Ethnic Diversity, Social Identity, and Social Withdrawal: Investigating Putnam’s Constrict Thesis

ABSTRACT

Since Putnam introduced his constrict thesis in 2007, many researchers have established that ethnic diversity lowers perceptions of social cohesion, at least in the short term. The connection between ethnic diversity and social behavior, however, is less certain. In this paper we draw on social distance and social identity theories to empirically test if ethnic diversity encourages behaviors linked to social withdrawal. Using data from a longitudinal panel study of urban communities in Australia, we examine the influence of social distance on neighborhood ties, neighborly exchange, and civic engagement and assess if an individual’s social identity (ethnic or civic) strengthens or weakens these relationships. We find individuals that endorse an ethnic identity are more likely to engage in social withdrawal behaviors. Withdrawal is also more likely in neighborhoods where individuals distort the presence of minorities.

KEYWORDS:

• Social distance
• social identity
• ethnic diversity
• immigration

Disclosure Statement

No potential conflict of interest was reported by the authors.

Notes

1. Putnam’s paper “E Pluribus Unum: Diversity and Community in the Twenty-first Century” has over 5000 citations according to Google Scholar.

2. For brevity, “ethnic diversity” hereafter refers to ethnic, racial, and cultural diversity. We recognize that racial diversity is a concept used in the U.S., referring to visible differences in skin color. In non-U.S. contexts where diversity is driven by immigration and ethnic differences (in particular language and religion), ethnic diversity is more the commonly used term. Given our research was conducted in Australia, an ethnically diverse, immigrant nation, we use ethnic diversity throughout this manuscript.

3. In this study, neighborhoods represent state suburbs. In Australia, state suburbs are similar to census tracts in the U.S. context, though in some cases, suburbs may be larger than census tracts as they are not determined by population. Throughout this paper we use the more familiar term “neighborhood”.

4. The survey did not directly ask the race of the respondent. Instead, we classified as “white” respondents who reported their ancestry as being Australian, Northern European, or North America. We note that this is an imprecise measure as people can be nonwhite within these categories, but it is the only way we can distinguish those from Anglo-Saxon backgrounds from other backgrounds as the Australian Census does not collect information on “race”. In our sample, 78% were classified as white. Thus, our neighborhoods were 22% nonwhite, with a standard deviation of 14.3%, ranging from 0% to 65% nonwhite.

5. These models showed a relatively good fit to the data. We included as predictors the expected number of ties, the quadratic version of this measure, and the cubic version (to capture possible nonlinearities). They explained about 50% of the variation in the actual number of intergroup ties in Melbourne neighborhoods.

6. We also estimated models with each of the interaction terms entered separately, and found similar results to those including all interactions simultaneously. We therefore present the model results including all interactions simultaneously.

Additional information

Funding

This work was supported by by the Australian Research Council [LP0453763; DP0771785; RO700002; DP1093960; DP1094589 and DE130100958].

Notes on contributors

Rebecca Wickes

Rebecca Wickes is an Associate Professor in Criminology and the Director for the Monash Migration and Inclusion Centre  at the School of Social Sciences, Monash University. She is the lead investigator of the Australian Community Capacity Study.  Dr Wickes’ research focuses on physical and demographic changes in urban communities and their influence on social relationships, inclusion and the concentration of social problems, in particular crime, social exclusion and public disorder.

John Hipp

John Hipp is a Professor in the departments of Criminology, Law and Society, and Sociology, at the University of California Irvine.  His research interests focus on how neighborhoods change over time, how that change both affects and is affected by neighborhood crime, and the role networks and institutions play in that change.  He approaches these questions using quantitative methods as well as social network analysis.

Jacqueline Laughland-Booÿ

Jacqueline Laughland-Booÿ is Senior Advisor (Research and Communications) in the Office of the Vice-Chancellor and President at Australian Catholic University, Australia. She is also an Adjunct Research Fellow in the School of Social Sciences at Monash University, Australia.  Her research interests are in the fields of life course studies and political sociology.