Yaroslav Yakymovych
I am a postdoctoral researcher at the Institute for Housing and Urban Research (IBF), Uppsala University, Sweden. In my research, I apply machine learning methods as well as traditional econometric techniques to topics in labour and urban economics. I am particularly interested in social mobility, worker outcomes following job displacement, routine-biased technological change and sickness absence.
CV (updated 05/2024)
Research
Medical Certificates and Sickness Absence: Who Stays Away From Work if Monitoring Is Relaxed?
Sickness insurance guarantees employees the right to take leave from work when they are sick, but is vulnerable to excessive use because monitoring of recipients’ health is difficult and costly. In terms of costs, it would be preferable to focus monitoring on individuals whose sickness absence it strongly affects. This paper studies targeted monitoring in the setting of a large-scale randomised experiment where medical certificate requirements were relaxed for some workers. I employ a machine learning method, the generalised random forest, to identify heterogeneous effects on the duration of workers’ sickness absence spells. This allows me to compute treatment effect estimates based on an extensive set of worker characteristics and their potentially complex relationships with each other and with sickness absence duration. The individuals who are most sensitive to monitoring are characterised by a history of extensive sick leave uptake, low socioeconomic status, and male gender. The results suggest that a targeted policy can achieve the same reduction in monitoring costs as took place during the experiment at a 51 percent smaller loss in terms of increased sickness absence. Monitoring all insured individuals is estimated to be inefficient, but the benefits of targeted monitoring are estimated to exceed the costs.
IFAU working paper Report in Swedish Link to latest version
Media coverage: Dagens ETC, Han vill se minskade krav på läkarintyg vid sjukfrånvaro – för vissa
The heterogeneous earnings impact of job loss across workers, establishments, and markets (with Susan Athey, Lisa Simon, Oskar Nordström Skans and Johan Vikström)
Using generalized random forests and rich Swedish administrative data, we show that the earnings effects of job displacement due to establishment closures are highly heterogeneous. We find as much heterogeneity within as across closing establishments, and within as across worker types defined by age and schooling. We display the potential of market-based policy interventions by showing that much of the heterogeneity across establishments is shared within markets. Several results suggest that the effect heterogeneity disfavors already vulnerable workers. Thus, targeted policy interventions may be justified to a larger extent than suggested by estimated average earnings effects.
IFAU working paper Report in Swedish
Consequences of Job Loss for Routine Workers
Routine-biased technological change has deteriorated labour market prospects for workers in exposed occupations as their work has increasingly been done by machines. Routine workers who lose their jobs in mass displacement events are likely to have been a particularly affected group, due to difficulties in finding new employment that matches their skills and experience. I compare the annual earnings, employment, monthly wages, days of unemployment and disposable incomes of displaced routine workers to those of displaced non-routine workers using matched employer-employee data from Sweden. The results show substantially larger losses among displaced routine workers, with the difference persisting in the medium to long term. Over ten years, displaced routine workers’ additional losses amount to 1.2 years of pre-displacement earnings, and they spend 250 more days in unemployment over the same period. A significant share of the earnings losses is passed through to disposable income, with routine workers’ additional disposable income losses corresponding to 0.8 years of pre-displacement income. Loss of occupation-specific human capital might explain the more adverse effects, as routine workers find new jobs less similar to those they had before becoming displaced. I do not find evidence that switching to a non-routine occupation reduces routine workers’ losses, but on the contrary, there are indications that switchers do worse in the short-to-medium run. The findings suggest that the most exposed individuals can suffer severe consequences from labour-replacing technological change, and that their losses are not compensated by the social insurance system.
IFAU working paper Report in Swedish Link to latest version
Understanding occupational wage growth (with Adrian Adermon, Simon Ek and Georg Graetz)
Using a new identification strategy, we jointly estimate the growth in occupational wage premia as well as 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 as well as due to shifts in life-cycle profiles. However, this increase is not apparent in raw wage data, because of strong sorting responses. The association of wage premium growth and employment growth is positive, suggesting that premium growth is predominantly driven by demand side factors. Our results are robust to allowing for occupation-level changes in returns to cognitive and psycho-social skills.
Families, Neighbourhoods and Children's Educational Outcomes (with Matz Dahlberg, Torsten Santavirta and Majken Stenberg)