Earnings of students who change universities

Abstract

Using data on Swedish university entrants, this study finds that earnings are significantly lower for students who change universities compared to students who do not change. Earnings differences decrease over time and over the earnings distribution. The pattern in the estimates seems consistent with non-transfer students having higher earnings because of their earlier labor market entry and transfer students catching up because of their additional human-capital investments. But by changing universities, individuals signal that they are more likely to jump between jobs, and some employers account for these factors when screening job applicants.

Keywords:

• college education
• university choice
• earnings distribution

JEL classifications:

• J24
• J31

Acknowledgements

We thank Lena Granqvist, Niklas Hanes, Eva Löfbom, Maria Noleryd, Olle Westerlund, Gunnar Wetterberg, Christina Wikström, Magnus Wikström, and seminar participants at Umeâ University, Saco, and the 2008 European Association of Labor Economists (EALE) meeting for helpful comments. We also thank two anonymous referees for constructive comments. We are especially grateful to Galina Pokarzhevskaya for skillful computer programming and constructive comments.

Notes

Kane and Rouse (1995) and Hilmer (2000) are two related studies. Kane and Rouse consider a particular type of transfer students (non-degree recipients who attended both two- and four-year colleges). Hilmer considers various types of transfer students and estimates inter-group variation in terms of college quality rather than the wage effects of college transfers.

The system was introduced in 2007 to align Sweden's system with other European countries. Before 2007, one credit point was equal to one week of full-time studies. Credits in our data correspond to the old system.

The estimated earnings differences are significantly larger than those reported in this study, when also students who have changed more than once are included in the analyses.

See, Holmlund (2009) for a discussion of these quality measures and their effects on earnings.