Language Training, Language Proficiency and The Effect of Language Proficiency on Earnings of Migrants in Germany

Open Access
Vlazny, Sarah Catherine
Area of Honors:
Bachelor of Arts
Document Type:
Thesis Supervisors:
  • Amanda Marie Mandzik, Thesis Supervisor
  • Russell Paul Chuderewicz, Honors Advisor
  • economics
  • Mincer
  • earnings
  • migrants
  • immigration
  • language
  • language proficiency
  • Germany
  • income model
  • instrumental variable
  • ethnic identity
  • cultural assimilation
  • probit model
This paper uses a probit model to determine the impact of taking a language course on German language proficiency for migrants in Germany. It also models the income returns to language proficiency for fully-employed male migrants. The results show that those who have enrolled in a language course are significantly more likely to be proficient at writing in German. The results also show that certain ethnic identity factors, such as whether the migrant has visited Germans in their home and whether or not they have been discriminated against because of their ethnicity are significantly important in predicting language proficiency. The results imply that those who feel more assimilated in Germany and who have experienced less ethnic discrimination are more proficient in the German language. The model of income returns to language proficiency uses the language the migrants speaks with their friends and whether or not the migrant has visited a German citizen in their home as instrumental variables to correct for potential endogeneity issues. After this correction, the results show that language proficiency does have significantly positive effect on earnings. However, after adding additional employment variables such as the size of the company and the training required for the job, language proficiency is no longer significant. A likely explanation is that there is simultaneous causality between being proficient at the German language and having a high level of training or working at a large company. Another potential explanation is the absence of a variable to control for whether the migrant works in an urban or a rural area, which could be highly correlated with company size, training required, and earnings.