Return to skills and labour market size

ABSTRACT

Several studies document that skills are strong predictors of earnings; however, less is known about the extent to which labour market size influences the return to skills. Using data from a unique representative survey recording the skill requirements of Hungarian firms, we show that social skills have higher returns in large urban labour markets. Surprisingly, this pattern cannot be observed for cognitive skills, while the return to manual skills slightly decreases with labour market size. Our estimates are robust to different agglomeration measures, additional controls and estimation methods; however, returns to skills seem to vary considerably across worker groups.

KEYWORDS:

• skill
• agglomeration
• labour market size
• wage

JEL:

• J24
• J31
• R12
• R23

ACKNOWLEDGEMENTS

The authors are grateful for the insightful comments of three anonymous referees, Zoltán Elekes, Gergő Tóth, Balázs Lengyel, Rikard Eriksson, and the seminar participants at IE-CERS, ELTE and Umeå University for their insightful comments. We also thank the participants at the 2018 Geographies of Innovation (GEOINNO) conference in Barcelona for their comments on a preliminary version of the paper. All remaining errors are the authors’ alone.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. Other explanations include worker sorting and firm selection.

2. For firms, where the total number of occupations did not exceed 10, all available occupations were included in the sample.

3. We have experimented with different combinations of the skill items and tried principal component analysis to create alternative skill indices, but due to the high pairwise correlations between skill items, none of these alternative approaches changed any of the results reported below. Running the models separately for the selected skill items, we find that skill items that are assigned to the same skill type show similar wage returns (see Appendix C in the supplemental data online).

4. Appendix C in the supplemental data online estimates additional models in which other individual-level variables (such as education and work experience) are allowed to have coefficients that vary with LAU-1 population. The main results of Table 1 continue to hold.

5. To distinguish between routine and non-routine occupations, we use the classification scheme developed by Becker et al. (2013), which reports the fraction of non-routine job tasks for two-digit ISCO-88 occupations. Where the share of non-routine tasks exceeds 50%, they are considered non-routine tasks.