Race and IQ: Genes That Predict Racial Intelligence Differences

For the past century, psychologists have recorded racial differences in intelligence test scores showing that Asians score higher than Whites who in turn score higher than Blacks. The causes of these intelligence differences have been heavily debated.

One challenge often put to the so-called “hereditarians,” those who say that the gap is significantly caused by genetics, is to name the specific genes which make some races smarter than others. Until recently, this has been impossible due to technological limitations and so hereditarians have relied on less direct evidence when making their case. However, in recent years, new research has come out which has pinpointed several genes that are probably involved in racial intelligence differences. This post will describe this research and explain why what has been found thus far provides powerful evidence for the hereditarian viewpoint.

The Genes

The research comes from 3 papers which looked at how 14 alleles (gene variants) which were previously associated with intelligence, or a proxy for intelligence, vary by race (Piffer 2013), (Piffer 2014), (Piffer and Kirkeggard 2014). In a sample of 101,069 10 of these 14 alleles were each found to predict higher than average educational attainment (Rietveld et al. 2013). The predictive ability of each allele was then retested again across 12 samples totaling 25,290 people. All 10 alleles were found to be associated with intelligence in multiple samples, though the the associations were not always statistically significant. Importantly, the samples consisted of only white people, which means that no genes arbitrarily associated with race will be falsely thought to associate with education just because race does. What were the alleles associated with biologically?

These genes were only shown to directly associate with education, but there is good reason to think that they predict intelligence as well. For one thing, intelligence highly correlates with education. Secondly, previous studies that have sought out alleles associated with education have found that they predict intelligence test scores even better than they do education.

The other four alleles come from more varied sources. The first is a version of the NPTN gene, which is involved in how the brain changes itself (neural outgrowth and synaptic plasticity). A particular allele of this gene has previously been found to predict lower IQ scores and less cortical thickness. The second allele comes from the FNB1L gene and has been associated with high intelligence across multiple studies. The third allele is a version of the CHRM2 gene and has been associated with high intelligence in 4 separate studies. Finally, in a meta-analysis of 77 previous studies, a version of the APOE4 gene has been found to predict better memory, perceptual speed, and general cognitive functioning. Each of the studies involving these four genes used different sets of controls and statistical adjustments. Because they have been found to associate with intelligence so consistently, a causal relationship between the allele and intelligence is likely.

The Databases

Data on the frequency of each allele across different racial populations was taken from three databses: ALFRED, HapMap, and 1000 Genomes. Each of these databases collects genetic data taken from samples all over the world. Combined, they have genetic samples from well over 100 distinct populations. They are highly reputable, and having access to three different databases allows these findings to be replicated multiple times. (The first ten alleles were tested across all three databases while the other set of four were only tested across ALFRED and 1000 Genomes.)

The Results

The 14 alleles were found to be patterned such that, based on this genetic data alone, Asians would be predicted to have the highest IQs followed by Whites and ending with Blacks. These differences were statistically significant and were replicated across all three databases. More extraordinary was the finding that all 14 alleles differed between Blacks and Whites in a way that would predict that Blacks would be less intelligent. This result strongly suggests that the hereditarian viewpoint is the correct one. The egalitarian viewpoint would predict that each allele should have, on average, a 50% probability of existing in a greater frequency among either Blacks or Whites. Therefore the probability of the first 14 alleles examined all favoring Whites would be a mere 1 in 16,284. Obviously, the probability of this happening under the hereditarian model is much higher.

Given the logic of science, this clearly suggests that the hereditarian viewpoint should be favored: we have two competing hypotheses one of which would make an outcome extremely unlikely and the other which would make it probable. We have found that said outcome has materialized and, on this basis, can declare one hypothesis, egalitarianism, highly unlikely; and the other hypothesis, hereditarianism, probable.

Below are charts showing the distribution of all 14 alleles across 4 racial groups. The data was taken from ALFRED but, as we have already seen, the pattern is consistent across other databases as well. (The first chart is for an allele associated with lower than average cognitive ability. All the others alleles are positively associated with intelligence.)


Piffer, D. (2014). Simple Statistical Tools to Detect Signals of Recent Polygenic Selection. Interdisciplinary Bio Central, 6, 1-6. doi:10.4051

Piffer, D. (2013). Factor Analysis of Population Allele Frequencies as a Simple, Novel Method of Detecting Signals of Recent Polygenic Selection: The Example of Educational Attainment and IQ. Interdisciplinary Bio Central, 1-31. doi:10.4051

Piffer, D., & Kirkegaard, E. (2014). The genetic correlation between educational attainment, intracranial volume and IQ is due to recent polygenic selection on general cognitive ability. Open Behavioral Genetics, 1-35.

Rietveld, C. et al (2013). GWAS of 126,559 Individuals Identifies Genetic Variants Associated with Educational Attainment. Science, 340, 1467-1471. doi:10.1126/science.1235488