Do ethnic networks influence circular migration? Evidence from the UK

ABSTRACT

This article identifies ethnic network effects among temporary migrants in the UK. Using microdata from Understanding Society and the UK Census, the empirical results show that ethnic networks change the individual probability for circular migration. These effects are strong and significant only for some ethnicities, when controlling for a set of socio-economic characteristics and when adding spatial variability

KEYWORDS:

• Circular migration
• ethnic networks
• social connections
• United Kingdom

JEL CLASSIFICATION:

• F22
• R23

Acknowledgment

The author thanks Mark Taylor and all the ISER department of the University of Essex for their comments and the data provision. A special thank to David McKenzie, Michel Beine and Gautam Tripath.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ University of Essex (2012)

² Minority Ethnic Boost Sample; See Berthoud at al. (2009).

³ The information for the percentage of the ethnic minorities is derived from Census 2001.

⁴ The local authority districts are the finer level of spatial disaggregation of the UK Government structure.

⁵ Indian Network is obtained by interacting the size of each ethnic network with each individual ethnicity. For the same reason the mixed ethnicities are excluded.

⁶ I do not observe the wage for the home country and one feasible way is to approximate it by estimating the potential wages of the returns given by the equation:where is a vector of individual characteristics such as age, , gender, are the year of schooling of each individual returned, and are dummy variables for ethnic groups (For a reference see Dustman, 2003). Whereas, for the host wages, I use the reported UK wages.

⁷ In the meantime, characteristics of the country origin may influence the decision for circular migration such as the occurrence of the social conflicts or economic crisis, push factors that are difficult to be captured in the data set. In this case, I introduce a proxy variable for the level of development of the origin country, HDI (Human Development Index 2001), whereas the log of geographical distance is used as a proxy for the migration cost.

⁸ Government Office Regions