Nativity, Ethnicity, and Residential Relocation: The Experience of Hmong Refugees and African Americans Displaced from Public Housing

ABSTRACT:

This article focuses on how the experiences of refugee public housing residents differ from those of other public housing residents when they participate in housing dispersal programs. An analysis of the spatial resettlement patterns and survey responses of Hmong and African-American public housing residents who were involuntarily relocated from public housing in Minneapolis, MN indicates the extent to which residents resettled in ethnically concentrated neighborhoods and their satisfaction with their new housing arrangements and neighborhoods. Research results suggest that Hmong did not settle in ethnically concentrated neighborhoods to the same extent as African Americans after relocation, and experienced lower levels of satisfaction in their new housing arrangements and neighborhoods. These findings indicate that the outcomes for residents involved in housing dispersal programs may depend in part on nativity status.

Notes

1 RCR created on HUD’s website on June 16, 2009 (https://hudapps.hud.gov/public/pic/Mtcsrcr?category=rcr_familystatus&download=false&count=0).

2 We report 1990 Census data here because the vast majority of residents had moved from the Sumner-Glenwood neighborhood at the time of the 2000 Census in anticipation of the redevelopment of the Sumner-Glenwood public housing project. Specifically, the 1990 Census indicates 3,336 residents in the Sumner-Glenwood neighborhood, but the 2000 Census indicates only 144 residents.

3 Of the 50 respondents dropped from this analysis, 4 were white, 2 were Latino, 1 was American Indian, 11 were Somali, 14 were Laotian, 15 were other Asian, and 3 were some other race or combination of races. We elected to drop these observations from our analysis since none of the other groups represented in the initial sample were numerous enough to stand alone in the various statistical analyses we conducted. In addition, since one of our major goals was to test the extent to which the residents relocated to ethnically concentrated neighborhoods, it did not make sense to collapse respondents from varied ethnicities into a common “other” category.

4 Following convention (see Logan et al., 2002) we use census tracts as a proxy for neighborhood. We use the combined central cities of Minneapolis and Saint Paul as our comparison scale. Using the entire metropolitan area would have produced very low thresholds for our enclave measures given the small numbers of Asians and African Americans in the Twin Cities region.

5 These variables related to the respondents’ assessment of the extent of different problems in their current and previous neighborhoods and the respondents’ monthly household income. None of these variables were missing data from more than 16% of observations.

6 The multiple regression method for data imputation uses nonmissing data to predict values for missing data. This method uses stochastic substitution (based on a normal distribution) to add a random value to the predicted value. One drawback to this method of imputation is that it probably reduces standard errors, resulting in smaller confidence intervals and probability values.

7 Some of the variables used as controls in the regression analysis were missing data for one or two observations. Rather than impute these values, we elected to use listwise deletion in these cases. Thus the sample size for some of the regression models was 143 or 144 instead of 145.

8 We use Asian racial categorization as a proxy for Hmong ethnicity because of the relative lack of diversity of the Asian populations in neighborhoods where Hmong settled. Data from the 2000 Census indicate that a majority of Asian residents in Minneapolis and St. Paul is foreign-born. In most of the census tracts where Hmong resettled after moving from Sumner-Glenwood, almost all Asians living in the tract were originally from Southeast Asia and were most likely Hmong. For example, in census tracts where more than one Hmong household resettled (72% of the Hmong households in our sample resettled in a census tract with at least one other Hmong household that had lived in Sumner-Glenwood), Southeast Asians accounted for almost 95% of foreign-born Asians. This ranged from a high of 100% (in one-third of such census tracts) to 81%. In only three such census tracts were Southeast Asians less than 90% of the foreign-born Asians.

9 We do not have access to much objective data on conditions in the site, with one exception. The City of Minneapolis reports that 49 crimes were reported in Sumner-Glenwood during 1999. Normalizing this statistic by population to obtain a crime rate per capita indicates a rate of 0.01 crimes reported per capita. In comparison, in 1999 the crime rate in the neighborhood directly to the North of Sumner-Glenwood was 0.19 crimes reported per capita. Other nearby neighborhoods report crime rates that were similarly higher than the rate in Sumner-Glenwood.

10 The Neighborhood Relocation Index consisted of six items (“How easy was it to find (1) a safe location, (2) a desirable area, (3) an affordable home, (4) adequate space, (5) a location near friends/family, and (6) a location near public transportation?”). The Neighborhood Problem Index consisted of five items (“In your neighborhood, how much of a problem is (1) graffiti, (2) public drinking, (3) drug use, (4) abandoned buildings, and (5) racial intolerance?”). The Neighborhood Socializing Index consisted of two items (“How often do you (1) say hello to your neighbors and (2) talk with a neighbor more than 10 minutes?”). The Neighborhood Satisfaction Index consisted of eight items (“How satisfied are you with (1) bus service, (2) safety, (3) the racial makeup, (4) the proximity to your place of worship, (5) the proximity to your friends, (6) the proximity to healthcare, (7) grocery stores, and (8) parks and playgrounds in this neighborhood?”).