In many countries of the world the extent of ethnic, cultural and racial diversity is increasing as a result of international migration. One frequently voiced concern is that as populations become more diverse, socioeconomic inequalities within countries will become greater. The present study presents new measures of ethnic, racial and religious diversity for 198 countries and territories. These measures were used as predictors of the Gini index in regression models with other predictors. Ethnic and religious diversity were found to be virtually unrelated to the Gini index. However, a high level of racial diversity independently predicts a high Gini index. The effect of racial diversity is robust and persists when controls for spatial and cultural autocorrelation are introduced.
Key Words: Income inequality; Gini index; Multiculturalism; fractionalization; Ethnic diversity; Racial diversity; Religious diversity; IQ.
(ProQuest-CSA LLC: ... denotes formula omitted.)
The ethnic, racial, cultural and religious diversity of national populations varies over a wide range. At one extreme are countries such as Sweden and Norway which are, or were until recently, very homogeneous on most or all of these dimensions. Other countries are extremely diverse. For example, we find very high levels of ethnic diversity in most countries of subSaharan Africa, and very high levels of racial diversity in most Latin American countries.
The possibility that aspects of cultural, ethnic or racial diversity are related to the level of income inequality in the country derives from two lines of reasoning. One hypothesis states that social solidarity is greatest in countries with a homogeneous population. It has been shown repeatedly that in private life, people associate preferentially with partners who are similar to themselves (DeBruine, 2002). According to genetic similarity theory (Rushton, 1989; Rushton and Bons, 2005), this tendency evolved as an extension of kin-selected altruism. Because the benefits of friendly cooperation accrue not only to the agent but also to his partner, the genetic payoff of cooperation is greater with a partner who shares some of one's own genes than with a stranger. According to Rushton this applies not only to relatives but also to those who are perceived as phenotypically similar to oneself.
In addition to a personal preference for people who are similar to oneself, humans identify with the groups to which they belong. Richerson and Boyd (2001) describe ethnocentrism as arising from "tribal instincts," which helped our ancestors to survive in a world that was dominated by lethal conflict between human groups. A preference for people of one's own ethnic or cultural group has been observed in places as diverse as the United States (Alesina and La Ferrara, 2002), Russia (Bahry et al., 2005) and Papua New Guinea (Bernhard et al., 2006).
Ingroup solidarity does not necessarily lead to outgroup hostility (Brewer, 1999; de Figueiredo and Elkins, 2003), but both are found universally in human societies. In the United States, the participation in social activities is reduced by high levels of ethnic and especially racial diversity in the community (Alesina and La Ferrara, 2000). If cultural, ethnic or racial divisions within a country or community reduce social solidarity, a likely consequence will be unwillingness of the rich to share their wealth with the poor, at least in cases where groups differ noticeably in wealth.
Indeed, at least one survey in the United States found that public support for welfare spending diminishes as the share of local welfare recipients from other racial groups rises (Luttmer, 2001). This preference does seem to affect policies at the community level. In urban areas, the shares of spending on productive public goods such as education, roads, sewers and trash pickup were found to be inversely related to the community's ethnic fragmentation (Alesina et al. …