The Allocation of Talent under Perfect and Imperfect Information (Job Market Paper) w. Asbjørn Juul Petersen
We develop a sorting model that incorporates both self-selection by applicants, who assess their own skills, and evaluation by admissions agents, who rely on noisy signals of skills. Building upon the theoretical framework of the Roy model, our analysis examines how the presence of noise in skill signals influences admission strategies and the wage distribution. Our analysis reveals two novel behavioral effects on admission procedures. First, institutions may optimally assign positive weight to skill signals that are uncorrelated and unproductive in the relevant sector—a phenomenon we term talent hoarding. This talent hoarding effect disrupts comparative advantages, reallocates talent towards restricted sectors, and diminishes overall efficiency. Furthermore, as the noise in signals intensifies, the admissions agent is incentivized to increase the number of admitted applicants, thereby reducing reliance on admission rules and promoting more informative self-selection—a behavioral effect we label talent separation. Under relatively lenient assumptions, talent separation improves efficiency. Evidence from Danish administrative data reveals empirical patterns consistent with the predicted talent-hoarding effect, and a structural model of Denmark's education system highlights that the two behavioral effects can have a substantial impact on the distribution of wages.
How vocational training matters for social mobility and why children of immigrants are losing out w. Mathias Fjællegaard Jensen
We show that vocational training is a key driver of social mobility for sons of locals, but less so for sons of immigrants in Denmark. To understand why, we examine whether access barriers can explain sons of immigrants' underrepresentation in traditional, high-return vocational fields. We develop a stylized theoretical model in which students choose between two vocational tracks: old traditional and new digital tracks, facing group-specific search frictions. When traditional sectors are harder to access for immigrant youth, our model predicts earlier switching into new digital tracks and potential welfare losses from misallocation. Using population-wide administrative data, we show that sons of immigrants are 30\% more likely to graduate from ICT/digital vocational tracks. This difference appears as a skill-neutral parallel shift, suggesting that the inequality is not driven by underlying ability gaps but by access barriers (e.g., discrimination or lack of parental networks) in traditional vocational education. Digital tracks exhibit lower mean earnings but a significantly steeper skill gradient, consistent with digitalisation acting as a compensatory but incomplete substitute for access to traditional vocational sectors. Our findings suggest that access barriers in vocational training may result in suboptimal sorting across educational tracks, limiting the economic mobility of sons of immigrants.
Does High School Grade Variability Explains Long Run Labour Market Income? w. Asbjørn Juul Petersen
We study how variability in high school performance predicts long-run labour market income. Using Danish administrative data on complete grade distributions for all general high-school graduates from 2001 to 2006, we measure within-student grade variability as the standard deviation of final grades and relate it to long-run labour market income. Conditional on identical GPA and a long list of background controls, a one–standard deviation increase in grade variability predicts a 4.8 percent lower income. The result holds for various specifications and even in a twin fixed effects analysis. Decomposing the effect, we find that the negative relationship between grade variability and income is primarily driven by higher unemployment and fewer hours worked. The findings highlight performance variability as an overlooked dimension of human capital with substantial predictive power for long-run economic success.
Can Nominating Minority Candidates Backfire? w. Karl-Emil Svenningsen Bendtsen
We develop a model of candidate nomination under open-list proportional representation to analyze when adding more candidates of a given type promotes or hinders their electoral success. Nominating additional candidates generates a trade-off: While more candidates can increase the chance of representation by not putting all eggs in one basket, it simultaneously spreads votes more thinly across those candidates. The model formalizes the conditions under which each force dominates. We refer to the setting as soft when the more–draws effect dominates and as hard when vote–spreading is stronger. We generally find that the relationship is determined by the intensity of competitiveness in the election. We test the vote-spreading predictions by leveraging institutional variation from large-scale municipal and county amalgamations in Denmark and Finland, which generate systematic local advantages for candidates. Across both contexts, we find that being local yields a strong electoral premium; however, this premium declines sharply as more locals are nominated. We find that dedicated type-specific votes cannot become a disadvantage for the individual belonging to a type, although they can be diluted. Our findings highlight a fundamental trade-off in nomination strategies and inclusion rules. Expanding candidate pools broadens representation, but excessive nominations may reduce collective success.