MacArthur SES & Health Network
MacArthur SES & Health Network


printable version

Residential Segregation

Summary of measures provided by Dolores Acevedo-Garcia, Kimberly Lochner, and Ichiro Kawachi to the Social Environment working group. Last revised December, 2001.

Chapter Contents

  1. Measurement Approaches
  2. References

Measurement Approaches

The table below provides an overview of approaches for measuring residential segregation, poverty concentration and related concepts in sociological literature.

Concept name Concept definition (measure) Source

Metropolitan area economyy
Metropolitan area mean income Average household income (Jargowsky, 1997)
Metropolitan area income inequality Coefficient of variation of the household distribution of income, i.e. standard deviation of household income divided by mean household income (Jargowsky, 1997)

Geographic concentration of poverty
MA poverty rate % of total number of MA persons that is poor (below federal poverty line) (Massey and Denton, 1993)
High poverty neighborhood Neighborhood where poverty rate³ 40% (Jargowsky, 1997)
Neighborhood poverty rate (NPR) % of MA total population that resides in high-poverty neighborhoods (Jargowsky, 1997)
Concentration of poverty (affluence) % of MA poor (affluent) population that resides in high poverty (affluence) neighborhoods (Jargowsky; Waitzman and Smith, 1998)
Poverty concentration Exposure to poverty across neighborhoods (based on isolation index)

For a given MA (j) and a racial/ethnic group (m)′, the index of exposure to poverty (EPjm)′, is given by

where xi jm, ti j and Xjm are the number of members of the group m in census tract i, the total population of census tract i, and the number of members of the minority group m for the entire MA j; eji is the number of persons living in poverty; N is the total number of census tracts in MA j. For example, EPjm = 0.15 indicates that in MA j, the typical member of group m lives in a census tract where 15% of the population lives in poverty.
(Massey and Denton, 1993)

back to top


Residential segregation by race/ethnicity
Dissimilarity The dissimilarity index (D), which may be interpreted as the proportion of the minority racial/ethnic group of interest (m) that would need to move across sub-units in order to achieve an even distribution, is given by

where ti and xi are the total population and minority proportion of areal sub-unit (i.e. census tract) i, and T and X are the population size and minority proportion of the whole geographic area, i.e. MA (j), which is subdivided into N areal sub-units.

Ranges from 0, no residential segregation, to 1, complete residential segregation.
(Massey and Denton, 1988)
Isolation The isolation index (P), which measures the extent to which a member of a racial/ethnic group (m) is likely to be in contact with members of this same group (as opposed to members of other groups), is given by

where x, X and t are defined as above; e.g. Pjm=0.6 indicates that in MA j, the average member of group m lives in a census tract where the probability that (s)he will have contact with another member of group m is 0.6.

Ranges from the overall proportion minority in the entire MA, no residential segregation, to 1, complete residential segregation.
(Massey and Denton, 1988)

back to top

Concentration Concentration refers to the relative space occupied by a minority group in a geographic area. If a group occupies a small share of the total area, it is said to be residentially concentrated. A simple concentration index (C) can be derived from an application of the dissimilarity index defined above

where xi and Xi are defined as before; and ai equals the land area of sub-unit i and A is the total land area of the geographic area (i.e. MA) j. This index may be interpreted as the share of minority members that would have to move across sub-units in order to achieve a uniform density of minority members over all units.

Ranges from 0, no residential segregation, to 1, complete residential segregation.

Massey and Denton have proposed two more complex indices of concentration (the absolute concentration index, ACO, and the relative concentration index, RCO).
(Massey and Denton, 1988)
Centralization Centralization refers to nearness to the center of the urban area, which in the largest and oldest US MAs is often characterized by dilapidated housing and socioeconomic deprivation. The absolute centralization index is given by

where the N areal sub-units are ordered by increasing distance from the central business district, C is the cumulative proportion of X in sub-unit i, and A is the cumulative proportion of land area through sub-unit i.

Ranges from 1 to -1. Positive values in indicate tendency of group X to live close to the center of the MA; negative values indicate tendency to live in the outlying areas; 0 denotes a uniform distribution throughout the MA.
(Massey and Denton, 1988)
Clustering Clustering is the extent to which areal sub-units inhabited by minority members adjoin one another, or cluster, in space. The preferred measure of this dimension is the index of spatial proximity (SP) given by the average of intergroup proximities (Pxx, Pyy)

where T, X and Y are the population size, minority proportion, and majority proportion of the whole geographic area. To illustrate, the measure of spatial proximity for group X, i.e. the average proximity between members of group X, is given by

where cij represents a distance function between areas i and j; and x and X are defined as before.

SP equals 1, when there is no differential clustering between X and Y, and is greater than 1 when members of each group live closer to one another than to each other.
(Massey and Denton, 1988)

back to top


Residential segregation by class
Neighborhood sorting index (NSI) and (Jargowsky, 1997)
Neighborhood distribution of
income
where is the standard deviation of the neighborhood income distribution and is the standard deviation of the household income distribution. The neighborhood distribution of income is the distribution of households by the mean household income of the neighborhood in which they live—each neighborhood is weighted by the number of household it contains. The household distribution of income is the distribution of households by their own income. If there was no income segregation, all neighborhoods would have the same mean income and , and thus NSI, would be zero. At the other extreme, if there was perfect income segregation, all households would live in neighborhoods where the mean income approximates their own. In this case, would approach
, and thus NSI would approach 1.

NSI² is the proportion of the total variance in household income among rather than within neighborhoods.
 
Dissimilarity of the poor (affluent)¹ Proportion of poor (affluent) families that would have to move in order to achieve an even socioeconomic distribution throughout the metropolitan area.

See dissimilarity index above

Thresholds:

Poor: below federal poverty line

Affluent: $75,000 (top 12% of the 1989 family income distribution)
(Waitzman and Smith, 1998)
Isolation of the poor (affluent)¹ Exposure to poverty (affluence) across neighborhoods (defined by Massey and Denton as poverty concentration).

See isolation index above

Thresholds:

Poor: below federal poverty line

Affluent: $75,000 (top 12% of the 1989 family income distribution)
(Waitzman and Smith, 1998)

1 The dissimilarity and the isolation index are not independent of the mean and variance of the income distribution. For this reason, Jargowsky has proposed the use of NSI and NSI².

back to top

References

Jargowsky, P.A. (1997). Poverty and place: ghettos, barrios, and the American city. Russell Sage Foundation, New York.

Massey, D.S. and N.A. Denton (1988). The dimensions of residential segregation. Social Forces, 67:281-315.

Massey, D.S. and N. A. Denton (1993). American apartheid: Segregation and the making of the underclass. Harvard University Press, Cambridge, Mass.

Waitzman, N.J. and K.R. Smith (1998). Separate but lethal: the effects of economic segregation on mortality in metropolitan America. Milbank Quarterly, 76:341-73.

Addendum

Dolores Acevedo-Garcia and Kimberly Lochner have written a chapter on residential segregation and health to appear in Neighborhoods and Health (Ichiro Kawachi and Lisa Berkman, Editors) to be published by Oxford University Press. The chapter provides a review of the area for the reader who wishes more in depth coverage of the subject than is provided in the Web site chapter. A pdf is available of the chapter and of the chapter appendix (which contains the information in this web chapter); the authors request that no citations be made from either pdf.

 

MACSES
UCSF Home About UCSF Search UCSF UCSF Medical Center