thing happens in the last seven pages of The Intelligence of Nations: it becomes impossible to alter current trends. (Emphasis hereafter is mine.)
It is inevitable that throughout Western Europe the numbers of non-Europeans will increase as a result of immigration and their greater fertility, the continued arrival of asylum seekers, illegal entrants and marriages with Europeans.
Africa is experiencing a population boom, and “inevitably large numbers [of Africans] will seek a better life in Europe and many will succeed.” They will try to claim asylum, but most will be economic migrants. However, “most of those whose asylum claims are rejected will remain because it will be impossible to deport them.”
Donald Trump was elected President of the United States in part because he promised to build a wall to stop illegal immigration from across the Mexican border. Lynn and Becker, however, write:
It is doubtful whether such a wall, if it is ever built, would be effective because Hispanics could continue to enter the United States by tunneling under it, by boat or by air, including flying to Canada and crossing into the US. Further dysgenic immigration into the United States is unstoppable.
The same impossibility of keeping out immigrants holds for Europe, where “even if Spain closed its borders to migrants, they will find other ways to enter Europe.”
Australia and New Zealand are two Western countries not suffering the effects of mass low-IQ immigration. They too are doomed, however, because
their indigenous peoples with low IQs will continue to grow because of their high fertility. It will not be possible to reduce this and it will have some adverse effect on their national IQs.
The authors also wait until the last pages of their book to mention the devastating dysgenic consequences of female careerism, but finally acknowledge that
the principal problem is the large number of highly educated high IQ career women who remain childless. It is probably impossible to introduce policies to increase the fertility of these women who have been educated out of their reproductive function.
As elsewhere, the authors do not explain why harmful social policies cannot be altered. Such an impossibility did not exist, however, as recently as the nineteenth century. As I pointed out some years ago in a review of Allan Carlson’s Third Ways [TOQ 8:3, Fall 2008]:
In Britain between 1842 and 1914 “substantial gains in material standards were achieved by the working class, accompanied by the movement of women from wage-earning to domestic pursuits.” Similarly, in Belgium there was “a thorough transformation in the family life of workers between 1853 and 1891, based on a withdrawal of married women from the labor market and a dramatic rise in the real incomes of men.”
In the U.S., the 1924 immigration restriction law in the U.S, upheld over President Truman’s veto in 1952, effectively restricted immigration until the immigration law of 1965 was passed.
Today, however, we said to be the helpless playthings of an inscrutable destiny. Women are fated to continue foregoing motherhood until Western Civilization is entirely extinguished. After all, it’s the current year.
Imagine my surprise, then, when the authors, turning their attention to China, write:
There is also likely to be an increase of intelligence in China with a reduction in air pollution which at present is impairing intelligence in many towns and cities.
That’s right: rising levels of air pollution in China are neither “inevitable” nor “unstoppable.” The Chinese are not fated to continue mindlessly polluting the air they breathe until they are reduced to a nation of blithering idiots. They are going to make a 180° turn and reduce air pollution so that their IQs rise. In view of this amazing ability to intervene in their own destiny, it is hardly surprising that Lynn and Becker conclude their book with a prediction that China will overtake the West and “emerge as the world’s superpower in the second half of the twenty-first century.”
Of course, the underlying problem with this entire concluding section of The Intelligence of Nations is that all the authors’ professional expertise in psychometry furnishes them with no more ability to predict the future than my Aunt Fanny. Like most forecasters, Lynn and Becker merely project current trends into the future (with the unexplained exception of air pollution in China). The problem with this is that “current trends” never continue. For example, immigration could be stopped immediately if there was the political determination to do so. Anyone can show where we are presently headed, but that is not the same as knowing where we will be at any given time in the future. History largely consists of trend-altering events; one might even define “historical event” as an occurrence which interrupts a pre-existing trend.
A civilization’s normal response to invasion is to send out an army to prevent it. Invaders have generally known this, and not attempted invasion before carefully considering their own strength and arming themselves extensively. Today’s invasion is historically unique in being virtually unarmed and depending instead on the sympathy of the population being invaded. This ought to make it far easier to defeat, rather than “inevitable” and “unstoppable.” But for that, of course, Western Man would have to be willing to take an active part in his own history.