Metro Area Residential Segregation Data: A First Look At Metro NW Areas Leading Into Fair Housing Month.

In advance of the Fair Housing month of April, I have taken a first look at residential segregation data posted on the US 2010/Brown University website HERE.  

The main conclusion from this initial analysis is that NW metro areas (including metro areas in Oregon) had low to moderate levels of Black and Hispanic segregation in 2010 and the vast majority of areas improved from 2000 to 2010. 

The data I focused on was 2000 and 2010 Census data for White-Black and White-Hispanic segregation in metro areas as measured by a dissimilarity [D] index (Wikipedia description HERE). 

The Brown website explains that:

" A high [D] value indicates that the two groups tend to live in different tracts. D ranges from 0 to 100. A value of 60 (or above) is considered very high. It means that 60% (or more) of the members of one group would need to move to a different tract in order for the two groups to be equally distributed. Values of 40 or 50 are usually considered a moderate level of segregation, and values of 30 or below are considered to be fairly low."

The two tables I have prepared HERE:

1. List dissimilarity [D] data for 24 Northwest metro areas.
2. Include the rankings of these NW metro areas for both 2000 and 2010 (Not all metro data is available yet for 2010, so the rankings in the "2010 to date" column signal this ranking is only for 328 metro areas where data is available [ 395 eventual metro areas expected]). Rankings within the NW metro areas are also included.
3. Include the % change in the dissimilarity index from 2010 to 2000. 

Some observations

White-Black/Black-White Segregation

1. In almost ALL cases segregation in NW Metro areas is LESS than the national Metro median to date value of 45.6 for 2010.
2. For 2010, ALL NW D metro values are less than 50; 3 metro area values are in the 30's; and 17 metro D values are in 20's or below. 
3. Bend has one of the LOWEST rates of segregation in the county, ranking 326 out of 328 Metro areas to date in 2010 with a D value of 16.4. 
4. 22 NW metro areas saw a DECREASE in segregation from 2000 to 2010. Every Oregon metro area saw a decrease in segregation from 2000-2010.
5. The two NW metro areas that saw an increase in segregation from 2000 to 2010 still had low D values; Boise-Nampa's D value was 24.6 and the Logan ID-UT D value was 28.7.
6. The Portland metro segregation ranking for 2010 to date is 205th; Segregation declined by 14% from 2000-2010.
7. Within Oregon metro areas, Salem had the highest rate of segregation, and Eugene/Springfield the lowest. NOTE again that segregation in all Oregon metro areas was low to moderate. 

Hispanic-White/White-Hispanic Segregation

1. There were 5 NW metro areas, including Salem in Oregon, where the dissimilarity was GREATER than the National median to date for 2010. However in these areas, EXCEPT for Salem, segregation decreased from 2000 to 2010 and Salem's segregation remained at a moderate D value of 42.
2. There were also 5 NW metro areas where segregation increased from 2000 to 2010, BUT for 2010, ALL NW D metro values are less than 50; 7 metro area D values are in the 30's;and 13 Metro D values are in 20's or below.  
3. Coeur d'Alene has the LOWEST rate of 2010 segregation in the county to date, ranking 328 out of 328 metro areas. [D value was 9.6].
4. Portland's segregation ranking for 2010 to date is 177th.; Segregation remained virtually the same from 2000-2010 (34.2 vs 34.3).
5. Within Oregon metro areas Portland had the highest rate of segregation, and Bend the lowest, NOTE again that  segregation in all Oregon metro areas was low to moderate. 

Caveats: 

1. Dissimilarity is only one measure of segregation. Other measures may produce different results; this analysis should serve merely as a starting point for analysis of 2010 segregation.
2. Segregation within cities in an metro area may differ from metro level segregation. Metro central city segregation may differ from suburban levels of segregation. (City level data is not yet available for analysis).
3. National rankings for metro areas currently only include 328 metro areas to date for 2010. NW metro rankings will change once data for all 395 metro areas is available, and better comparisons between 2000 and 2010 metro rankings can then be made. Metro median D values will also change once all metro data is available.
4. I encourage others, including Fair Housing organizations, to take an independent look at the Brown and Census 2010 data and offer their own "take" on that data. 

Originally created and posted on the Oregon Housing Blog.