Immigrating, Assimilating, Cashing in? Analyzing Property Values in Suburbs of Immigrant Gateways

Abstract

Assets like properties determine opportunities. Many immigrants have begun to bypass cities and move straight to suburbs. Until the recent house price crash, suburbs had been perceived as locations where appreciation rates were high, but this perception might no longer hold true. Not many studies have focused on suburban house prices with regard to race, ethnicity, and nativity. This study fills the gap in terms of nativity. This study uses data from the 2000 US Census and the 2005–2009 American Community Survey to perform regression analyses to analyze immigrant gateways as delineated by Singer (2008) with regard to median values of owner-occupied housing units and the factors that influence median values, while also differentiating between inner cities and suburbs. Results show that there are differences in terms of values and appreciation rates among suburbs of immigrant gateways, indicating different economic integration patterns.

Keywords:

• suburbs
• suburban
• house prices
• immigrants

Acknowledgements

The author thanks participants of the Association of Collegiate Schools of Planning (ACSP) conference in July 2008 in Chicago, IL and of The Diverse Suburb: History, Politics, and Prospects conference in October 2009 in Hempstead, NY, in particular Nancy Denton and Andrew Wiese, for constructive comments.

Notes

¹ The terms ‘foreign born’ and ‘immigrant’ will be used interchangeably in this study. Technically, ‘immigrant’ refers to someone who was awarded legal permanent resident status in the USA (Chiswick & Sullivan, 1995).

² Homeownership rates for many but not all Asian groups are similar to those of non-Hispanic Whites (Bankston and Hidalgo, 2007; Listokin & Listokin, 2001; Painter and Yu, 2003, 2004).

³ The author thanks Referee 1 for making this point.

⁴ The Public Use Microdata Sample data set does not provide property values.

⁵ Baltimore-Towson, MD; Buffalo-Niagara Falls, NY; Cleveland-Elyria-Mentor, OH; Detroit-Warren-Livonia, MI; Milwaukee-Waukesha-West Allis, WI; Philadelphia-Camden-Wilmington, PA-NJ-DE; Pittsburgh, PA; St. Louis, MO-IL (Singer, 2008).

⁶ Boston-Cambridge-Quincy, MA-NH; Chicago-Naperville-Joliet, IL-IN-WI; New York-Northern New Jersey-Long Island, NY-NJ-PA; San Francisco-Oakland-Fremont, CA (Singer, 2008).

⁷ Houston-Baytown-Sugar Land, TX; Los Angeles-Long Beach-Santa Ana, CA; Miami-Fort Lauderdale-Miami Beach, FL; Riverside-San Bernardino-Ontario, CA; San Diego-Carlsbad-San Marcos, CA (Singer, 2008).

⁸ Atlanta-Sandy Springs-Marietta, GA; Dallas-Fort Worth-Arlington, TX; Las Vegas-Paradise, NV; Orlando, FL; Phoenix-Mesa-Scottsdale, AZ; Washington-Arlington-Alexandria, DC-VA-MD (Singer, 2008).

⁹ Denver-Aurora, CO; Minneapolis-St. Paul-Bloomington, MN-WI; Portland-Vancouver-Beaverton, OR-WA; Sacramento-Arden Arcade-Roseville, CA; San Jose-Sunnyvale-Santa Clara, CA; Seattle-Tacoma-Bellevue, WA; Tampa-St. Petersburg-Clearwater, FL (Singer, 2008).

¹⁰ Austin-Round Rock, TX; Charlotte-Gastonia-Concord, NC-SC; Greensboro-High Point, NC; Winston-Salem, NC; Raleigh-Cary, NC; Durham, NC; Salt Lake City, UT (Singer, 2008).

¹¹ Obviously, there is variety in the foreign-born proportion per Census tract. Ideally, one would create a threshold to include only Census tracts that have a certain proportion of foreign-born residents. Alternatively, one would compare Census tracts that have low proportions with Census tracts that have high proportions of foreign-born residents. Unfortunately, either approach limits the number of observations for inner-city and suburban tracts, resulting in possible selection biases.

¹² The vast majority of the immigrant gateway types are located in only one or two of the four Census regions. Thus, it was not possible to include a single Census region variable in a standardized (i.e., comparable) regression model for the regression models of the different immigrant gateway types.

¹³ Heteroskedasticity is a situation where there are considerably unequal variances in the dependent variable for the same values of the independent variable. Thus, the statistical assumption of equality of variances is violated. Standard errors are biased, t- and F-tests are invalid, and coefficients with a high variance are inefficient (Hamilton, 1992; Vogt, 2005). In order to address these statistically undesirable issues, weighted least squared regressions were run.

¹⁴ The data sets do not provide the arrival date of immigrants, utilized in Clark's (2003) model at the state level. Nevertheless, the data sets provide the following variables related to immigrants: persons of Mexican origin; persons of Puerto Rican origin; persons of Cuban origin; persons of Dominican origin; persons of Guatemalan origin; persons of Honduran origin; persons of Nicaraguan origin; persons of Panamanian origin; persons of El Salvadorian origin; persons of other Central American origin; persons of Colombian origin; persons of Ecuadorian origin; persons of Peruvian origin; persons of other South American origin; persons of other Hispanic/Latino origin. Future research efforts will focus on these groups in immigrant gateways.

¹⁵ The author thanks Referee 2 for making this point.