sparks & sparks spatiotemporal persistence of residential segregation sssa 2013
DESCRIPTION
Recent research on residential segregation has focused on defining trends in areas outside of traditional metropolitan areas, including so-called new destinations and established immigrant gateways. Moreover a trend of decreasing black-white segregation between 1970 and 2009 has been described by recent work by Iceland and colleagues. In this paper, we go beyond the approach used in this recent work to investigate the county-level patterns of change in residential segregation between 1990 and 2010 using the most recent decennial census data available. We consider both black-white and Hispanic-Non-Hispanic patterns of segregation using measures of three dimensions of segregation: Evenness, Exposure and spatial clustering. Furthermore, we use tools of exploratory spatial data analysis and Geographic Information System (GIS) visualization to highlight areas of the country experiencing the most change over this period. This will allow us to see sub-regional trends in the dynamics of segregation, and understand the nature of segregation beyond the traditional black-white dichotomy, especially in areas of recent Hispanic immigration.TRANSCRIPT
COREY S. SPARKSP. JOHNELLE SPARKS
DEPARTMENT OF DEMOGRAPHYTHE UNIVERSITY OF TEXAS AT SAN
ANTONIO
A Spatial Analysis of Changing Segregation Patterns in the United
States Between 1990 and 2010
Introduction
Between 1970 and 2009, black-white residential segregation declined on average across the country
Recently, Iceland and colleagues (2013) showed that within this time period, the western region of the US showed the greatest decline in black-white segregation, while the Midwest showed the lowest decline
Generally, these trends are for metropolitan areas, with no reference given to non-metro areas
Reasons for Regional Patterns of Residential Segregation
Iceland et al (2013) point out that the size of the minority population itself is often associated with higher levels of segregation Also size of the metro area This is especially seen in the NE and MW
In other areas, more multi-ethnic populations may serve as a buffer to black-white patterns of segregation Historic patterns of black population migration have also affected
the overall temporal trend in segregationAlso, patterns of economic activity and growth in
housing markets have affected segregation patterns Government, military and higher education decrease segregation Suburbanization also decreases segregation
How segregation is measured
MSA versus Place? Parisi and colleagues (2011) point out how racial
segregation can occur at many different levels of geography, not just the MSA
This is primarily because of recent trends of population decentralization into small towns and metro-fringe areas
Macro and micro segregation They go on to show how micro-segregation, or segregation
between areas within cities, only accounts for ~50% of the variation in segregation, while macro, or between place segregation is also important
They go on to point out the importance of considering segregation in smaller communities
Goals
First, we go beyond the approach used in this recent work to investigate the county-level patterns of change in residential segregation between 1990 and 2010 using the most recent decennial census data.
Secondly, we use tools of exploratory spatial data analysis and Geographic Information System (GIS) visualization to highlight areas of the country where segregation has continued to persist, despite the national trend.
Thirdly, we consider non-metro, as well as metro segregation in our analysis, and examine if the described trends hold for these areas as well.
Data
US Decennial Census Summary File 1 1990, 2000, 2010 Population tabulated by Race
Census tract level
NHGIS County Shapefile n=3109 counties, 3 time periods
Methods
Calculate two indices of segregation Evenness – Dissimilarity Index Exposure – Interaction Index
Apply Exploratory Spatial Analysis Methods Getis-Ord G* - Examine spatial clustering of high and
low values of segregation indices
Examine persistence of clustering across time Examine spatial patterns of segregation change
between 1990 and 2010
Spatial Trends in SegregationDissimilarity
Spatial Trends in SegregationDissimilarity
Spatial Trends in SegregationDissimilarity
2010 Dissimilarity G*
Spatial Trends in SegregationChange in Dissimilarity
Spatial Trends in SegregationInteraction
Spatial Trends in SegregationInteraction
Spatial Trends in SegregationInteraction
Spatial Trends in SegregationChange in Interaction
Changes in Segregation Across the Urban-Rural Continuum
Persistence in Segregation Across the Urban-Rural Continuum
Conclusions
Patterns of change Largest changes in segregation occurred in large metro
areas and small urban areas outside of metro counties Smallest changes occurred in most non-metro counties Large metros also were most likely to be persistently
segregated (D>.6) over 1990 to 2010 Most nonmetro areas were most likely to change from
segregated to desegregatedResults support propositions from Parisi and
colleagues Most pronounced changes are in areas outside of
traditional metro areas Micropolitan population changes