CensusScope is a product of the Social Science Data Analysis Network.
ABOUT MEASURES OF EXPOSURE
Exposure is an additional way of describing the
relationship between races. Unlike the dissimilarity index
is a comparison of the degree of segregation between two groups, when we
consider exposure we examine how the composition of the average metro
city) resident's neighborhood varies according to that person's race.
Broadly speaking, exposure measures a given group's exposure to all racial
groups, including itself, in the form a weighted average depicting the
racial composition of the neighborhood of the average person of a given
race. This can then be compared to the racial distribution for the city
or metro area as a whole in order to examine whether races are relatively
evenly spread out over the city or, conversely, clustered into
neighborhoods where one racial group tends to be disproportionately
represented while others are underrepresented.
In the table below, the first five columns represent
the average racial composition of the neighborhood of a person of a given
race. The rightmost column shows the racial composition of the metro area
or city as a whole.
Under conditions of perfect integration, the racial
distribution for a given neighborhood or census block will be the same as
for the city as a whole. Typically, however, this is not the case. A more
likely scenario is the one described in the graphic above, where white
residents tend to live in neighborhoods in which there are a
disproportionate number of whites, black residents tend to live in
neighborhoods where there are disproportionate numbers of blacks and so
on. This means that, in most cases, the exposure of whites to whites and
of blacks to blacks will be higher than the degree of exposure suggested
by the total metro racial distribution. At the same time, the exposure of
whites to blacks and other minorities will often be lower than the degree
suggested by the total metro racial distrubution.
For example, consider a hypothetical city that has 1500
white residents and 500 black residents for a racial distribution of 75%
white and 25% black.
The city is then divided into two equally sized
neighborhoods. 900 whites and 100 blacks live in the first
neighborhood, while 600 whites and 400 blacks live in the second
neighborhood. The racial composition of the first neighborhood is 90%
while and 10% black, and the racial composition of the second
neighborhood is 60% white and 40% black.
The exposure measures used on CensusScope.org take into
account two factors: first, what portion of the metropolitan area's total
population of a given racial group lives in a specific neighborhood, and
second, the racial makeup of that neighborhood. The concentration of the
given racial group in a neighborhood affects how heavily the racial
composition of that neighborhood will be weighted when calculating the
exposure index for the metro as a whole.
In our example, 60% of the total white population has
the experience of living in a neighborhood that is 10% black, and 40% of
the white population has the experience of living in a neighborhood that
is 40% black.
To calculate exposure, we take into account the
experiences of the entire white population, while weighting the
experiences according to what percent of the white population has them.
The more people who have a given experience, the more strongly it will be
weighted in the exposure index. In our instance, since fully 60% of the
white population lives in a neighborhood that is 10% black, that will be
weighted more heavily in the index than the experience of the other 40% of
the white population who live in a neighborhood that is 40% black. The
average white person (who represents no single individual, but rather a
composite of the total white population) in our hypothetical city lives in
a neighborhood that is 22% black.
Since exposure paints a complete picture of the
racial makeup of a neighborhood, they will always sum to 100%, and can
include as many racial or ethnic groups as desired.
Exposure between two groups can be derived using the
where P1 = city -wide population of Group 1
P1i = neighborhood i population of Group 1
P2i = neighborhood i population of Group 2
Ti = neighborhood i total population
n = number of neighborhoods in city