Elsevier

Intelligence

Volume 33, Issue 3, May–June 2005, Pages 273-284
Intelligence

Educational and ecological correlates of IQ: A cross-national investigation

https://doi.org/10.1016/j.intell.2005.01.001Get rights and content

Abstract

The new paradigm of evolutionary social science suggests that humans adjust rapidly to changing economic conditions, including cognitive changes in response to the economic significance of education. This research tested the predictions that cross-national differences in IQ scores would be positively correlated with education and negatively correlated with an agricultural way of life. Regression analysis found that much of the variance in IQ scores of 81 countries (derived from [Lynn, R., & Vanhanen, T. (2002). IQ and the wealth of nations. Westport, CT: Praeger]) was explained by enrollment in secondary education, illiteracy rates, and by the proportion of agricultural workers. Cross-national IQ scores were also related to low birth weights. These effects remained with national wealth, infant mortality, and geographic continent controlled (exception secondary education) and were largely due to variation within continents. Cross-national differences in IQ scores thus suggest that increasing cognitive demands in developed countries promote an adaptive increase in cognitive ability.

Introduction

There is much continuing scientific debate about the meaning and significance of IQ scores (Dickens & Flynn, 2001, Sternberg et al., 2001). One of the key problems is that while IQ scores are strongly genetically heritable, and related to various biological measures including brain size, and reaction times (Lynn & Vanhanen, 2002) they are also quite strongly affected by environmental change, as reflected, for example, in the pronounced secular changes in IQ scores during the 20th century known as the Flynn effect after a researcher who may not have discovered the phenomenon but has written extensively on it (Flynn, 1987). (Such effects are clearly environmental because of the implausibility of an alternate explanation in terms of genetic change). One critical such environmental change in respect to variation in IQ is education. Years at school accounts for approximately two-thirds of the variation in IQ scores. This association likely involves reciprocal causation given that more intelligent people opt for more extensive education, that level of schooling is highly heritable, and that educational interventions designed to raise intelligence generally produce modest effects (Ceci, 1991, Neisser et al., 1995). This paper investigated the cross-national association between education and IQ—as well as ecological differences underlying variation in education, specifically the importance of agriculture as an occupation.

The history of IQ tests is bound up with the problem of predicting educational success. Although modern IQ tests have been validated against “real-world” outcomes from earning capacity to law-abiding, such effects are either quite small, or are obtained in studies lacking appropriate controls for education, socioeconomic status, family structure, personality, and so forth. For example, intelligence tests account for 31% of the variation in overall job performance and are thus the best available predictor in personnel psychology (Schmidt & Hunter, 1998). On the other hand, the correlation between tests of cognitive ability and earning capacity may become trivially small when controls are included for family structure and other background variables (Devlin et al., 2002, Griffin & Ganderton, 1996, Hout, 2002).

Intelligence tests are useful at predicting both academic success and practical outcomes, such as job evaluations (Lynn & Vanhanen, 2002, Schmidt & Hunter, 1998, Sternberg et al., 2001). Success in some cognitively complex tasks, from street children running a business, to racing experts betting on horses, to career success among doctors, may be poorly predicted by intelligence test scores, however (Ceci & Liker, 1986, McManus et al., 2003, Sternberg et al., 2001), although such research has been criticized on methodological grounds (Detterman & Spry, 1988, Sternberg et al., 2001).

If IQ tests measure the cognitive capacity for schooling (as well as a more general capacity to solve problems and achieve economic success), then the fact that these capacities increase with economic development (the Flynn effect) is really no more surprising than the phenomenon of athletic ability increasing with training, at least if one assumes that modern societies are more cognitively demanding (Greenfield, 1998). Nor does it challenge the recognition of heritable individual differences in the capacity to benefit from education-propensities that may have little opportunity to be expressed in a non-literate environment.

The new research paradigm of evolutionary social science suggests that humans are capable of rapid (i.e., in one or two generations) psychological and behavioral adjustment to changing ecological conditions (Barber, submitted for publication). Children raised in poverty experience increased psychological stress that is now known to sculpt the developing brain, for example, and are more vulnerable to anxiety, depression, and impaired immune function throughout their lives as a result (Teicher, Andersen, Polcari, Anderson, & Navalta, 2002). Assuming that anxiety has a protective function in threatening environments, this phenomenon can be seen as a case of adaptive plasticity of the developing brain. Recent evidence suggests that the psychological stress of domestic violence also reduces children's IQ scores and academic performance (Delaney-Black et al., 2002, Koenen et al., 2003).

Increased academic aptitude as a consequence of environmental complexity and increased emphasis on education is another kind of adaptive plasticity during psychological development, in the sense that it can facilitate social success. From this perspective, one would expect intellectual capacity to change in response to the economic significance, and prevalence, of education in a country. This prediction is certainly not unique to evolutionary social science but there are distinct advantages to working within a natural science perspective—including parsimony, more direct relevance to neighboring sciences, breadth of generalization, and novelty of predictions—and these advantages apply even in the case of recent phenomena such as the learning of mathematics (Geary, 1995).

The economic gains of literacy are modest in traditional agricultural societies where farming practices are dictated by time-honored routines propagated via the oral tradition. Literacy is essential for the extensive record-keeping required by a market economy, however, and it is no accident that the rise of universal education accompanied the progressive increase in market economies facilitated by the Industrial Revolution (Justman & Gradstein, 1999). If education increases as economies move away from agriculture, and if educational opportunity increases cognitive abilities, then it is possible to make the following predictions about the cross-national distribution of IQ scores:

  • 1.

    IQ scores should be lower in countries where large numbers of people make their living from agriculture.

  • 2.

    IQ scores should be higher in countries having extended educational opportunities (here operationalized as enrollment in secondary school).

  • 3.

    IQ scores should be lower in countries where more of the population is illiterate.

Such predictions have never been tested in cross-national research due both to lack of suitable data and to skepticism about the cross-national validity of IQ tests (Sternberg et al., 2001; see Method section). The predictions were tested using the IQ data of Lynn and Vanhanen (2002) for 81 countries.

Section snippets

Sample of societies

The countries studied comprised the 81 nations for which IQ data were provided by Lynn and Vanhanen (2002). The average 1996 gross national product (GNP) of these countries was $2514 and this did not differ significantly from that of all 185 countries in the world with populations over 50,000: M=$2020, 95% confidence interval $1660–2847.

Dependent variable

The dependent variable was the IQ score associated with each nation. Lynn and Vanhanen (2002) derived their data from published sources, as described in detail

Results

Correlations among the dependent and independent variables are shown in Table 1 along with their means and SDs. It can be seen that national IQs were significantly negatively correlated with illiteracy, agricultural labor, infant mortality, and the incidence of low birth weights. They were significantly positively correlated with secondary education and GNP.

Regression analyses are shown in Table 2. The first analysis without regional controls found significant negative effects of agricultural

Discussion

All three of the predictions were strongly supported by the data. Countries where large numbers of people made their living from agriculture had lower IQ scores (prediction 1). IQ scores were higher in countries with extended education as indexed by secondary school enrollment (prediction 2). Countries with high levels of illiteracy also had reduced IQ scores (prediction 3). These findings suggest that cross-national differences in IQ scores may be largely determined by ecology, with

Acknowledgments

I am grateful to Earl Hunt and to anonymous reviewers for many detailed and helpful comments that made a substantive contribution to this paper.

References (43)

  • S.J. Ceci et al.

    A day at the races : A study of IQ, expertise, and cognitive complexity

    Journal of Experimental Psychology: General

    (1986)
  • T.C. Daley et al.

    IQ on the rise: The Flynn effect in rural Kenyan children

    Psychological Science

    (2003)
  • V. Delaney-Black et al.

    Violence exposure, trauma, and IQ and/or reading deficits among urban children

    Archives of Pediatric and Adolescent Medicine

    (2002)
  • D.K. Detterman et al.

    Is it smart to play the horses? Comment on “A day at the races: A study of IQ, expertise, and cognitive complexity” (Ceci and Liker, 1986)

    Journal of Experimental Psychology. General

    (1988)
  • B. Devlin et al.

    Intelligence and success: Is it all in the genes?

  • W.T. Dickens et al.

    Heritability estimates versus large environmental effects: The IQ paradox resolved

    Psychological Review

    (2001)
  • R. Fernandez-Ballesteros et al.

    Sociohistorical changes and intelligence gains

  • J.R. Flynn

    Massive IQ gains in 14 nations: What IQ tests really measure

    Psychological Bulletin

    (1987)
  • J.R. Flynn

    Asian Americans: Achievement beyond IQ

    (1991)
  • J.R. Flynn

    IQ gains, WISC subtests and fluid g: G theory and the relevance of Spearman's hypothesis to race

    Novartis Foundation Symposium

    (2000)
  • J. Fox

    Regression diagnostics

    (1991)
  • Cited by (72)

    • Neuropsychological, electrophysiological and neurological impairments in patients with obsessive compulsive disorder, their healthy siblings and healthy controls: Identifying potential endophenotype(s)

      2016, Psychiatry Research
      Citation Excerpt :

      All evaluations took almost 3–3.5 h for each participant. Patients were evaluated with DSM-IV, Structured Interview for DSM Axis I (SCID-I) for diagnosis of psychiatric disorders (Ettelt et al., 2007; de Vries et al., 2013) and Yale Brown Obsessive Compulsive Rating Scale (Y-BOCS) for scanning OCD symptoms and disease severity (Barber, 2005), Hamilton Depression Rating Scale (HDRS) was used for detecting comorbid depression (APA, 1994). Clinical Global Impression Scale (CGIS) for the severity of the disease (First et al., 2002) and Global Assessment of Functioning Scale (GAF) for the assessment of functioning (Goodman et al., 1989) were applied to patients.

    • Intelligence, human capital and HIV/AIDS: Fresh exploration

      2015, Intelligence
      Citation Excerpt :

      While the positive correlation between intelligence and health has been substantially documented in many studies, only some have tried to present causal analyses. Barber (2005) has linked it to the example of low birth weight; Lynn and Vanhanen (2006) to undernourishment and Meisenberg and Lynn (2012) to malnourishment. Many studies have also assessed the relationship with: (i) infant mortality (Barber, 2005; Lynn & Vanhanen, 2006; Kanazawa, 2006; Templer, 2008; Rushton & Templer, 2009; Reeve, 2009); (ii) maternal mortality (Lynn & Vanhanen, 2006; Reeve, 2009), life expectancy (Lynn & Vanhanen, 2006; Kanazawa, 2006; Lynn, Meisenberg, Mikk, & Williams, 2007; Ram, 2007; Templer, 2008; Rushton & Templer, 2009; Reeve, 2009) and (iii) HIV/AIDS1 (Templer, 2008; Rindermann, Sailer, & Thompson, 2009; Rushton & Templer, 2009; Reeve, 2009).

    • The burden of disease and the IQ of nations

      2013, Learning and Individual Differences
    View all citing articles on Scopus
    View full text