Abstract
We explore matched employer–employee data for a new destination of international migrants in Europe—Portugal. We conclude that the difference between the earnings of immigrants and natives with similar personal characteristics is for the most part due to the characteristics of the matches they form, immigrants being penalized on two different counts: absence of match-specific human capital and occupational downgrading. Moreover, we show that non-random sorting across workplaces has a significant detrimental effect on immigrants’ wages. This is the flip side of joining migrant-crowded workplaces.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Matched employer–employee data are not without problems, the most severe being the absence of information on the workers’ family status, and the exact date of arrival to the host country.
We note from the outset that our analysis focuses on formal employment. There is very limited evidence on the incidence of informal work arrangements amongst immigrants in Portugal. Carvalho (2007) reports survey results indicating that such arrangements are more prevalent in agriculture, construction, and retail and wholesale trade. He also finds that non-regular forms of employment account for 30% of total immigrants’ employment, which is above the estimated size of the shadow economy in Portugal (around 23% of the GDP according to Schneider et al. 2010). No data are available on workers’ income in the informal sector, immigrants’ or natives’.
Public administration and non-market services are excluded.
The nationality of the worker is the only information available that can be used to identify migrant workers. For that reason, in this article we take the word immigrant as synonymous with non-national citizen. Data permitting, our preferred option would have been different, as we must recognize that we are considering as immigrants all non-nationals independently of where they were born. We acknowledge that some of the workers that we are counting as immigrants may have already been born in Portugal. However, given the fact that large inflows of migrants are new to Portugal, we believe that this data limitation is likely to have a minor impact and does not reduce the validity of the results. Besides, the robustness checks we perform indicate that this is not a real problem in this case.
In most model specifications we use one variable—‘concentration of immigrants’—lagged one year. For that reason, all results reported refer to the period 2003–2008.
Whenever a worker was present in one wave of the QP data set more than once we kept only the record corresponding to the establishment where he or she was working the greater number of hours.
See the Serviços de Estrangeiros e Fronteiras (SEF) 2009 annual report.
Each worker is administratively assigned to one out of eight possible skill levels: top executives, intermediary executives, supervisors, highly skilled professionals, skilled professionals, semi-skilled professionals, unskilled professionals, and apprentices. Assignment is determined on the basis of the worker’s qualification and the nature of the tasks being performed (Decreto-Lei 121/78).
A special reference must be made to Chinese workers, as 43.6% of the men in this group and 44.8% of the women are paid at the legal minimum wage.
Amongst those using the same approach while focusing on other performance indicators (test scores), see Kahn (2004).
The set of regressors corresponding to specification 1 includes controls for the worker’s age (and age squared), immigrant status, a proxy for time since arrival in the country (and the same proxy squared), schooling achievement, and time and region effects. Specification 2 further controls for three employer characteristics: size, industry, and immigrant concentration at the workplace. Finally, specification 3 adds controls for worker tenure (and tenure squared), and skill-categories. For a definition of all variables see Appendix B.
Selectivity into labour force participation is a well-known problem in immigration studies, especially in the case of women. Given the nature of our data, there is nothing we can do to control for this type of selection other than estimating the models by worker fixed effects. Fixed effects estimation is not an option in this case as it would eliminate the parameter of interest because the corresponding variable—immigrant status—does not change over time. Hence, we simply acknowledge the problem and refrain from making causal interpretations of the estimates obtained with the women subsample.
Our results are also consistent with those that would be obtained by computing the Oaxaca–Blinder decomposition of the raw wage differential between migrants and natives. Although the focus of this article is not on measuring wage discrimination against migrants, we note that if we account for all the observable characteristics of workers, employers, and matches included as covariates in specification 3, the Oaxaca–Blinder procedure would attribute 76% of the wage differential between male immigrants and natives (85% in the case of women) to differences in characteristics. Gelbach’s procedure, which produced the results in Table 2, actually nests the Oaxaca–Blinder decomposition.
The full set of results with the restricted samples are available from the authors upon request.
Throughout, we refer to the immigrants’ wage penalty as the absolute value of the coefficient of the immigrant status dummy variable in the wage equation.
Immigrant concentration at the level of the establishment is measured as the log of the share of workers in the establishment that are non-national citizens. To minimize potential endogeneity problems we use the first lag of this variable measured at the establishment level, meaning that for worker i in establishment j in year t the immigrant concentration variable measures the proportion of immigrant workers in establishment j in year t − 1 even if worker i was working at a different establishment in year t − 1 or he or she had not entered the country at that time.
Full results are presented in Appendix D.
Nepotism, i.e., a positive preference for co-workers belonging to minority groups, could arguably have produced the same effect, but the estimated coefficients are far too large to believe that this is the case. In any case, nepotistic behaviour is less common among low-wage workers. This type of behaviour is more frequently referred to (although not in the context of wage studies) amongst highly skilled white-collar workers in intellectual or scientific occupations. The results of quantile regression estimation (Fig. 3) indicate otherwise—the magnitude of the effect of workplace concentration for the pooled sample of natives and immigrants increases in absolute value as we move to the right of the wage distribution.
The establishment fixed effects results presented in Appendix C corroborate this idea. Actually, when an establishment fixed effect is added to specification 3, the immigrant wage penalty is reduced by around 0.04 log points for men and 0.02 log points for women.
For women, the average share of immigrants in the workplace is 30.1%, 29.0% for the Africa and East Timor group, and 28.8% for the Brazilian group.
We report only the results for males.
For Chinese immigrants the estimate is not statistically significant at the 1% level up to the fifth decile, even though this can be due to the fact that this is the group with the smallest usable sample (5,845 observations). The size of each nationality subsample is the same as the total number of observations reported in Table A.1 (Appendix A).
References
Adsera A, Chiswick BR (2007) Are there gender and country of origin differences in immigrant labor market outcomes across european destinations? J Popul Econ 20(3):495–526
Aydemir A, Skuterud M (2008) The immigrant wage differential within and across establishments. Ind Labor Relat Rev 61(3):334–352
Borjas GJ (1999) Heaven’s door: immigration policy and the American economy. Princeton University Press, Princeton
Borjas GJ (2000) Ethnic enclaves and assimilation. Swed Econ Policy Rev 7(2):89–122
Borjas GJ (1994) The economics of immigration. J Econ Lit 32(4):1667–1717
Calvó-Armengol A, Jackson MO (2004) The effects of social networks on employment and inequality. Am Econ Rev 94(3):426–454
Carrington WJ, Lima PJF (1996) The impact of 1970s repatriates from Africa on the Portuguese labor market. Ind Labor Relat Rev 49(2):330–347
Carvalho LX (2007) Os Limites da Formalidade e o Trabalho Imigrante em Portugal. ACIDI, Lisbon
Chiswick BR (1979) The economic progress of immigrants: some apparently universal patterns. In: Fellner W (ed) Contemporary economic problems. American Enterprise Institute, Washington DC, pp 357–399
Chiswick BR, Miller PW (2002) Immigrant earnings: language skills, linguistic concentrations and the business cycle. J Popul Econ 15(1):31–57
Cohen S, Eckstein Z (2002) Labor mobility of immigrants: training, experience, language and opportunities. Int Econ Rev 49(3):837–872
Dustmann C, Glitz A, Schönberg U (2011) Referral-based job search networks. IZA discussion paper no. 5777
Fontaine F (2008) Why are similar workers paid differently? the role of social networks. J Econ Dyn Control 32(12):3960–3977
Eckstein Z, Weiss Y (2004) On the wage growth of immigrants: Israel, 1990–2000. J Eur Econ Assoc 2(4):665–695
Friedberg RM (2000) You can’t take it with you? Immigrant assimilation and the portability of human capital. J Labor Econ 18(2):221–251
Gelbach JB (2009) When do covariates matter? And which ones, and how much? University of Arizona WP, pp 09–07
Green DA (1999) Immigrant occupational attainment and mobility over time. J Labor Econ 17(1):49–79
Jasso G (2008) A new unified theory of sociobehavioural forces. Eur Sociol Rev 24(4):411–434
Kahn LM (2004) Immigration, skills and the labor market: international evidence. J Popul Econ 17(3):501–534
Pereira SC (2003) The impact of minimum wage on youth employment in Portugal. Eur Econ Rev 47(2):229–244
Portugal P, Cardoso AR (2006) Disentangling the minimum wage puzzle: an analysis of worker accessions and separations from a longitudinal matched employer-employed data set. J Eur Econ Assoc 4(5):988–1013
Schneider F, Buehn A, Montenegro CE (2010) New estimates for the shadow economies all over the world. Int Econ J 24(4):443–461
SOPEMI (2002) Trends in international migration. OECD Publications Service, Paris
Venturini A (2004) Postwar migration in Southern Europe, 1950–2000. An economic analysis. Cambridge University Press, Cambridge
Weiss Y, Sauer RM, Gotlibovski M (2003) Immigration, search and loss of skill. J Labor Econ 21(3):221–251
Acknowledgements
CEF.UP—Center for Economics and Finance at University of Porto—is supported by the Fundação para a Ciência e Tecnologia (FCT), Portugal. We would like to thank Alessandra Venturini, Barry Chiswick, Carmel Chiswick, Catalina Amuedo-Dorantes, Guillermina Jasso, Paulo Guimarães, and three anonymous referees for comments on earlier versions of the paper. Financial support from Fundação para a Ciência e a Tecnologia (Research grant no. PIQS/ECO/50044/2003) is deeply appreciated. We also thank the Gabinete de Estratégia e Planeamento do Ministério do Trabalho e da Solidariedade Social, which kindly allowed us to use the data.
Author information
Authors and Affiliations
Corresponding author
Additional information
Responsible editor: Klaus F. Zimmermann
Electronic Supplementary Material
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Carneiro, A., Fortuna, N. & Varejão, J. Immigrants at new destinations: how they fare and why. J Popul Econ 25, 1165–1185 (2012). https://doi.org/10.1007/s00148-011-0387-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00148-011-0387-3