Human Genetics 101 - Every single behavioural trait ever documented among human beings is heritable
"Twins have a special claim upon our attention; it is, that their history affords means of distinguishing between the effects of tendencies received at birth, and those that were imposed by the special circumstances of their after lives." — Francis Galton, 1875
Genetic inheritance is passed down vertically within families. Genetic difference can be measured in a number of ways. One way, as Galton observed in the 19th century, is to look at how closely identical (monozygotic; referred to as “MZ” in behavioural studies) twins resemble each other on a specific trait compared to same-sex fraternal (dizygotic; DZ) twins (a control group). With random mating identical twins share ~100% of their genes and fraternal twins only ~50%. Under any reasonable assumption environments will be equally similar for fraternal (DZ) as for identical (MZ) twins. If genetics made no difference to a trait the trait’s correlation for identical and fraternal twins would be the same. If identical twins show a stronger correlation than fraternal twins, e.g., 0.4 versus 0.2, that implies that the transmission of the trait from parents to children is AT LEAST 50% genetic. For an example, observe the MZ and DZ correlations, compared with the genetic (A) estimate, in the table below:Table source: Heritability of Positive Psychological Traits
Proponents of the blank slate, the predominant idea among intellectuals that individual human nature is simply a product of the environment and upbringing — the cornerstone worldview of the Enlightenment — long questioned the validity of such twin studies, but GCTA analysis all but put an end to this.
A newer and more advanced technique, known as Genome-wide Complex Trait Analysis (GCTA), allows estimation of heritability due to common SNPs (as identified by genome-wide association studies (GWASs)) using relatively small sample sizes. It examines pairs of unrelated individuals and computes the correlation between pairwise phenotype similarity and genotype similarity (relatedness). This newer, more accurate method finds similar estimates of genetic inheritance on human traits as previously measured by the old (but evidently still completely valid) twin studies. Meaning, those who still doubt the validity of twin studies, and subsequently the influence of genes, now have another kind of evidence to contend with. It has been applied to height, intelligence, and many medical and psychiatric conditions:Source: Estimation and Partition of Heritability in Human Populations Using Whole-Genome Analysis Methods
Another method used to measure the role of genes is the outcomes for adopted children compared to their biological and adoptive parents. Studies of criminality, using either twins or adoption methods, reveal a strong genetic connection, e.g., the chance an adoptee would end up with a criminal record when neither set of parents had one is relatively low. When only the adoptive parent had a criminal record this chance rises only very slightly - if at all. However if only the biological parent had a criminal record the chance of the adoptee having a criminal record rose much more. If both sets of parents had a criminal record the chance of the adoptee having such a record was again higher. Therefore, genetic influences on criminal propensities are much greater than environmental influences. Below is a table from Bouchard’s Sources of Human Psychological differences: Minnesota Study of Twins Reared Apart, a renowned longitudinal study of twins:Above: MZA = Monozygotic (identical) twins raised apart; MZT = Monozygotic twins raised together. On the personality traits the MZA and MZT correlations are almost identical — even more so than for intelligence. Evidently, the shared family environment does not make twins more similar in personality, and only slightly more similar in IQ.
The Shared Environment and the Unique "Environment"
In behavioural genetics studies, traits are measured by their influence from genetics effects (genes), the shared environment, and the unshared (unique) environment. In twin studies, the shared environment, (e.g. parental influence, education, wealth, culture, community) would signify events that occur to both twins, affecting them in the same way. This force is found to have an insignificant (or even statistically zero) lasting effect over how an individual turns out, on seemingly any trait analysed. Parents leave no lasting effect on adult outcomes, aside from what they bequeath to their children genetically— it’s the genes people share, not examples set by their parents, that explains the relationship between parents and their biological children.
It is the unshared, or unique environment—events that occur to one twin but not the other, or events that affect either twin in a different way—that seems to influence traits in a way that is attributable neither to genetic nor familiar variation. This can include measurement errors and perhaps other random variables such as (but not limited to) the effects of chance – new mutations, odd prenatal effects, developmental noise in brain development, events in life with unpredictable effects and perhaps anything else that is not subject to controlled manipulation. Very little is known about this miscellaneous, unknown and misleadingly titled "environmental" factor.
Not only does the research in human behavioural genetics demonstrate that genes impacts human behaviour, it finds that virtually every single behavioural trait ever documented among human beings is heritable. From height and BMI, to personality and IQ: Image source: Estimation and Partitioning of Heritability in Human Populations Using Whole Genome Analysis Methods (GCTA Methods)
A large fraction of the changes that can be observed in human behaviour occurred in the historical period, the majority of these changes are genetic in nature. Traits linked to the most important life outcomes, such as intelligence, have high heritabilities (~0.80). Some of this research is interpreted in the posts below:
Additional papers: Marital stability/divorce risk , ,  (twin studies); Substance abuse [review of meta analyses]; Smoking [meta analysis]; Promiscuity [twin study]; Body Mass Index [GCTA]; Various psychopathic traits [meta-analysis].
If you have a trait which is additively influenced by many essentially random variables (most complex human traits are polygenic; no single gene of large effect, instead many genes of small effect), by the central limit theorem you expect a quasi-normal (Gaussian) distribution in the population. Most of the interesting human traits, such as racial differences in intelligence, personality, and so forth, follow a Gaussian distribution. In more layman’s terms, quantitative traits (such as intelligence) are normally distributed within populations, meaning, people of every population can theoretically be found at every level of the trait’s distribution, it’s just a question of probability, and differences are always more pronounced at the edge of the distribution, just like any other quantitative trait.
The “mean” to which offspring will regress (regression toward the mean is a well documented phenomenon in population genetics that was originally discovered by Galton), is the mean of their ancestors. Unrelated individuals cannot impact the heritability of a given trait, i.e., there is one distribution, not many - each group regresses toward their own mean (of their ancestors), not that of the overall mean.
Thus, a moderate difference in means between two populations makes a big difference in the fraction that exceeds a high threshold: that’s just a consequence of the shape of the distribution, which falls off more and more rapidly as you get farther from the mean. So the difference in the mean between two groups has many implications. For one, it implies that between-population differences in IQ, for instance, become much larger at the highest levels of ability:Source: The g Factor: The Science of Mental Ability
Marriage is assortative. This means that people marry those who are like themselves in heritable traits; tall people are likely to have tall mates while short people are likely to have short mates, people with more years of education have more highly educated spouses, and so forth. Traits such as political and religious attitudes show a particularly strong “pull” of genetic similarity in assortative mating [see previous link]:Thus even fraternal twins share more than 50% of their genes. The higher correlation of traits between identical as opposed to fraternal twins is thus based on even less than a 50% difference in genetic material. Genes must explain even more, e.g., when not accounting for assortative mating, a couple, each from a family of orthodox Christians (again; religious attitudes is a trait for which assortative mating is particularly strong) tend to inflate the estimated influence from parents (the shared environment), at the expense of the genetic estimate, in studies that would be examining the heritability of religiosity. (The previous table shows a small shared environmental component influencing religiosity (and other traits), but when accommodating for the effects of assortative mating and measurement error; even a small amount of influence from the shared environment can be accounted for).
Moreover, many studies don’t partition out measurement error, that is, inaccuracies in the assessment of the traits to be analysed, which has the effect of attenuating the heritability estimate. Other studies, which use methods to get around that problem, find heritabilities for personality traits in the 0.7-0.8 range, as is found for IQ, mental disorders, height, BMI, etc.
Noted palaeontologist, Stephen Jay Gould, spent his career arguing against the concept that selective pressures have resulted in human population differences. Since everything suggests this, he kept busy. During the period in which humans spread out of Africa, findings indicate that human evolution greatly accelerated, with at least 7% of the human genome changing over the last 40,000 years, which is consistent with population-specific gene frequencies (i.e., the genetic clustering of socially constructed human racial groups). Not only are humans still evolving, but studies of selection and differentiation of human populations are alive and well. For instance, there is direct evidence of recent selection impacting the French who settled Québec, or rather, there are varying gene frequencies (genetic differences) present among two closely related contemporary human populations (French and French Canadians). Selection pressures on the broader scale have more notable, long lasting effects on global human variation, (i.e., racial differences). For instance, a single gene is a long chain of nucleotides, and a single nucleotide mutation can significantly alter how an entire gene works. Therefore, a mutation of large effect, which differs by only 0.1 from one population to another, can make a big difference. See: the misleading “all humans are 99% the same” and “more genetic variation in humans is found within populations" (Lewontin) statistics.