Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in University of Tübingen ANNE STEINBACHER ♯ University of Tübingen May 2012Abstract This paper analyzes the role of ethnic communities in shaping the recent immigration boom to Spain. We find that ethnic communities exerted a strong positive effect on the scale and a strong negative effect on the skill structure of this immigration. Unlike previous studies, we explicitly acknowledge similarities among final migration destinations and thus partly relax the independence of irrelevant alternatives assumption. We strengthen our causal interpretation by controlling for observed and unobserved heterogeneity in bilateral migration costs, and by adopting an instrumental variables approach. Our results suggest that previous estimates of the scale effect are upward-biased by approximately 50%.

We offer fresh evidence on the effect of migrant networks on two essential aspects of migration: (1) the total scale of migration and (2) the skill composition of migration. Our analysis is for the remarkable case of Spain, which experienced a full‐blown immigration boom from the mid‐1990s up to the Global Financial Crisis. To accommodate flexible substitution patterns across alternative migrant destinations, we use a three‐level nested multinomial logit model. We find a strong positive network effect on the scale of migration and a strong negative effect on the ratio of high‐skilled to low‐skilled migrants. Simplifying restrictions on the structure of cross‐destination substitutability are rejected by the data. (JEL F22, J61)