Labour Market Outcomes for Immigrants and the New Zealand-born 1997-2009
Conclusions
In this paper, unit record survey data from the 1997-2009 NewZealand Income Survey (NZIS) is used to examine how labour market outcomes and returns to human capital vary in the immigrant and non-immigrant populations in NewZealand and whether these returns vary over time and across business cycles. First, regression analysis is used to estimate the relationship between education and employment, hours worked and wages for immigrants and the NewZealand-born separately for men and women, controlling for other factors that are related to individual wages. These relationships are allowed to vary over time to see whether any variation is systematically related to business cycles or other macroeconomic variation.
Then, the same empirical approach is used to examine the relationship between education and labour market outcomes for immigrants and NewZealand-born with different ethnic backgrounds and from different source countries, again separately for men and women. This part of the analysis extends previous work to simultaneously consider the relationship between immigrant status and ethnicity, and labour market outcomes. This may help to illustrate the extent to which different outcomes for immigrants are likely related to their immigrant status per se, rather than other pertinent characteristics, such as their ethnicity.
The four main findings from the first part of the analysis are as follows.
First, the relationship between education and both employment and hours worked conditional on employment is the same for immigrants and NewZealand-born of the same gender. On the other hand, immigrants earn relative more in terms of hourly wages than the NewZealand-born for each year of education.
Second, differences in characteristics do not explain the observed lower employment rates for both male and female migrants and the shorter average work hours of male migrants and longer average work hours of female migrants. Accounting for differences in characteristics indicates that the average wage for immigrants is 4-8percent lower than the average wage for an equivalent NewZealand-born individual.
Third, while differences in employment rates occur across the educational distribution, the observed differences in hours worked and the estimated immigrant wage gap occur mostly among less-educated workers. In fact, immigrants with university education and above generally have quite similar hours worked and wages as the NewZealand-born with the same level of education, although they have lower employment rates.
Fourth, there is a surprising lack of variation in differences in employment rates, hours worked, and real wages for NewZealand-born and immigrants between 1997 and 2009. This suggests that, at least over this period, the business cycle did not have differential impacts on labour market outcomes for migrants and the NewZealand-born.
Further examining how labour market outcomes vary for immigrants and NewZealand-born with different ethnic backgrounds and from different source countries reveals the following findings.
First, differences in characteristics explain most of the large employment gap, particularly for men, between NewZealand-born Māori, NewZealand-born Pasifika, and NewZealand-born Pākehā/Europeans. On the other hand, almost none of the large observed gap in employment rates for Foreign-born Pacific, Asian-born Asians, Pacific-born Asians, and Foreign-born Other relative to NewZealand-born Pākehā/Europeans is explained by differences in characteristics.
Second, unlike for employment rates, little of the wage gap for NewZealand-born Māori and NewZealand-born Pasifika relative to NewZealand-born Pākehā/Europeans is explained by differences in characteristics. Similarly, once controlling for the different average characteristics of each group, large wage gaps are found for NewZealand-born Asians, Foreign-born Pasifika, Asian-born Asians, Pacific-born Asians, and Other-born Asians.
Overall, these results indicate that when looking at employment gaps, it is immigrant status as opposed to ethnicity that is driving poor outcomes for Asians and Pacific Islanders in NewZealand. Interpreted along with the evidence from the first analysis that immigrants at all levels of education have lower employment rates than equivalent NewZealand-born, these results are consistent with Asian and Pasifika immigrants having worse job networks or higher reservation wages, perhaps because of different family obligations or less access to informal childcare, or being affected by labour market discrimination.
It is also possible that lower employment rates occur because these immigrants have lower effective human capital than equally educated NewZealand-born, perhaps because of worse English language skills, but in this case one would expect to find bigger employment differences among less-educated migrants since this is presumably the group for whom poor English language skills are most common.
On the other hand, Māori, Pasifika and Asians are found to have much lower wages than individuals with European or Other ethnicity regardless to whether they are immigrants. Interpreted along with the evidence from the first analysis that it is only less educated migrants that have much lower wages than equivalent less educated NewZealand-born, these results suggests that labour market discrimination may play an important role in wage setting among less educated Māori, Pasifika and Asians in NewZealand. However, other possible explanations, such as lower quality education, worse job networks, or occupational segregation could also explain these findings.
Unfortunately, the NZIS is not well suited for distinguishing in an empirical sense the relative importance of each of these possible explanations for worse labour market outcomes among individuals in different ethnic/immigrant groups in NewZealand. Hence, further work with different data needs to be done to differentiate between these possible reasons. For example, detailed data on individual job search behaviour could be used to examine whether equivalent NewZealand-born Pākehā/Europeans, members of minority ethnic groups, and immigrants have similar job-finding rates. Similarly, longitudinal employee-employer data could be used to see whether promotion and retention rates differ by ethnicity and immigrant status for otherwise similar employees or whether certain groups of workers appear to be paid less than what they contribute to firm productivity (for example, Hellerstein etal 2002).
List of Figures
Figure 1: Relationship Between Years of Education and Employment for NZ-Born and Immigrants by Gender and Year
Figure 2: Relationship Between Years of Education and Hours of Work for NZ-Born and Immigrants by Gender and Year
Figure 3: Relationship Between Years of Education and Wages for NZ-Born and Immigrants by Gender and Year
Figure 4: Actual and Predicted Employment Rate Gaps for NZ-Born and Immigrants by Gender and Years of Education
Figure 5: Actual and Predicted Hours Work Gaps for NZ-Born and Immigrants by Gender and Years of Education
Figure 6: Actual and Predicted Wage Gaps for NZ-Born and Immigrants by Gender and Years of Education
Figure 7: Relationship Between Years of Education and Employment for NZ-Born and Immigrants by Ethnicity and Gender
Figure 8: Relationship Between Years of Education and Hours of Work for NZ-Born and Immigrants by Ethnicity and Gender
Figure 9: Relationship Between Years of Education and Wages for NZ-Born and Immigrants by Ethnicity and Gender
List of Tables
Table 1: Observable Skills Among NZ-Born and Immigrants by Gender and Year
Table 2: Labour Market Outcomes for NZ-Born and Immigrants by Gender and Year
Table 4: Predicted Labour Market Outcomes for NZ-Born and Immigrants by Gender and Year
Table 5: Observable Skills by Among NZ-Born and Immigrants by Ethnicity and Gender
Table 6: Labour Market Outcomes for NZ-Born and Immigrants by Ethnicity and Gender
Table 8: Predicted Labour Market Outcomes for NZ-Born and Immigrants by Ethnicity and Gender
