By Arnold Kling
- DNA is the only thing that makes a substantial difference, accounting for 50 per cent of the variance in psychological traits. The rest comes down to chance environmental experiences that do not have long-term effects.
- —Robert Plomin, Blueprint: How DNA Makes Us Who We Are1
After decades of studying the heritability of psychological traits, Robert Plomin argues that analysis of DNA can help predict individual characteristics. By the same token, interventions in our environment, including parenting styles and variations in education, have little or no long-term effect.
Plomin’s ideas have the potential to disturb many standard political views. For example, conservatives argue that poverty can be cured by encouraging people to finish high school and get married before having children. Progressives argue that poverty can be alleviated by spending more on education. But if Plomin is correct, then attempts to use environmental interventions to influence behavior or ability are equally unlikely to prove successful.
One of the many interesting claims Plomin makes is that heritability tends to increase with age. One might think that environmental influences would tend to accumulate and hence matter more as we get older, but the opposite appears to be the case. For example, Plomin writes that
- … the heritability of weight increases from about 40 percent in early childhood to about 60 percent in adolescence to about 80 percent in adulthood.
Plomin speculates that this due to the gradual development of humans. As we reach adulthood, “we grow into our genes.”
- It is difficult to detect any differences in childhood for individuals who are later diagnosed as schizophrenic. It is likely that genes that contribute to the disorganized thinking, hallucinations and paranoia characteristic of schizophrenia do not have their effect until the brain has developed to a high level of symbolic reasoning in early adulthood.
Plomin says that we often attribute behavior to the environment when it really comes from genetics. For example, when we observe that the offspring of divorced parents are themselves more likely to become divorced, he writes
- … the link between divorce in parents and divorce in their children is forged genetically, not environmentally. For a sample of 20,000 adopted individuals, the likelihood of divorce was greater if their biological mother, who did not rear the individual, had later in life become divorced than if the adoptive parents who reared them had become divorced.
- The heritability of divorce is about 40 percent across studies. This is a long way from 100 percent, meaning that non-genetic factors are also important. However, the major systematic factor affecting divorce is genetics. In contrast, no environmental predictors of divorce have been identified in research after controlling for genetics.
He goes on to write that,
- … evidence for genetic influence has been found for home environments such as chaotic family environments, for classroom environments such as supportive teachers, peer characteristics such as being bullied, neighbourhood safety, being exposed to drugs, work environments, and the quality of one’s marriage.
Plomin asserts that the relationships between these environmental factors and psychological characteristics is not causal. So, for example, when a trait is correlated with having a chaotic family environment, this is due to a genetic connection between having that trait and experiencing such an environment. He calls this “the nature of nurture.” This analysis is based on studies comparing identical with fraternal twins, studies of twins raised apart, and studies examining correlations of traits with patterns in DNA.
One of the challenges with predicting psychological characteristics is that the dependent variable (the trait to be predicted) may not be reliably measured. For example, some critics of personality psychology will point out that when you measure, say, extroversion, in an individual at two different points in time or using two different survey instruments, the results can differ.
Another problem, which Plomin emphasizes, is that we are given binary data about individuals based on diagnoses. Someone either is diagnosed as schizophrenic or not. Someone is either diagnosed as bipolar or not. Yet he considers it likely that the underlying traits are normally distributed, and the diagnoses are applied to people at one end of the spectrum.
Imagine an index in which a low number indicates a low probability of manifesting schizophrenia and a high number represents a high probability. It could be that the overall tendency is for people to have a middle number on this index, and schizophrenics have a number that is two or three standard deviations above average. It could also be that having a number that is two or three standard deviations below average is also harmful, in ways that we have not yet systematically cataloged.
Plomin offers a helpful analogy. Suppose that we labeled everyone with height more than two standard deviations above average as being afflicted with a disorder called “giantism.” Then we would have a binary variable associated with this disorder, even though the underlying trait is normally distributed.
Having an unreliable dependent variable makes it harder to find correlations. Plomin alludes to this. But he does not point out that if more reliable measures could be found, this would increase the likelihood of finding both stronger genetic influences and more robust environmental influences.
Plomin is excited by polygenic scores, a recent development in genetic studies. Researchers use large databases of DNA-sequence individuals to identify combinations of hundreds of genes that correlate with traits.
- The most predictive polygenic score so far is height, which explains 17 percent of the variance in adult height… height at birth scarcely predicts adult height. The predictive power of polygenic scores is greater than any other predictors, even the height of the individuals’ parents.
One can view this 17 percent figure either as encouraging or not. It represents progress over attempts to find one or two genes that predict height, an effort that is futile. But compared to the 80 percent heritability of height it seems weak.
Plomin is optimistic that with larger sample sizes better polygenic scores will be found, but I am skeptical. Unless there are unexplored areas in the existing data sets, such as non-linearities or interaction effects, my guess is that there are diminishing returns to enlarging the sample size.
I wish that Plomin had done a better job of anticipating criticisms of various sorts. For example, “Flynn” does not appear in the book, even though the Flynn Effect, which is a finding that IQ has risen over time, suggests that environmental effects do in fact matter for psychological traits.
In fact, it is difficult to dismiss the significance of what might be termed macro-environmental effects. For example, Plomin wishes to claim that genes account for all the systematic variance across individuals in years of schooling. But clearly there are differences across countries or across generations. It is not plausible to deny any systematic environmental effects whatsoever.
As I see it, Plomin’s claims for the potential of polygenic scores to explain psychological traits are speculative. His claims may turn out to be correct, or they may not. Either way, the ongoing research is very important.
 Rober Plomin, Blueprint: How DNA Makes Us Who We Are. MIT Press, 2018.
*Arnold Kling has a Ph.D. in economics from the Massachusetts Institute of Technology. He is the author of several books, including Crisis of Abundance: Rethinking How We Pay for Health Care; Invisible Wealth: The Hidden Story of How Markets Work; Unchecked and Unbalanced: How the Discrepancy Between Knowledge and Power Caused the Financial Crisis and Threatens Democracy; and Specialization and Trade: A Re-introduction to Economics. He contributed to EconLog from January 2003 through August 2012.
Read more of what Arnold Kling’s been reading. For more book reviews and articles by Arnold Kling, see the Archive.