Also of interest to note, in Table A-2 when estimating the Baseline model for the white
sub-sample, living in the Southern region and possessing a liberal ideology are not significant
predictors as they were in the model entire sample.
As hypothesized, the fully specified model in column 3 indicates not only that an
individual’s characteristics matter, but that the community they reside in continues to matter and
shape trust levels. Individuals residing in communities with a larger percentage of African
Americans are less likely to trust African Americans, regardless of individual demographic
factors.
Trusting Whites
The level of trust that individuals posses for African Americans is important, however we
are concerned with the trust levels that African Americans possess towards whites. As we
hypothesize, does racial discrimination coupled with racial stigmatization impact African
Americans ability to trust whites? Does the level of trust vary if the individuals are residing in
the South where race relations have historically been the most strained? Are Blacks living in
areas with a large population more or less likely to trust whites?
In this section we repeat the ordered logistic and the HLM models, with a dependent
variable measuring the level respondents trust whites. We measure the level of trust of whites
with the following survey question: ““How much do you trust white people?” with responses
ranging from 4 (trust them a lot) to 1 (trust them not at all).
Alesina and La Ferraa (2000) find
that racial fragmentation and racial inequality are positively correlated and that racial
fragmentation leads to a 12 percent decrease in trust levels between blacks and whites. Racial
fragmentation as a whole decreases the probability of trusting others by three percentage points
in their study, almost 10 percent of the mean value. This is larger than the effect of having a
traumatic experience in the last year, and almost the same size as the effect as being divorced or
19
Convention