Does IQ Reflect A Person’s Socioeconomic Status?

Russell T. Warne

Jun 12, 2025

The SAT is supposed to measure aptitude, but what it actually measures is parental income, which it tracks quite closely. (Deresiewicz, 2014, paragraph 33)

Social scientists have long known that the best predictor of test scores is family income ... Standardized tests are best at measuring family income. (Ravitch, 2016, p. SR8)

Correlational Evidence

The evidence is clear: wealthier individuals tend to score higher on intelligence and academic tests. This is true, both in adulthood (Herrnstein & Murray, 1994) and childhood (Zwick, 2002). As a result, some people – like those quoted above – have argued that the tests of g are actually little more than tests of someone’s socioeconomic status. Others have argued that differences in wealth or socioeconomic status cause differences in performance on tests of g. In other words, they believe that money makes people smarter, or that it can buy higher scores through test preparation classes, better schools, or home life advantages (e.g., L. Brody, 2018; Zwick, 2002). As a result, some skeptics of intelligence testing believe that these tests reflect economic advantages more than any cognitive abilities.

Like many ideas debunked in this book, the idea that IQ scores (and other test scores) are determined by socioeconomic status is based on a grain of truth: the correlation between the two is positive. However, it is a weak relationship. In a study of nearly 1.3 million college-bound teenagers, the correlation between SAT scores and parental income was just r = .10; in a sample of almost 35,000 college students, the correlation between SAT scores and parental income was r = .23 (Camara, 2009). Slightly stronger is the correlation of r = .30 between parental income and IQ at ages 14–22 in a representative sample of young Americans (Rindermann & Ceci, 2018, supplemental file p. 3). If tests of g really were proxies for socioeconomic status, the correlations would be much stronger. Discussing the same evidence, Mackintosh (2011, p. 29) stated that the correlations are so weak that the idea that intelligence tests measure socioeconomic status “is a singularly foolish assertion.” I agree.

Controlling for confounding variables (e.g., school quality, number of books in the home, technological access) only makes the relationship between family income and intelligence weaken. This was most apparent in a recent study with data from 19 samples in 7 countries on 4 continents in which the correlation between the two variables went from r ≈ .25 to r ≈ .12 after controlling for confounding variables (Rindermann & Ceci, 2018). This means that the relationship between income and intelligence test scores can be partially explained by other variables, though controlling for these other influences does not make the correlation between IQ and socioeconomic status disappear completely.

Another piece of statistical evidence indicating that intelligence tests are not proxies for socioeconomic status comes from educational testing. Scores on tests of g – like college admissions tests – are the best predictors of academic performance (see Chapters 18–20). Statistically controlling for socioeconomic status has almost no impact on the ability of test scores to predict grades (Sackett et al., 2008), and even after controlling for childhood socioeconomic status, IQ has a moderately strong positive correlation with later income and educational success (Kuncel & Hezlett, 2010; Murray, 1998, 2002). This indicates that the correlation between IQ and academic performance is mostly independent of socioeconomic status – even though all three variables are positively correlated with one another.

Even if intelligence test scores are not total reflections of socioeconomic status, there exists the possibility that socioeconomic status could cause a boost in intelligence test scores. The best evidence for this comes from adoption studies, which generally show that children in adopted families (which tend to be middle or upper class) often score at or above average on intelligence tests (e.g., Spinks et al., 2007). The best study, from Sweden, showed that adopted children had IQ scores that were an average of 4.41 points higher than the scores of their biological sibling who had been raised by their biological parents (Kendler, Turkheimer, Ohlsson, Sundquist, & Sundquist, 2015). Even this evidence, though, does not indicate that income causes increases in intelligence test scores because low, middle, and high socioeconomic status families vary in many ways besides income (Protzko, Aronson, & Blair, 2013). For example, in middle-class homes, mothers talk to their children more, and children watch less television (Elardo & Bradley, 1981). Likewise, children in poor households are less healthy and less likely to live in a two-parent home (Brooks-Gunn & Duncan, 1997). Indeed, children in homes with positive social characteristics that are unrelated to income tend to have higher IQ scores (Cleveland, Jacobson, Lipinski, & Rowe, 2000). Thus, it is very possible that the correlation between test scores and socioeconomic status may not be entirely due to the impact of family income on intelligence.

Genetics: Setting Limits on the Influence of Environment

In addition to the many ways that homes from different socioeconomic statuses vary, it is important to consider one factor which limits the causal impact that income can have on intelligence test scores: genetic influences on intelligence. An interdisciplinary science called behavioral genetics studies the influence of genes on psychological traits and behaviors. One technique of behavioral genetics is to use correlations between family members’ scores on a trait to determine the trait’s heritability, which is the degree to which trait differences among people are due to genetic differences. The exact methods are too complex to explain here (see Plomin, DeFries, Knopik, & Neiderhiser, 2012, for a detailed explanation), but if the correlation between family members’ traits is stronger for close relatives than for more distant relatives, it is an indication that the trait is genetically influenced. For example, Bouchard and McGue (1981, p. 1056) reported that IQ scores are correlated r = .15 for cousins and r = .47 for siblings. Because siblings share – on average – 50% of their genes and cousins share only 12.5% of their genes, the stronger correlation among siblings indicates that intelligence is genetically influenced.

Heritability is abbreviated as h² and ranges from 0 (indicating that differences in a trait are solely due to environmental differences) to 1 (indicating that genetic differences are the only influence determining differences in a trait). When h² is low, environmental variables are more important than genetics in determining trait variability; when heritability is close to 1, then genes are a powerful influence on a trait, and the environment has very little impact in determining differences on the trait. In reality, h² h2values of 0 or 1 are very rare; for almost every trait, heritability is between these two extremes (Plomin, DeFries, Knopik, & Neiderhiser, 2016).

Correlations in IQ scores among relatives show evidence of a genetic influence on intelligence. For identical twins (who share 100% of their genes), the correlation between their IQ scores is r = .86; the fact that this value is not r = 1.00 indicates that genes are not the only factor determining IQ scores. Likewise, adoptees and their non-biological relatives (who share 0% of their DNA) have IQ scores that are correlated r = .19 to .24 (Bouchard & McGue, 1981, pp. 1057–1058). This positive correlation indicates that the environment these family members share has an influence on IQ scores, though the environmental influence is not as strong as the genetic influence on IQ scores.

Heritability for intelligence tends to be around .50 (i.e., about 50% of IQ score differences are due to genetic differences), though there are differences from study to study. Generally, studies of children tend to produce lower heritability (and therefore higher environmental/non-genetic influence), often as low as .20. Studies of adults produce higher heritability – sometimes above .80 (Bouchard, 2004, 2014; Deary, 2012; Hunt, 2011). This indicates that the importance of genes increases as people age (Plomin & Deary, 2015). In other words, intelligence differences among adults are more genetic in origin, whereas in young children, environmental variables matter more.

What does genetics have to do with understanding the impact of socioeconomic status on intelligence? Heritability sets limits on the influence that non-genetic variables can have on intelligence in typical environments (Hunt, 2011). If the differences in IQ scores are 50% caused by genetic differences (as indicated by a h² value of .50), then all environmental variables combined must account for no more than 50% of people’s differences in intelligence. While this does not exactly tell scientists the strength of the impact of income on intelligence, it does limit the magnitude of that impact. Because genetics is as important for determining intelligence as all other environmental variables combined, it is not possible for differences in IQ scores – or scores on other tests that measure g – to be determined solely (or mostly) by income or socioeconomic status differences.

Added Complexity: Genetically Influenced Environments

There is another finding from behavioral genetics that makes it difficult to argue that intelligence test scores are solely caused by socioeconomic differences: many “environmental” variables are also heritable. In other words, genes can influence the environment that people find themselves in. This idea was postulated long ago (e.g., Pearson, 1903, pp. 179–180), but in the past generation the evidence has mounted that it is correct (Vinkhuzen, Van Der Sluis, De Geus, Boomsma, & Posthuma, 2010). As far as socioeconomic status is concerned, heritability is high enough that genes are a major influence on income differences: .42 in one highly cited study (Rowe, Vesterdal, & Rodgers, 1998), which is typical in developed countries (Plomin, 2018, p. 100). Moreover, some of the genes that influence socioeconomic status also influence intelligence (Marioni et al., 2014; Trzaskowski et al., 2014). Therefore, some of the impact that socioeconomic status has on intelligence is ultimately genetic in origin – even though many people consider socioeconomic status and related variables (e.g., income, poverty) to be environmental variables.

It is not always clear what causes this genetic influence on environmental variables. One leading theory is that people’s genetic proclivities lead them to choose environments they feel comfortable in. For example, a child who has a genetic propensity to enjoy reading may choose to spend more time in their school’s library. This may explain why heritability is higher for adults (who have more freedom to choose their environments) than for children (Plomin & Deary, 2015). Another possibility is that people in the surrounding environment, such as parents, teachers, or employers, may respond to a person’s behavior and foster development in areas where the person has a (genetically influenced) interest or talent. This might be what occurs when a parent buys a trumpet or violin for a child who shows interest in music; the access to an instrument will then amplify the child’s genetic propensity for musical talent. In all likelihood, both theories have some truth to them.

Most children are raised by one or both biological parents, which means that a child shares 50% of their genes with at least one caregiver. As a result, genes have multiple ways of influencing the development of both individual traits – like intelligence – and “environmental” variables – like socioeconomic status. An important implication of these shared genes is that a correlation between parental behavior and child outcomes may be genetically caused. For example, it is known that the number of books in a home is correlated with a child’s school performance and that more educated parents tend to buy more books (e.g., Rindermann & Ceci, 2018). This does not prove, though, that buying more books causes children to do well in school. Instead, the same genes that may make a parent succeed in school may also influence them to buy more books; half of those genes from each parent are then passed on to the child. These genes that the child receives may then cause the child to read at home and also do well in school. What is apparently an environmental influence (of the number of books on a child’s school success) may be an entirely genetic phenomenon.

Shortcomings of Heritability Estimates

Although h² values provide extremely strong evidence that a person’s intelligence level is partially influenced by their genetic heritage, there are a few limitations that anyone dealing with heritability should know. The first is that h² estimates can only apply to the population and environments under investigation. Most studies of heritability occur in wealthy, industrialized countries, and these results may not apply to other nations. This is well illustrated in heritability studies conducted in impoverished countries. In one study in Sudan, the h² for intelligence in a sample of 10-year-old children was between .13 and .17 (Toto et al., 2019), which is about half of the h² value seen in wealthy nations for that age group. This indicates that genes are a less influential cause of differences in intelligence than environmental variables in Sudan. On the other hand, the h² for IQ in a sample of Nigerian adolescents aged 11–18 was .50 (Hur & Bates, 2019), which is consistent with the h² values for adolescents in Western countries. Thus, heritability values do not apply to environments that were not part of the study. Moreover, h² values say nothing about biological immutability of a trait (Cronbach, 1975).

To a lesser extent, a problem with heritability study samples is that they tend to consist of more middle- and upper-class individuals than a representative sample would have. This is especially true of adoption studies because the poorest families in industrialized nations are usually not allowed to adopt children, nor are parents who have a history of violence, drug problems, or other dysfunctional behaviors. Therefore, the range of environments in these studies is reduced (Mackintosh, 2011), which makes the influence of genetics appear inflated (Nisbett et al., 2012). Adoption studies also cannot investigate the impact of abuse, neglect, and threats to physical safety because governments and adoption agencies try to prevent children from being placed into these extremely negative environments. Therefore, when behavioral geneticists produce a study that genes are a powerful influence on intelligence, it is important to consider the population and the environment that the study was conducted on. Often, the results of behavioral genetics studies will indicate that genes are important – if a person already lives in an industrialized nation in a home where basic needs are met. It is not clear how well these results apply to individuals in severe poverty or in highly unfavorable environments.

Another shortcoming of heritability statistics is that they do not state what portion of a person’s IQ originates from their genes and what portion originates from their environment (Tal, 2009). That is because heritability is a group-level statistic of variance that refers to the genetic influences on the variability of a trait among a group of individuals (Hunt, 2011). Therefore, if h² is .50 (i.e., 50%), and someone has an IQ of 100, it does not make sense to say that 50 points of their score come from genetics and 50 points come from environmental influences. Despite this apparent drawback, h² values do provide important information about groups as a whole.

Conclusion

It is abundantly clear that IQ scores do not merely reflect an examinee’s socioeconomic status. Indeed, the correlations between the two variables were never strong enough for that to be a plausible interpretation of the data. A more important influence for determining intelligence is a person’s genetic makeup. Among individuals living in middle- or upper-class homes in wealthy, industrialized nations, heritability is approximately .50. Among people outside this group, heritability may or may not be the same (though it is probably not zero). At least some of the correlation between socioeconomic status and IQ scores is probably due to genetic factors. Adoption studies do show that children adopted into middle- and upper-class families have a boost to their intelligence, but it is not clear how much of this is due to wealth and how much is due to the other characteristics of these homes. The research on this topic is complex, and new discoveries are certain to unfold.

From Chapter 11 of "In the Know: Debunking 35 Myths About Human Intelligence" by Dr. Russell Warne (2020)

We hope you found this information useful. For further questions, please join our Discord server to ask a Riot IQ team member or email us at support@riotiq.com. If you are interested in IQ and Intelligence, we co-moderate a related subreddit forum and have started a Youtube channel. Please feel free to join us.