1. Background

Earl B. Hunt writes that economists traditionally view differences in wealth between nations in terms of human capital, which is a general term for the abilities of the workforce. However, some researchers have argued that differences in average intelligence between nations also play a role. Richard Lynn and Tatu Vanhanen published the books IQ and the Wealth of Nations and IQ and Global Inequality, which led to further investigations from other researchers, some of them highly critical of Lynn and Vanhanen's methods and conclusions.^[1]

According to Hunt, international studies of intelligence are important because they measure which populations possess the cognitive skills that are necessary in a post-industrial world. He also writes that genetics cannot be ruled out as a possible cause, but that education surely plays a major role, so one should not conclude that human capital in poor countries can never be improved.^[2]

2. Studies of National Cognitive Ability

2.1. "Average IQ vValues in vVarious European cCountries"

The 1981 article "Average IQ values in various European countries" by Vinko Buj is the only international IQ study that over a short time period has compared IQs using the same IQ test. It was probably done in the 1970s in the capital cities or in the biggest town in 21 European countries and Ghana. Rindermann (2007) states that it is of dubious quality with scant information regarding how it was done. The correlations with the other measures of national intelligence, except the PISA student assessment study, are good.^[3][4]

2.2. Lynn and Vanhanen

2006 estimates by Lynn and Vanhanen. By Olivello - This vector image includes elements that have been taken or adapted from this file:, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=84115639

In the 2002 book IQ and the Wealth of Nations, and IQ and Global Inequality in 2006, Richard Lynn and Tatu Vanhanen created estimates of average IQs for 113 nations. They estimated IQs of 79 other nations based on neighboring nations or by other methods. They also created an estimate of "quality of human conditions" for each nation based on gross national product per capita, adult literacy rate, fraction of the population to enroll in secondary education, life expectancy, and rate of democratization. Lynn and Vanhanen found a substantial correlation between the national IQ scores they created and these various socioeconomic factors. They conclude that national IQ influences these measures of well-being, and that national differences in IQ are heavily influenced by genetics, although they also allow for some environmental contributions to it. They regard nutrition as the most important environmental factor, and education a secondary factor.^[5]

Several negative reviews of the book have been published in the scholarly literature. Susan Barnett and Wendy Williams wrote that "we see an edifice built on layer upon layer of arbitrary assumptions and selective data manipulation. The data on which the entire book is based are of questionable validity and are used in ways that cannot be justified." They also wrote that cross country comparisons are "virtually meaningless".^[6]

2.3. Wicherts, Dolan and vVan dDer Maas' aAnalysis

In 2009 Jelte M. Wicherts, Conor V. Dolan, and Han L.J. van der Maas conducted a new analysis of IQ in sub-Saharan Africa, which was critical of many of Lynn and Vanhanen's methods.^[7] Wicherts et al. concluded that Lynn and Vanhanen had relied on unsystematic methodology by failing to publish their criteria for including or excluding studies. They found that Lynn and Vanhanen's exclusion of studies had depressed their IQ estimate for sub-Saharan Africa, and that including studies excluded in "IQ and Global Inequality" resulted in average IQ of 82 for sub-Saharan Africa, lower than the average in Western countries, but higher than Lynn and Vanhanen's estimate of 67. Wicherts at al. conclude that this difference is likely due to sub-Saharan Africa having limited access to modern advances in education, nutrition and health care.^[8]

2.4. International sStudent aAssessment sStudies

Rindermann (2007) states that the correlations between international student assessment studies and measures of national IQ are very high. Using the same statistical method used to measure the general intelligence factor (g) he finds evidence for that the "student achievement assessments and intelligence tests primarily measure a common cognitive ability". The international student assessment studies have the advantages of standardized testing over a short time period. A disadvantage is that unlike IQ-data collections, it does not include older people or more developing nations.^[3] An advantage of using international student assessments instead of educational assessment is that "across countries and time, educational degrees are difficult to compare".^[9] Compared to international student assessments, "literacy as the ability to read and write texts is a much too basic competence" and as a result "these three educational measures [literacy rate, years spent attending school, and highest achieved degree] usually show lower correlations compared to more complex ability measures".^[9]

The international student assessment studies, the TIMSS, PIRLS, and PISA, are highly correlated with each other: "between TIMSS and PIRLS: r = .94 (N = 54), between TIMSS and PISA: r = .89 (N = 58), between PIRLS and PISA: r = .82 (N = 49)".^[10]

Rindermann's analysis found many of the same groupings and correlations found by Lynn and Vanhanen, with the lowest scores in sub-Saharan Africa, and a correlation of .60 between cognitive skill and GDP per capita. According to Hunt, due to there being far more data available, Rindermann's analysis was more reliable than those by Lynn and Vanhanen. By measuring the relationship between educational data and social well-being over time, this study also performed a causal analysis, finding that a nation's investment in education leads to increased well-being later on.^[11]

In 2013, Rindermann compared the results of three culture-reduced IQ tests collected in Tanzania in 1999 and 2000 (APM, MRT, LPS), WISC-IV working memory and verbal comprehension scale tests of blind, visually handicapped and non-blind multiethnic students aged 10 to 16 years in South Africa, and student assessment studies conducted in 14 African countries between 1964 and 2009 (primarily TIMSS, PIRLS and SACMEQ) with his own regression analysis, which used the Human Development Index and skin brightness as, respectively, potential nurture-based and nature-based predictors of cognitive ability. After adjusting for the Flynn effect and using 2010 estimates as the baseline, his predicted IQ for the African majority nation samples varied between 68 and 78, with an average IQ of around 75. This was similar to Rindermann and Te Nijenhuis (2012)'s average IQ estimate for Bali in Southeast Asia and other developing regions. According to Rindermann, the resulting IQ estimates are predicated on a number of contributing factors, including properly administered tests, the degree to which testing instructions are understood, sample bias, school enrollment rates, mean annual IQ grow at school and per age year, a higher Flynn effect among African samples, age correction, and greater or lesser familiarity with testing norms.^[12]

2.5. Other iIndicators of cCognitive aAbility

Rindermann (2018) uses a variety of other cognitive measures for a number of purposes: to examine cognitive ability in past time periods and to provide additional support for traditional ability measures such as IQ testing and student assessments.^[13] The more qualitative criteria discussed include:

1. historical measures of behavioral irrationality, since intelligence and rationality are positively correlated at r=.70.^[14] Examples of this irrationality include magical thinking, anthropomorphous thinking, and excessive use of cruelty
2. civilizational criteria such as those proposed by anthropologist John Baker, which Rindermann criticizes as "Some criteria are doubtful...and many are very basic and show only loose connection to cognitive ability"^[15] though other criteria such as the use and sophistication of cartography he views as more "cognitively informative".
3. the efficiency of time use and the use of more or less accurate timekeeping

3. Limitations and Criticisms of the Data Sets

3.1. IQ-dData cCollections

Rindermann (2007) writes that the mixture of many different tests and the not always clear representativeness of the samples seem to be the most serious problems. Furthermore, the measurement years vary, which is problematic because of the Flynn effect. Using the same adjustment for all nations is likely sometimes incorrect because since the 1970s developing nations have seen higher increases than the developed world. The method of averaging neighboring countries for an estimation for the many nations that did not have measured IQs, while having a high correlation (0.92) with the measured results in the case of the 32 nations that changed from the estimated to the measured categories between the two books, is likely problematic because some research indicates that absence of IQ tests indicates conditions such as poverty or war that may affect IQs. "In addition, some errors in the data have been observed".^[3]

As noted above, the article "A systematic literature review of the average IQ of sub-Saharan Africans" (2009) argued that a number of studies showing higher IQ values for sub-Saharan Africa had been excluded by "IQ and Global Inequality". Regarding four studies comparing and finding agreement between Lynn's estimated national IQs and the student assessment tests, they disagree regarding sub-Saharan Africa but write "these four studies appear to validate national IQs in other parts of the world."^[8] Wicherts and colleagues (2010)made several examinations of unrepresentativeness and stated: "In light of all the available IQ data of over 37,000 African testtakers, only the use of unsystematic methods to exclude the vast majority of data could result in a mean IQ close to 70. On the basis of sound methods, the average IQ remains close to 80."^[16] Consequently, some later studies using IQ data have checked their results against data from both sources.^[17][18]

The claim that the tests are culturally neutral and unbiased has been criticized.^[19][20][21]

3.2. International sStudent aAssessment sStudies

Rindermann (2007) writes that data from many developing nations are missing which is the case for more nations than for IQ data. The Flynn effect has to be adjusted for. In some nations school attendance is low. Even for the same test national organizers sometimes differ in implementation and exclusion rates differ.^[3]

4. Correlations with National IQ

Hunt writes that despite the limitations of Lynn and Vanhanen's data, their conclusions about the correlation between IQ and measures of social well-being probably are correct. This is partially driven by large differences in prosperity and intelligence test scores between regions of the world. The countries with the top highest tested IQ scores are the industrialized developed East Asian countries of Taiwan (104), Singapore (108), Japan (105), South Korea (106) and China (105), followed by North American and European countries such as the United States (98) and United Kingdom (100). Intermediate scores are found in South American countries such as Venezuela (84), Peru (85), and Honduras (81), as well as in South Asian countries such as Pakistan (84), India (82), and Bangladesh (82) and in Middle Eastern and North African countries such as Iran (84), Egypt (81) and Lebanon (82). The lowest recorded IQ scores are found in sub-Saharan African countries such as Somalia (68), Sudan (71), and Ethiopia (69). ^[22] However, substantial correlations between intelligence test scores and measures of well-being also exist when the analysis is limited to developed countries, where the IQ results are more likely to be accurate.^[1]

Hunt and Wittman (2008) conclude that although the correlation between national IQ and economic well-being is clear, the direction of causality between them is more difficult to determine. They suggest some methods which could be used to determine the direction of causality in future studies.^[23] A 2013 regression analysis by Gregory B. Christainsen examined the question of how much national IQ influenced national wealth, and how much the causality occurred in the opposite direction. This analysis concluded that the origin regions of the participants primarily influenced national wealth, rather than the reverse.^[24]

5. Causes of National Differences

Since the 20th century, there have been worldwide continual increases in measured IQ. This rise has been correlated with degrees of rising education levels, and as such may provide a partial explanation for observed differences in average IQ scores between nations. One recent study suggests that the Flynn effect has ended in some developed nations such as Denmark and Norway , and suggests that this is due to immigration from countries with lower education levels or due to changes in education policies within the countries.^[25] Wicherts et al. have suggested that national differences in IQ could be because African countries have not yet experienced the improvements that cause the Flynn effect in the developed world, such as improvements in nutrition and health, and educational attainment.^[8]

Controversially, the two books argued for a large genetic explanation. Such a role of genetics may or may not be related to race which is itself a controversial topic. Lynn argued further for this in the books Race Differences in Intelligence: An Evolutionary Analysis (2006) and The Global Bell Curve: Race, IQ, and Inequality Worldwide (2008).

Kanazawa (2008) also argues for a genetic role. He writes that cold climate and harsh winters (the study uses mean annual temperature) as well as environment novelty (the study uses three different measures of distance from the ancestral environment in sub-Saharan Africa: ordinary distance and differences in latitudes and longitudes) have been proposed as important factors behind the genetic evolution of human intelligence. The study found independent support for both theories and argues that they together explain half to two-thirds of variance in national IQ.^[26] This study has been criticized by Wicherts et al. (2012) for multiple alleged flaws, including assuming that the Earth is flat, and that humans migrated in the most direct possible route, even if that route involved crossing the ocean. They conclude that these flaws "...should have precluded publication of the paper as constituted at the time of review."^[27]

In contrast, Wicherts, Borsboom, and Dolan (2010) criticized this and some other evolutionary studies for problems such as ignoring or assuming that the Flynn effect is equal worldwide and assuming that there have been no migrations and changes in climate over the course of evolution. "In addition, we show that national IQs are strongly confounded with the current developmental status of countries. National IQs correlate with all the variables that have been suggested to have caused the Flynn Effect in the developed world."^[28]

Eppig, Fincher, and Thornhill (2010) states that distance from Africa, temperature, and most importantly by a large margin, prevalence of infectious disease predict national IQs. Education, literacy, GDP, and nutrition were not important as independent factors (however, the prevalence of infectious diseases is likely greatly affected by these factors). The authors argue that "From an energetics standpoint, a developing human will have difficulty building a brain and fighting off infectious diseases at the same time, as both are very metabolically costly tasks" and that "the Flynn effect may be caused in part by the decrease in the intensity of infectious diseases as nations develop."^[29]

Becker and Rindermann (2016) found a moderate correlation between genetic distances, particularly with respect to geographic latitude, and cross-national IQ differences, even when controlled for factors such as Human Development Index, indicating an evolutionary explanation.^[30] Woodley of Menie et al. (2016) found that the frequencies of eight SNPs predictive of IQ correlated negatively with national fertility, and furthermore that the SNP meta-gene negatively predicts national fertility when HDI is controlled, "suggesting the existence of a selection pressure" that may reduce global g by 0.253 points per decade.^[31]