science can explain but not undo the works of the reptile brain...,

phys.org |  Figure 2 now shows on the horizontal axis the share of total income inequality due to differences between racial groups. Under this dimension, the two cities turn out to be actually very different. The share of total inequality due to differences between races is twice as large in Houston as in San Francisco. This in turn is related to the level of trust in the two cities. In San Francisco, where the probability of meeting an individual of a different race but similar income level is relatively high, the level of trust is higher than in Houston, where belonging to a different race is also likely to be associated with a difference in income.

This same pattern of apparent similarity, which is in reality masking an additional dimension of heterogeneity, is repeated over different pairs of cities in the US My empirical analysis documents the pattern in a systematic way, exploiting answers from 20,000 respondents to the US General Social Survey (GSS) between 1973 and 2010. The survey contains a variety of indicators on the respondents' political views, social behavior and socioeconomic characteristics. Crucially, it also asks respondents whether they think that most people can be trusted. I match their answers to this question to their socioeconomic and demographic characteristics, and to the level of racial diversity, total income inequality and racial income inequality in the MSA of residence.

I start out by showing that racial diversity and total income inequality have a statistically significant, negative effect on individual measures of trust, a result that is consistent with previous studies. But I then find that these effects become statistically insignificant once I account for the income inequality between racial groups, which instead remains negatively and significantly associated to the level of trust of the respondent.

I then show that the negative impact of racial income inequality on trust is larger in more racially fragmented communities, and that members of minority groups reduce their trust towards others more, when racial income inequality increases. These results are consistent with a simple framework in which individuals can be similar in both race and income, and trust towards others falls at increasing rates as individuals become different in both dimensions.

Overall, my results suggests that racial diversity is more detrimental when associated with income disparities between races and that, similarly, income inequality is more harmful when it has a marked racial connotation. This in turn suggests that policies aimed at reducing income disparities along racial lines might be particularly effective in increasing the level of social participation and trust in US communities.