Skill Substitution, Expectations, and the Business Cycle

This paper studies how labor market conditions around high school graduation affect postsecondary skill investments. Using administrative data on more than six million German graduates from 1995-2018, and exploiting deviations from secular state-specific trends, I document procyclical college enrollment. Cyclical increases in unemployment reduce enrollment at traditional universities and shift graduates toward vocational colleges and apprenticeships. These effects translate into educational attainment. Using large-scale survey data, I identify changes in expected returns to different degrees as the main mechanism. During recessions, graduates expect lower returns to an academic degree, while expected returns to a vocational degree are stable.

💡 Research Summary

This paper investigates how labor‑market conditions at the moment of high‑school graduation shape post‑secondary skill investments in Germany. Using administrative records for more than six million graduates spanning 1995‑2018, the author constructs commuting‑zone panels of enrollment and graduation shares for three mutually exclusive pathways: traditional universities (general‑skill, academic degrees), universities of applied sciences/other colleges (mixed general‑specific skills), and apprenticeships (highly specific, firm‑based training). The empirical strategy exploits deviations of state‑level unemployment rates from long‑run quadratic trends, thereby treating these deviations as exogenous shocks. By separating the national component of unemployment from the state‑specific component, the analysis distinguishes between macro‑level expectation effects and local‑level outside‑option effects.

The main findings are strikingly pro‑cyclical. A one‑percentage‑point increase in the state unemployment rate reduces first‑time college enrollment by 2.10 percentage points. The decline is concentrated in traditional universities, where attainment falls by 1.04 percentage points. Conversely, enrollment in vocationally oriented colleges rises by 0.40 percentage points, and apprenticeships become a more attractive alternative. In other words, adverse labor‑market conditions trigger a substitution from general‑skill academic education toward specific‑skill vocational pathways.

To uncover the mechanism, the author merges large‑scale survey data on expected earnings from different degrees. National unemployment spikes lower the expected return to an academic degree, while the expected return to a vocational degree remains essentially unchanged. The effect is strongest for low‑performing male graduates without college‑educated parents, whose perceived value of a vocational degree even rises during recessions. State‑level unemployment, by contrast, mainly influences the availability of apprenticeships and other outside options, with negligible impact on expectations.

Credit constraints and capacity limits are examined and found to play a minor role. Only a small share of students receive federal aid (BAföG), and tuition fees are low; variations in these factors do not explain the observed enrollment shifts. The author therefore concludes that expectation adjustments are the primary driver.

A theoretical framework extends a Roy‑type sorting model to include ability, parental income, tuition and psychological costs, and differentiates between expected college premiums and non‑college earnings. The model predicts that a shock to unemployment reduces the expected premium of a general‑skill degree more than that of a mixed or specific‑skill pathway, shifting the ability threshold at which students opt for each option. This prediction matches the empirical patterns.

The paper contributes to three strands of literature. First, it challenges the conventional “counter‑cyclical college‑going” narrative derived mainly from Anglo‑Saxon contexts, showing that in Germany college enrollment is pro‑cyclical and that students substitute toward apprenticeships rather than “safe‑harbor” community colleges. Second, it disentangles macro‑ versus local labor‑market shocks, demonstrating that national cycles affect expectations while local cycles affect outside options. Third, it adds to the growing body of work on belief formation and the business cycle by providing direct evidence that macro‑economic conditions reshape expected returns to education, with divergent effects across degree types.

Robustness checks include accounting for the Bologna Process reforms, controlling for regional industry composition, using alternative specifications of the unemployment trend, and testing alternative outcome variables (major choice, GPA). Results remain stable. The author discusses external validity, noting that the German system’s mix of publicly funded higher education and regulated apprenticeship markets resembles many OECD countries, suggesting broader relevance.

Policy implications are clear: during downturns, simply increasing financial aid for university study may be insufficient; instead, strengthening the quality and signaling of vocational pathways and managing expectations about academic returns could better align skill supply with labor‑market needs. Targeted information campaigns for low‑performing youths could further mitigate adverse skill substitution effects.