Using German administrative data, I identify the causal effect of conscription on men's early career wage. My quasi-experimental design exploits a discontinuity in the probability of having served induced by the school start cutoff in connection with the suspension of conscription in 2011. Comparing men born in a narrow window around the cutoff, I find that conscription has a positive effect on wage for men of high educational background. Falsification tests using men and women of different birth years show that the effect is unique to men for whom the school start cutoff induces a discontinuity in the probability of having served. I test several mechanisms: draft avoidance behavior and the selection into broad occupational sectors do not explain my result. Rather, conscription causes men to select into jobs requiring more teamwork, which are on average higher-paying. In connection with qualitative survey evidence of the literature, my results suggest that men under conscription acquire social skills.

This paper provides evidence of costly perseverance in the field. In a setting where consultants select and pursue projects autonomously, I show that perseverance is related with fewer successfully completed projects as well as lower sales and commissions. Using rich firm data on individual job activity, I shed light on the task-specific behavioral mechanisms: In fast markets, the main channel of costly perseverance is a lower number of projects started; in slower markets, costs arise primarily from pursuing projects in a more narrow and isolated way. The survey questions driving costly perseverance point to the consultants' failure to incorporate negative signals and opportunity costs into their effort allocation. Using heterogeneity within and between consultants' task assignment, I show that perseverance is more costly in explorative tasks, as opposed to well-defined tasks characterized by mere exploitation.

We identify the impact of agency on women's education and family formation outcomes. The reform we investigate lowered the age of legal adulthood from 21 to 18 in 1975 West Germany. Before the reform, students of the academic school track became legal adults only 1-2 years after obtaining the general university entrance qualification (Abitur). Thus, they needed parental consent to make important decisions such as enrolling in university after school. After the reform, students became legal adults before obtaining university entrance qualification, and thus had full agency to shape the transition from school to higher education. Focusing on women of the academic school track, we define treatment assignment as being a legal adult at the predicted date of finishing school, which we calculate on the basis of birth month and federal state. Our difference-in-differences model exploits variation in school start cutoff dates that vary between federal states. Using data from the German Microcensus, we show that women assigned to treatment are more likely to obtain a university degree and less likely to marry and have children. Reaching the age of majority earlier thus has lasting effects on major life events and the population composition as a whole.