Essay I: Are workers with poor outside opportunities less responsive and more susceptible to negative demand shifts in routine occupations? To answer this, I create and estimate an occupation specialization index (OSI) using Swedish register data and machine learning tools. It measures the expected difference in utility between a worker's occupation and his best non-routine outside option. This determines the loss he is willing to tolerate to avoid switching. Low-OSI employees disproportionately left routine work. Their future wage growth was akin to comparable workers initially in non-routine occupations. By contrast, routine specialists largely stayed put and experienced lower wage growth than generalists and non-routine specialists.
Essay II (with Adrian Adermon, Georg Graetz, and Yaroslav Yakymovych): Using a new identification strategy, we jointly estimate the growth in occupational wage premia and time-varying occupation-specific life-cycle profiles for Swedish workers 1996–2013. We document a substantial increase in between-occupation wage inequality due to differential growth in premia. The association of wage premium growth and employment growth is positive, suggesting that premium growth is predominantly driven by demand side factors.
Essay III (with Peter Fredriksson, Lena Hensvik, and Oskar Nordström Skans): We provide two pieces of evidence that workers' capacity to extract rents from match-specific productivity hinges on their outside options. Using a measure of match quality, derived from the relationship between workers' multidimensional skills and job-specific skill requirements, we show that: (i) wages within ongoing matches are more closely aligned with match quality following an improvement of local labor market conditions; (ii) wages of job-to-job movers are positively related to the match quality in the previous job, even when controlling for previous wage.
Essay IV (with Mats Hammarstedt, and Per Skedinger): In a field experiment we study the causal effects of previous work experience and language skills when Syrian refugees in Sweden apply for low-skilled jobs. We find no evidence of sizeable effects from experience or completed language classes on the probability of receiving callback from employers. However, female applicants were more likely than males to receive a positive response. As a complement to the experiment, we interview a select number of employers.