Latino Spatial and Structural Assimilation: Close Intergroup Friendships among Houston-Area Latinos

Abstract

This article examines the cross-sectional association between spatial and structural assimilation among Latino adults in Harris County, Texas, USA. Prior studies have produced little evidence of any significant association between the racial/ethnic composition of Latinos' neighbourhoods and their social networks, particularly net of controls for socio-economic status (SES). But prior studies have failed to distinguish social ties that are either weak or exceptional from strong, systematic ties that may be more indicative of social integration, ignored the racial/ethnic diversity in Latinos' neighbourhoods and networks, or both. Negative binomial and multinomial logit analyses predicting close intergroup friendships among Latinos were conducted using data from the 2006 Houston Area Survey. Net of acculturation and SES, Latinos who lived in neighbourhoods that provided greater exposure to non-Hispanic whites were more likely to have non-Hispanic whites among their three closest friends than those who lived in Latino enclaves. Neighbourhood diversity, particularly among non-white residents, was also associated with diverse friendship networks among Latinos. Implications for the relationship between Latino spatial and structural assimilation are discussed.

Keywords:

• Latino Americans
• Residential Integration
• Structural Assimilation
• Spatial Assimilation
• Interracial Friendship

Notes

[1] Technically, Hispanic is a broader term than Latino, since the former includes all Spanish-speaking peoples (including immigrants from Spain), whereas Latino refers specifically to people of Latin American ancestry. In practice, this distinction is unimportant, since less than one percent of Harris County Hispanics were of Spanish ancestry.

[2] There were no significant differences across waves in the proportions of US-born Latinos who reported having any Asian or black friends. Consequently, the proportion reporting any intergroup friendships was only slightly lower among US-born Latinos in 2006 as compared to 2003, and this difference was not statistically significant (z = 1.28, p = .20).

[3] I assumed that the 2005–2009 ACS estimates provided the best estimate of the midpoint of this period (i.e., 2007) and then used linear interpolation with the percentage from the 2000 Census as the starting value and the ACS estimate as the 2007 ending value.

[4] I also estimated models that controlled for whether respondents were second generation (i.e., born in the United States to at least one immigrant parent) or third generation or beyond (i.e., born in the United States to non-immigrant parents). These measures were strongly associated with having non-Latino friends in bivariate analyses, but did not significantly predict friendship network composition in multivariate models that also controlled for English proficiency, so I omitted them from the analyses presented below.