Spillovers from immigrant diversity in cities

Abstract

Theory and evidence suggest that people born in different countries complement each other in the labor market. Immigrant diversity could augment productivity by enabling the combination of different skills, ideas and perspectives, resulting in greater productivity. Using matched employer-employee data for the USA, this paper evaluates this claim, and makes empirical and conceptual contributions to prior work. It addresses the potential bias from unobserved heterogeneity among individuals, work establishments and cities. The paper also identifies diversity impacts at both city and workplace scales, and considers how relationships vary across different segments of the labor market. Findings suggest that urban immigrant diversity produces positive and nontrivial spillovers for U.S. workers. This social return represents a distinct channel through which immigration may generate broad-based economic benefits.

Urban Immigrant Diversity and Inclusive Institutions

“…Though results vary by country and approach taken, a wealth of studies find evidence consistent with the idea that immigrant diversity augments worker and firm productivity (Ottaviano and Peri 2006;Nathan 2011Nathan , 2015Kemeny 2012;Bakens, Mulder, and Nijkamp 2013;Bellini et al 2013;Longhi 2013;Lee 2014;Suedekum, Wolf, and Blien 2014;Alesina, Harnoss, and Rapoport 2016). Recent contributions have sought to address bias from confounding factors, including nonrandom worker selectivity or sorting (Bakens, Mulder, and Nijkamp 2013;Kemeny and Cooke 2015;Trax, Brunow, and Suedekum 2015;Elias and Paradies 2016). In the case of Trax, Brunow, and Suedekum (2015) for Germany, and Kemeny and Cooke (2015) for the United States, the positive association between urban immigrant diversity and productivity remains after accounting for sorting behavior, diversity in the workplace, and other sources of heterogeneity at the individual, workplace, and city scales.…”

“…With few exceptions, existing empirical studies at the urban scale have not measured such institutions, instead testing more generally for the existence and direction of spillovers from diversity. These studies mostly conclude that diversity and productivity are robustly positively related, leading researchers to suggest that immigrant diversity in cities generates tangible economic benefits (Ottaviano and Peri 2006;Nathan 2011;Bellini et al 2013;Suedekum, Wolf, and Blien 2014;Kemeny and Cooke 2015;Trax, Brunow, and Suedekum 2015;Elias and Paradies 2016). Hence, one reason to examine the moderating role of institutions is that it provides the opportunity to estimate a model of diversity's economic impacts that is, compared to much of the extant empirical work, closer to theory.…”

“…First, it offers new and improved measures of local institutions, triangulating across these in order to have more confidence that they capture the underlying latent variable of interest. Second, in keeping with recent contributions in the literature, notably Bakens, Mulder, and Nijkamp (2013), Trax, Brunow, and Suedekum (2015), Kemeny and Cooke (2015), and Elias and Paradies (2016), it uses an empirical approach that accounts for a wide range of potential confounding factors. Third, it unpacks overall estimates to consider how institutions and diversity operate differently for natives and foreign-born workers.…”

“…Recent contributions have sought to address bias from confounding factors, including nonrandom worker selectivity or sorting (Bakens, Mulder, and Nijkamp 2013;Kemeny and Cooke 2015;Trax, Brunow, and Suedekum 2015;Elias and Paradies 2016). In the case of Trax, Brunow, and Suedekum (2015) for Germany, and Kemeny and Cooke (2015) for the United States, the positive association between urban immigrant diversity and productivity remains after accounting for sorting behavior, diversity in the workplace, and other sources of heterogeneity at the individual, workplace, and city scales.…”

“…We build two distinctive indicators: a 8 Kemeny and Cooke (2015) demonstrate that estimates of the relationship between diversity and wages are not strongly dependent on choosing the standard fractionalization measure, as against an entropy index, decompositions of the fractionalization measure in the manner of Alesina, Harnoss, and Rapoport (2016), or measures that distinguish the simple share of foreign born from the range of countries present.…”