Attitudes toward blacks by race/ethnicity

Drug Use Trajectory participants were asked: "Based on your experience, a person could be counted on to do what they say if that person was Black." Answers ranged from "strongly agree" (1) to "strongly disagree" (5). Here are the means by race/ethnicity (sample size = 1,795):

Mean "can't count on blacks" score

Whites 1.61
Hispanics 1.90*
Black Hispanics 2.28*
Blacks 2.88*

*significantly higher than whites

Blacks--both Hispanic and non-Hispanic--are much less likely to agree that blacks can be counted on. The gap in attitude between blacks and whites is huge--1.1 standard deviations. Whites have nicer views of blacks than Hispanics do, and they have a much more positive view of blacks than blacks have of themselves. So much for the idea that white attitudes make blacks feel bad about themselves.

A very strong association between percent white and trust

As a follow-up to the last post, I calculated across the nine General Social Survey regional divisions the correlation for the past decade between trusting people in general and the percent of the population that is white. It is .82--a very strong link.

Falling level of trust

General Social survey respondents have been asked the following question each survey year since 1972: "Generally speaking, would you say that most people can be trusted or that you can't be too careful in life?"

The graph shows that trust among Americans has been slipping for four decades (sample size = 34,790). The red section shows the percent who say that people can be trusted: this has dropped from around 50 percent in the early 70s to less than 40 percent currently. The trend has paralleled the post-1965 ethnic diversification of the country.

Falling trust is not a good sign since a flourishing society depends on a certain amount of confidence that contracts can be made and will be honored.

I'm putting together a cross-national data set and plan to identify the social correlates of ethnic diversity. I predict the results will not be pretty.

Race, gender, and trust

MIDUS respondents were asked if they agreed with this statement: "I know that I can trust my friends, and they know they can trust me."  Answers ranged from strongly disagree (1) to strongly agree (7). Here are the means (N = 3,997):

Mean trust score

White men 5.16
Black men 4.89
White women 5.33 
Black women 4.59*

*significantly lower than white women

The real gap here is between white and black women: it's almost two-thirds of a standard deviation, a large difference. White women, on average, are much more trusting of their friends.

Religion and trust

N = 8,866

N = 4,612


In What's So Great about Christianity?, Dinesh D'souza claims that religion makes people more trusting. Trust is an important ingredient in a well-functioning society. There's has been a lot of discussion in the Steveosphere about how diversity undermines trust in your neighbors, even among members of the same ethnic group.

The General Social Survey offers little evidence for D'souza's contention. In the top graph, you do get a little bump among people who attend church nearly every week or more frequently, but the lower graph on belief in a deity shows that those who know God exists are the least trusting category.

I'll have to look at it, but I doubt there is a positive cross-national correlation between religiosity and trust.


UPDATE: According to this study of 105 countries, religiosity is negatively correlated with trust.

Like we're surprised--Northern states are most trusting

Gallup asked almost 180,000 Americans if they would trust someone who found their wallet to return it. The darkest states on the map indicate the most trusting states. Gallup concluded that states with higher of levels of well-being and access to services are more trusting. They seem to be hinting that collectivism breeds trust.

Collectivist California and New York are two of the most untrusting states in the country. And D.C. has half the level of trust seen in Utah--that bastion of liberalism.

It goes with out saying that percent white and ethnic homogeneity were not mentioned.

Race, ethnicity, and trust: Americans (GSS respondents) were asked if people can be trusted. Here are the percent who answered yes by ethnic/racial group:


Percent trusting (N = 27,313)

Norwegian 56.0
Finnish 54.4
Swiss 53.8
Danish 52.3
Belgian 52.1
English 51.1
Swedish 50.9
Scottish 48.3
Yugoslavian 48.4
Russian 47.6
Jewish 45.7
Irish 45.4
French 45.2
Japanese 44.4
Chinese 44.2
German 44.0
Czech 44.0
Austrian 43.6
Hungarian 42.9
Polish 42.8
Dutch 41.2
French Canadian 40.9
Lithuanian 40.5

All Americans 39.6

Greek 39.3
Italian 38.2
Arab 36.4
Portuguese 34.5
Asian Indian 32.8
Rumanian 31.9
Spain 31.6
Amerindian 26.2
West Indian 26.1
Mexican 24.0
Filipino 21.3
Black 16.7
Puerto Rican 16.2


The pattern is very clear for Americans of differing ancestries: those of northwestern European descent are the most trusting; Asians and central and eastern Europeans are in the middle; southern Europeans, Asian Indians, and Arabs are low; Amerindians, Mexicans, Filipinos, blacks, and Puerto Ricans are the least trusting.

Differences are large. For example, Americans of Norwegian descent are more than three times as trusting as black Americans.

In a recent analysis, I showed that trust in the political process is associated with greater interest in politics, and in the comments TGGP linked to a post by n/a which reveals the same kind of European and American pattern in civic engagement that we see with the trust numbers above.

In the Big Five factors of personality, trust is a subcomponent of the broader personality trait Agreeableness. In a study that removed the heritable component of Agreeableness from Trust, a substantial unique heritability remained: .31, and .50 after reliability was corrected. (Jang et al. 1998. Heritability in facet-level traits in a cross-cultural twin study. Journal of Personality and Social Psychology, 74, 1556-1565.)

Trust and wealth are correlated: Recently, I used World Values Survey data to estimate the level of trust in countries in several regions of the world. But how close is the connection between trust and economic success? I calculated the correlation between the percent saying that people can be trusted and per capita GDP (taken from Nationmaster) across 82 countries. It's .45, and .51 if we omit Luxembourg (which is a real outlier on GDP).

It almost looks like a case of necessary but not sufficient. Several poor countries are trusting, but of the wealthy countries, only Singapore, Finland, and France are distrustful. And it could be that in some of these poor countries that are very different culturally from the West, there may be translation problems, or respondents might be giving socially acceptable answers.

More on trust: As a follow-up to the last post, I wondered about the level of trust in Asian and Muslim countries. Based on World Values Survey data, here are the percent who say that people can generally be trusted:


Percent who say you can trust people

World 29.6

Iran 65.3
Saudi Arabia 53.0
Indonesia 51.6
Iraq 47.6
Egypt 37.9
Pakistan 27.9
Morocco 23.5
Bangladesh 22.2
Azerbaijan 20.5
Turkey 12.6
Algeria 11.2

China 55.2
Japan 42.9
Taiwan 38.2
Vietnam 41.1
India 37.9
South Korea 32.1
Philippines 7.0

Muslim countries are all over the place--they are about as diverse as the whole sample of countries. Numbers like that in Iraq surprise me, given what Steve Sailer has written about how consanguinity in Muslim countries undermines trust beyond the family.

India is the trusting one among South Asian countries. The Phillipines is the exception in Southeast Asia, and East Asia is trusting.

NAMs don't trust: Hispanics and especially blacks tend to be suspicious people. According to the GSS, 40% of all Americans say that people can generally be trusted, but only 17% of blacks, 24% of Mex-Ams, and 16% of Puerto Ricans feel that way. Many of them are quick to think that someone is taking advantage of them; that someone is discriminating against them. Now, this could be due to their experience in America. Being dominated by another people could easily lead to distrust. Or it could be an ethnic trait.

The World Values Survey asked people all around the world the same question. Thirty percent of all respondents agreed that most people can be trusted. Here are the percentages who agreed in countries with high percentages of blacks or Hispanics:


Percent who say people can be trusted

Dominican Republic 26.4
Nigeria 22.1
South Africa 19.1
Zimbabwe 11.9
Tanzania 8.1
Uganda 7.6


Mexico 28.9
Chile 22.5
Argentina 20.3
Venezuela 14.8
El Salvador 14.6
Puerto Rico 12.4
Columbia 10.8
Peru 8.2

All black and Latino countries sampled have below-average numbers. Now, some might explain this lack of trust as the legacy of colonialism or a reaction to exploitative social conditions, but whatever the case, it seems to be a widespread, reliable characteristic. This does not bode well for long-term race relations in America. Trust is an essential ingredient for a thriving society.

Are Jews the most distrustful of all American ethnic groups? After reading Steve Sailer's insightful piece tonight contrasting the Cuban- and Jewish-American lobbies, I wondered if Jews as a group are more suspicious of people, and perhaps lack of trust comes from a history of negative relations with others. As an example, Ashley Montagu once said, if you meet a gentile, it's a good operating hypothesis that he is an anti-Semite.

The General Social Survey asked 26,201 Americans if people can be trusted. I list the percent who said they cannot be trusted by ethnic group:


Percent who say that people cannot be trusted

Puerto Ricans 79.2
Blacks 79.0
Other Spanish 73.1
Mexican 70.7
Filipino 70.4
Amerindian 69.9
West Indian 69.5
Spanish 63.3
Asian Indian 59.6
Portuguese 58.5
Italian 57.0
Arab 56.9

All Americans 55.9

French Canadian 55.0
Dutch 54.7
Lithuanian 54.3
Greek 54.1
Hungarian 52.6
German 52.2
Polish 51.9
Irish 50.7
French 50.5
Austrian 49.4
Czechs 49.0
Belgian 48.8
Japanese 48.7
Chinese 47.8
Scottish 47.6
Russian 47.3
Romanian 47.2
Jewish 46.9
English/ Welsh 44.6
Danes 43.4
Swiss 41.1
Finns 40.7
Norwegian 39.6


Jews are not distrustful--just the opposite. They are at the bottom of the list, along with those whose ancestors are from northern Europe.

At the top of the list, there are a number of poor and non-white minority groups. The only nonwhites in the bottom half are Japanese and Chinese American--wealthy minority groups like American Jews.

People from southern Europe tend to be distrsutful, but the really striking pattern in my mind is the high scores of all the Hispanic groups. Francis Fukuyama has argued that social trust is necessary for economic prosperity, and you can see that the high-trust groups are successful. The suspiciousness of Hispanics may prevent them from developing the kind of cooperativeness needed for upwardly mobility. This is one more fact that leads one to conclude that mass immigration from the South means more long-term poverty for America.

Who's more distrustful: smarts or dulls? I'm aware of two opposite images of smart people: 1) the toughminded cynic who is suspicious of others; and 2) the airhead who always seems naive about people (one incident of this was E.O. Wilson being thoroughly surprised at the violent reaction to his Sociobiology).

The General Social Survey asked 1,302 respondents if they think of themselves as a trusting person. Answers ranged from "very trusting" (=1) to "very distrustful" (=4). Here are the means by IQ level:


Mean distrust score by IQ level

High IQ (125) 1.82
Medium IQ (98) 1.50
Low IQ (71) 1.31

SD .67


The first image is supported here: Intelligent folks are almost a standard deviation more suspicious of others than slow people. As Erving Goffman claimed, people present an image of themselves to others that is better than the reality. Our public images are dishonest, we're all liars about ourselves, and dumb folks buy it while smarter people see through it.