Geneticist James J. Lee is a professor of psychology at the U. of Minnesota, where he is the successor to Tom Bouchard, leader of the famous “Minnesota Twins” study. In the Proceedings of the National Academy of Science, Lee has published a commentary on economic historian Gregory Clark’s new landmark study of the class system in England:

The heritability and persistence of social class in England

James J. Lee
Department of Psychology
University of Minnesota Twin Ci@es

The extent to which hereditary abilites affect social class has been a matter of interest since classical times. In 1918, Ronald Fisher published a framework for answering such questions about the inheritance of continuously varying traits and thereafter applied that framework to the “mental and moral qualities” that he presumed to be at work in the stratification of human populations. In this issue Clark reports an application of Fisher’s methods to English pedigree data spanning centuries and achieves a tour de force of insight into the inheritance of social status.

Fisher showed that the correlation between relatives induced by genetics is equal to the heritability times a coefficient that depends on the genealogical relationship between the relatives and the extent of assortative mating (spousal resemblance in the trait). By finding several such correlations to the theoretical equations, one can obtain estimates of the key parameters (heritability, assortative mating). Clark applied this theory to several measures of social class in a remarkable dataset formed by combining the immense labors of amateur genealogists with publicly available records of births, baptisms, marriages, probates, and so forth. It is natural to suspect such data of biases; but Clark found, among other things, that members of different lineages paired at random show no correlations in any of the measured social outcomes.

We might distinguish three possible kinds of models explaining the generational persistence of the abilities underlying social status. The first is a simple autoregressive model that has sometimes been used in social science, where the correlation in ability or skill across n generations is essentially rn, the correlation between parent and child to the power of the
number of steps in the genealogy (2 for grandparent and grandchild, 3 for great-grandparent and great-grandchild, and so on). Since most single-generation correlations in observable traits are well short of 1/2, this model implies a very rapid decay in resemblance across generations.

The second kind of model acknowledges the genetic contribution to ancestor-descendant resemblance but posits random pairing of fathers and mothers, which yields the correlation (1/2)n times the heritability. This is still a rapid decay across generations. The third kind of model allows mating to be assortative, in which case the correlation becomes essentially [(1 + m)/2] n times the heritability, m being the correlation between the genetic values of spouses.

Strong assortative mating is reasonably regarded as a qualitatively different regime altogether, since with a high enough value of m (e.g., 0.57) regression to the mean across generations is dramatically slowed and allows the common descendants of a Victorian ancestor to show quantitatively significant genetic resemblance even today. This is because such a scheme of strong spousal matching ensures that the other ancestors of the individual’s descendants tend to transmit DNA of like effect. Such persistent correlations between very distant relatives are what drive Clark’s inferences of strong assortative mating.

One naturally wonders whether the results might be explained by environmental sources of familial resemblance that are confounded with genetic relatedness. Clark reports compelling ancillary evidence, however, against explaining away his findings in such a way. … Second, he found that status was transmitted equally through paternal and maternal lines, exactly as expected if the responsible causal factors are the genes transmitted by fathers and mothers alike.

In the English class system, property tends to be passed down to the first-born male. So the correlation between male-line relatives is higher. But genetic relationships are close to equal and English social relationships are relatively equal on the father’s and mother’s side. The English are not the Chinese who cut off relationship’s with the wife’s family.

An interesting test of Clark’s theory that genetics rather than inherited wealth drive the English class system would be to study how much primogeniture matters. Are the descendants of first-born males privileged? Most English would say “Yes,” although this can be a tricky question to research since the first born doesn’t need a profession (he’s going to inherit Downton Abbey and go into Parliament) while the the latter born sons need professions (the cleverest into law, the bravest into the Army, the most pious into the Church, etc.).

[Comment at Unz.com]