Working Paper: CEPR ID: DP15412
Authors: Samuel Dodini; Michael F. Lovenheim; Kjell G. Salvanes; Alexander Willn
Abstract: This paper extends the literature on monopsony and labor market concentration by taking a skillbased approach and estimates the causal effect of monopsony power on labor market outcomes. The prior literature has focused on industry and occupation concentration and likely overstates the degree of monopsony power, since worker skills are substitutable across different firms, occupations and industries. Exploiting linked employer-employee data that cover the universe of Norwegian workers over time, we find that our skill-based monopsony measure shows lower degrees of monopsony power than the conventional industry-and occupation-based measures. However, we also find that the gender gap in concentration is substantially larger using the skillbased measure relative to the occupation- or industry-based measures. Using mass layoffs and establishment closures as an exogenous shock to labor demand, we find that workers who experience a mass separation have substantially worse subsequent labor market outcomes when they are in more concentrated skill clusters. Our results point to the existence of employer market power in the economy that is driven by the concentration of skill demand across firms.
Keywords: monopsony; skills; labor market concentration
JEL Codes: J23; J24; J42; J63
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| monopsony power driven by skill demand concentration (J42) | lower wages for workers (J31) |
| higher skill concentration (J24) | worse subsequent labor market outcomes for workers (J79) |
| 0.1 point increase in HHI of skill concentration (D29) | decrease in earnings by 9120 krone (J31) |
| monopsony power (J42) | larger gender gap in concentration (J79) |
| monopsony power (J42) | larger decline in male wages compared to female wages (J79) |
| higher skill concentration (J24) | reduces likelihood of skill upgrading (J24) |
| higher skill concentration (J24) | increases likelihood of skill downgrading post-separation (J62) |
| workers in concentrated skill clusters (J24) | less likely to move to another labor market after separation (J69) |