Race Differences in Ethnocentrism

by Edward Dutton

  A number of criticisms of Genetic Similarity Theory have been highlighted and responded to by Salter and Harpending (2015). Firstly, it may be argued that inclusive fitness can only operate between genealogical kin because their genes are identical by common descent (e.g. Mealy, 1985). However, it can be countered that inclusive fitness will operate regardless of the how the similarity arises and this is evidenced by the kind of data which Rushton has presented.

  Secondly, it might be argued that ethnic kinship is too slight to justify diverting effort from genealogical kin. However, this is simply untrue. The aggregate kinship within populations can be sufficient that it is adaptive to invest in ethnic kinship, as Salter (2007) has demonstrated. It is perfectly adaptive to contribute to collective goods, group defence and the punishment of free-riders.

  Thirdly, Grafen (1990) pointed out that within an ‘outbred population’ (one that was relatively genetically diverse) assortment by phenotypic selection could not be a form of kin selection, as Rushton had argued. This was because sharing the kinds of characteristics which Rushton highlighted would involve sharing miniscule proportions of genes. These genes would only occur frequently on the genome among genealogical kin. Grafen argued that the percentage of shared genes would be insufficient, beyond genealogical kin, for the investment of friendship, for example, to pay off in terms of inclusive fitness. But, as Salter (2007) has found, if we compare different ethnic groups — calculating their degree of genetic separation — then co-ethnics, relatively speaking, share a significant percentage of their genes; more than sufficient to make ethnic nepotism adaptive. Indeed, Salter (2002) has found in an ethnically divided population, people will share 15% more genes with co-ethnics than with others, meaning that investment in those with shared phenotypic traits would pay off even more. Moreover, it might be countered that even in an outbred population, if a person shared a large number of phenotypic traits with someone else — as Rushton found was true of friends or sexual partners — then the investment could potentially pay off. But it is true that this is more likely within an inbred population, and Salter and Harpending (2015) note that most populations are inbred to some extent. In this context, kinship is highly variable even among complete strangers, and this would be even more likely in small, relatively isolated populations. For people in these populations, Rushton’s model could apply even without the need to hypothesize collective goods and there would not need to be any ethnic conflict for the investment to pay off in terms of improved fitness. However, in more varied populations, altruism to strangers would not necessarily increase fitness.

  Fourthly, it has been argued by some cultural anthropologists that tribal people are essentially peaceful and kind, except for disruptions caused by colonialism. Before the development of agriculture, people were relatively immobile foragers who moved only very short distances and only interacted with people like themselves. As such, there would have been no selection in favour of ethnic or racial discrimination. The fundamental problem with this criticism is that Rushton has presented evidence that people can detect even the slightest genetic differences, even within families. Moreover, there is no reason to believe that pre-agricultural humans were anything like some anthropologists portray them as being. Evidence from ethnographies with surviving foragers (e.g. Chagnon, 1968; Chagnon, 2013) indicates that their lives are characterized by extreme violence, including frequent battles with other tribes (and, so, other ethnic groups), territory invasions, and genocide. As such, a capacity to discriminate against members of a different tribe — and by extension ethnicity — would be highly adaptive.

  Fifthly, it might be averred that Rushton’s model does not fully explain variations in ethnocentrism. As we have seen, ethnocentrism is only modestly heritable, the region of 0.3. Accordingly, environmental factors must be significant in explaining variations in ethnocentrism and, in particular, why ethnic solidarity varies in cultures over time. Ethnic solidarity is generally of a moderate or low intensity but can reach fever pitch due to even the most minor attack on the group or even due to a slur against its identity. As such, threat to the interests of the ethny appears to be a highly significant dimension to ethnocentrism, meaning that it cannot be explained solely in terms of directly increasing inclusive fitness. A more subtle approach is required.

  Further, it may be that a certain kind of personality is generally more ethnocentric and since personality alters throughout the lifespan, we would expect that the average age of a population would impact how ethnocentric it was. And this raises the broader question of why some populations are stereotyped as being particularly ethnocentric (e.g. MacDonald, 1996). Rushton’s model would explain this in terms of greater genetic similarity in comparison to outsiders (in other words, inbreeding) and historically having been in conflict with another group, leading to greater selection for ethnocentrism. However, this raises the question of by what mechanism this ethnocentrism has developed. Also, Rushton (1995) has looked at population differences in modal personality and so-called ‘Life History Strategy’. These may have some impact on the degree of a population’s ethnocentrism. We will explore all of these issues later in this book.

  Sixthly, research by Kurzban et al. (2001) has highlighted the degree to which ethnic nepotism is sensitive to cultural cues in a way which might be seen to challenge Genetic Similarity Theory. They found that the subjects of an experiment were less inclined to categorize fellow subjects in terms of race when the race of these subjects did not correlate with being or not being in their particular, experiment-manufactured group. However, this was not true of categorization by sex. This remained even when coalitions were composed of both males and females. This might be regarded as a challenge to the view that racial categorization is automatic and innate. But it can be countered that this does not seriously challenge the theory because Genetic Similarity Theory extends to intra-racial relations as well, based, often, around very weak ties. Also, it merely highlights the fact that — as already noted — humans can engage in mutual reciprocity and it is possible for this to be with somebody who is genetically very different, especially if such an alliance might be regarded as beneficial to one’s own fitness in some way. In these circumstances, we might expect that people would regard an individual’s genetic dissimilarity as less pertinent. An anecdotal example of this would be the way that, in the 1980s when I was a boy, those in England who might argue that ‘the blacks should all sent back to where they came from’ would sometimes add, ‘Except Frank. He’s one of us!’ The heavyweight boxer Frank Bruno (b. 1961) had been born in London to Caribbean parents but had married an English woman, and had a 95% knock-out rate in the forty out of forty-five fights that he won. Despite his genetic dissimilarity from the native English, he fought successfully for England, and displayed the attractive qualities of toughness, physical strength, and general amiability. As such, we can see how otherwise ethnocentric people would see a benefit to their group’s fitness in permitting him into their group, something that could potentially even outweigh (indeed, clearly outweighed) the negative side of this.

  That said, the research by Kurzban et al. does highlight the degree to which ethnocentrism can be affected by environmental variables. The coalitions established by Kurzban et al. would, presumably, not have been under a great deal of stress. The suggestion is that were the coalition placed under stress then it would likely break down into infighting and members would start to categorize and even assort along racial lines — along the lines of genetic similarity.

  7. Other Lines of Research

  Various other studies have validated many of the predictions of Genetic Similarity Theory.

  (A) Academic Citations

  The theory would predict that ethnic nepotism, or the influence of it, would be present in every area of life in which different ethnicities interacted. One obvious example of such an area is academia. Academia is international (especially in the hard sciences) and the medium of publication
is English. Genetic Similarity Theory would predict that academics would be more likely to take seriously research conducted by members of their own ethnicity than by outsiders and would more likely want to assist the credibility of members of their own ethnicity by citing them. This could be tested by examining the propensity for academics to cite people of their own ethnicity (as judged by the academic’s surname) in their research papers. Greenwald and Schuh’s (1994) large scale study of academic social science journals classified citing and cited academics, according to their surnames, as Jewish or non-Jewish. The author’s surname category was associated with a 40% increased likelihood of citing an academic with the same surname category. The authors noted that the overt leftist bias in social science adds credence to the view that this was probably an unconscious process.

  However, it could be countered that it is laborious and moderately more time-consuming having to cite people with foreign names that you don’t recognize and which you might therefore misspell. If you are using the Harvard method of citation, which I use here and which is standard practice outside of the humanities, you place the author’s name in parentheses and then scroll down the document to your reference section to add the reference. If the name is English, and you are English, you don’t have to then check again how the name was spelt and, accordingly, time is saved. It would be interesting to conduct a study like Greenwald and Schuh’s that controlled, for example, for length of the surname or whether or not the author had heard of the surname. Even in writing the above paragraph, I had to check how ‘Schuh’ was spelt, as by the time I scrolled down to my reference section I wondered if it might be spelt ‘Shuh’!

  (B) Trust

  Following Genetic Similarity Theory, we would expect that people would be more likely to trust members of their own ethnic group. This would be because the feeling of trust would facilitate altruism which would boost the individual’s inclusive fitness. In this regard, Salter (2002) has shown that trust, and the risky joint ventures underpinned by trust, are more common within ethnic groups than between ethnic groups. Indeed, Putnam (2007) has shown that in the USA in ‘ethnically diverse neighbourhoods residents of all races tend to ‘hunker down’. Trust (even of one’s own race) is lower, altruism and community cooperation rarer, friends fewer’. Thus, not only do people trust members of different ethnicities less than those of their own ethnicity but the presence of those of a different ethnicity reduces societal trust, as the community is no longer ethnically homogenous, making people — overall — less trusting. It is interesting that immigration makes even the native population trust their own members less. One possible explanation is that any random fellow co-ethnic now has the opportunity to defect and one cannot trust whether or not he will do so. This phenomenon has also been found in Melbourne (Healy, 2007).

  (C) The Welfare State

  We have already noted research finding that people are more likely to give money to beggars of their own ethnicity. Genetic Similarity Theory would predict that this would extend to any form of charity, including the compulsory system of wealth redistribution employed in welfare states. As such, we would expect that ethnic heterogeneity in a society would lead to people resenting their tax money going to non-co-ethnics and, thus, the whole society reducing the level of government expenditure. In this regard, Sanderson and Vanhanen (2004, p. 120) have found that ethno-linguistic diversity explains correlates at 0.3 with not supporting a system of welfare.

  (D) Economic Growth

  Genetic Similarity Theory would predict that ethnic heterogeneity would lead to a lack of social cohesion and consequently society would simply work less efficiently. It would have to deal with ethnic conflicts, organized crime along ethnic lines, more crime (due to lack trust between people) and so forth. Society would be less socially cohesive and we would expect that this would reduce the ability of the government to make rational economic decisions (Alesina et al., 1999). In this regard, it has been found that a nation’s ethnic diversity is generally negatively associated with its extent of economic growth, except in the richest 10% of countries (Masters & MacMillan, 2004).

  (E) Ethnic Conflict

  Genetic Similarity Theory would predict that the more ethnically diverse a nation was then the more ethnic conflict there would be, and the ferocity of ethnic conflict would be proportionate to the degree of genetic difference between the two conflicting ethnic groups, when controlling for socioeconomic factors. As discussed, Tatu Vanhanen has demonstrated that the extent of Ethnic Conflict (EC, defined on a scale up to severe ethnic massacres) correlates with the extent of Ethnic Heterogeneity (EH, defined as differences in visible race, language or tribe, and religion), when controlling for other variables such as socioeconomic development level and level of democracy. Vanhanen found that Ethnic Heterogeneity correlated with Ethnic Conflict within nations at 0.66. He also found that socioeconomic variables were very weak in explaining ethnic conflict. Level of democracy was weakly negatively associated (−0.2) with level of ethnic conflict. Vanhanen also noted that even a tiny degree of ethnic diversity is sufficient to cause low level ethnic conflict, manifested, for example, in the development of democratic but ethnically based political parties. For example, he observes that in his native Finland there is a 5% Swedish-speaking minority and that this has its own political party and institutions. It has been appeased only by Swedish being made Finland’s second official language as well as by various legal mechanisms that privilege the minority. Vanhanen highlights a number of outliers, where ethnic conflict is much lower than would be predicted by the levels of ethnic heterogeneity. He also notes possible reasons for their anomalous status, in particular institutional arrangements which appease ethnic minorities, strong autocracy, and a high percentage of ethnically mixed people. But, in general, it is clear that ethnic heterogeneity will lead to ethnic conflict, just as Genetic Similarity Theory would predict.

  (F) Genomics and Kin Recognition

  Salter and Harpending (2015) have employed the Human Genome Diversity Project database to test Genetic Similarity Theory. The database contains the genotypes of large numbers of individuals based on single nucleotide polymorphisms. The French sample confirms Salter’s (2007) view that two random co-nationals — in this case, French people — are only minimally related. Helping one’s nearest (stranger) kin would be worth only 2% of helping oneself, 4% of helping one’s child and so on. As such, in a small community, there would be little value in placing fellow ethnics before oneself or one’s family. The same is true of a Japanese sample. However, if these two samples were brought together then there is a clear benefit to inclusive fitness to helping a neighbour of the same ethnicity. In this context, a Frenchman would have a 0.06 kinship — equivalent to kinship with a great-grandchild — with any random Frenchman. Thus, offering a transient surplus to a fellow-Frenchman would increase a Frenchman’s own fitness by 12% while offering it to a Japanese person would decrease it by 12%. So, the capacity to racially discriminate — in Malthusian conditions — would lead to a fitness benefit of 24%; a significant difference. ‘Malthusian’ refers to preindustrial conditions wherein the growth of population was exponential and outpaced the production of food. Accordingly, there was a constant struggle for survival with the population level being kept in check by disease, famine, starvation and war. Clearly, this was a situation ripe for Natural Selection.19 In highly selective conditions, traits that favoured ethnic kin discrimination would be rapidly selected for and would spread throughout the population. The effect would be even stronger when the benefit was conferred on aggregated ethnic kinship via a collective good because this would not substantially damage anyone’s individual fitness. We can thus see how this would lead to the extreme ethnic-altruism associated with ethnic conflict, such as suicidal attacks on members of other ethnic groups. But even if groups were not in conflict, there would obviously be a fitness benefit — if one had a surplus, for example — to favouring those who were more
closely related even if not kin, and this would spread throughout the population.

  (G) Living Arrangements

  Clark and Tuffin (2015) surveyed New Zealanders in their early twenties to early thirties who were involved in house-sharing with others of a similar demographic. They note that it is increasingly popular in Western countries for young, single people to ‘house share’ in this way. They found that people tended to prefer to house share with those of the same ethnicity. This is in line with Genetic Similarity Theory as it shows that people are attracted to those who are more genetically similar to themselves and trust them to a greater extent.

  (H) Mothers Who Use Egg Donors

  It has been found that mothers who have babies using donor eggs create weaker bonds with their infants than do mothers whose infants are biologically their own. The mothers who have conceived via donor eggs make less eye contact with their infants and are less responsive to them than are the mothers whose infants are biologically theirs (Imrie et al., 2018).

  8. Conclusion

  We have seen that Rushton’s Genetic Similarity Theory has a great deal of evidence in its favour, allows correct predictions to be made about human behaviour, and stands up robustly to the assorted criticisms which have been levelled against it. The problem with Genetic Similarity Theory is that it is not an all-encompassing theory of differences in ethnocentrism. Although it allows us to understand why ethnocentrism exists, it does not tell us why there are group or individual differences in the extent of it. We will now look at other explanatory mechanisms for variation in the level of ethnocentrism.