Hispanic Location Patterns in Los Angeles, Miami and Dallas: Testing the Spatial-Assimilation Model Emília Paiva-Turra SYS 502  Spatial Data Analysis  Term Paper Prof. Tony Smith  Spring 2003 University of Pennsylvania  Department of City and Regional Planning 1 Table of Contents 1.INTRODUCTION 2 2.BACKGROUND 3 THE SPATIAL-ASSIMILATION MODEL 3 A TYPOLOGY BASED ON HISPANIC POPULATION 5 3.DATA 8 4.METHODS 9 G* AND GI* STATISTICS 9 GLOBAL AND LOCAL TESTS OF SPATIAL CONCENTRATION 12 COMPLEMENTARY TESTS: CHI-SQUARE AND ANOVA 13 5.RESULTS 13 6.SUMMARY AND CONCLUSIONS 17 REFERENCES 21 MAPS 22 2 1. Introduction The Hispanic1 population in the United States has grown fast over the past decades, becoming the largest minority group by 2000 (Suro 2002). Historically Hispanic migration has targeted few metropolitan areas – mainly Los Angeles, New York, Chicago and Miami. However this pattern has changed over the past two decades and cities with traditionally small Hispanic population have started to see a large influx of Hispanic immigrants (Frey and Farley 1996; Suro 2002). For decades researchers have attempted to understand the factors that influence the residential location patterns of minority groups, trying to explain the forces behind ethnic residential segregation (e.g. Massey 1985). The classic spatial-assimilation theory states that new immigrants tend to locate where they have social networks, creating ethnic enclaves through an ‘invasion’ and ‘succession’ process. Over time, acculturation and social mobility allow them to move into better areas, and they get assimilated by the native population (Massey 1985; Massey and Denton 1988, 1991; Alba et al. 2000; Lobo et al. 2002). However, mass migration can potentially cause significant changes in residential patterns (Alba et al. 2000), altering the invasion-succession-assimilation process (Denton and Massey 1991; Alba et al. 2000). Hispanic location patterns typically follow the spatial-assimilation model (Alba et al. 2000; Lobo et al. 2002). It remains uncertain whether the recent heavy migration has altered the traditional location patterns of Hispanics in the U.S. 1 In this paper I use the terms “Hispanics” and “Latino” interchangeably, reflecting both their popular use and the Office of Management and Budget (OMB) terminology standards for Census 2000. It is important to notice that this is an aggregation of people that share a common language but have very diverse birthplaces, national origins, legal status, socioeconomic class, and settlement histories (Suro 2002). 3 The objective of this paper is to verify if the spatial assimilation of Hispanics has changed in the face of strong immigration. I compare the location patterns of U.S. born Hispanics and recent Hispanic immigrants among metropolitan areas with high and low Hispanic growth rates. To test if the spatial-assimilation model explains Hispanic location patterns in eight selected Primary Metropolitan Statistical Areas (PMSA), I perform local and global tests of spatial concentration, using G* and Gi* statistics (Getis and Ord 1992). If the spatial-assimilation model explains Hispanic location patterns, we expect native Hispanics to be less concentrated in enclaves than Hispanic immigrants in all study areas. If native Hispanics are not less concentrated in enclaves than immigrants, the ‘assimilation’ stage predicted by the theory may not be happening. Also, if location patterns are different between traditional Hispanic gateways and places that only recently started receiving large inflows of Hispanics, the spatial-assimilation theory may not be holding under conditions of heavy migration. Whether Hispanic location patterns still follow the spatial-assimilation theory under intense immigration has important implications. Inequalities among residential neighborhoods are one of the key aspects of the American social system (Alba et al. 2000), and one of the major challenges of urban planning. Understanding the factors that drive the residential segregation of Hispanics is fundamental to support policies that will bridge the gap between this ethnic minority and the majority group. 2. Background The Spatial-Assimilation Model The spatial-assimilation model states that immigrants initially settle in inner-city disadvantaged areas, because of their lack of resources (Alba et al. 2000; Lobo et al. 4 2002). As they become established, immigrants create a social network for relatives and friends, starting a migration chain that concentrates ethnic groups into specific neighborhoods (Massey and Denton 1987). The increase of immigrants sparks a systematic out-migration of the native population, who perceives a social distance between themselves and the incoming group (Massey 1985). Eventually, the new population outnumbers the original one, creating an immigrant enclave (Lobo et al. 2002), and completing the ‘invasion’ and ‘succession’ stages described by the model (Denton and Massey 1991). Over time, as immigrant families experience social mobility and acculturation, they will leave the enclaves, searching for better amenities and services, which are usually available in majority dominated neighborhoods (Massey and Denton 1988). According to the model, this ‘assimilation’ stage is a natural consequence of acculturation and social ascendance, but depends to a large extent on the acceptability of the ethnic minority by the majority group (Alba et al. 2000). Under the spatial-assimilation model, socioeconomic and acculturation variables, such as household income, nativity/length of residence in the U.S., and linguistic isolation, should be important variables in explaining locational inequalities (Alba et al. 2000). Massey and Denton (1987) argue that even though social mobility and acculturation alone may result in assimilation, in the U.S. this process is directly related to suburbanization, since most racial and ethnic integration has occurred in the suburbs. However, the heavy flow of recent immigration may have altered the operation of spatial assimilation (Denton and Massey 1991;