Labor Supply vs. Demand of Older Workers: A Structural Approach with Stated Preferences
joint with
Max Groneck
and
Johanna WalleniusMany workers retire abruptly from full-time employment, a pattern that standard life-cycle models with smooth preferences over leisure and consumption struggle to reconcile. This suggests that factors other than those governing preferences for labor supply -- such as institutional and demand-side forces, including firms' limited willingness to offer part-time work to older employees -- play an important role. Distinguishing between these supply- and demand-side factors is challenging, as both shape observed labor supply behavior. We address this challenge by developing a new identification strategy that combines observational data with stated preferences for part-time work from strategic survey questions. Our preliminary results reveal that prime-aged men do not want to work part-time, exhibiting negative willingness to pay to reduce hours worked. However, their willingness to pay for part-time work becomes positive and sizable close to retirement age. In contrast, female reservation wages are sharply rising in hours worked, indicating that women do not want to work full-time. Viewed through the lens of our structural life-cycle model, these patterns are consistent with a large labor supply elasticity for women, and a slightly smaller elasticity for men.
Presentations: Young Swiss Economists Meeting 2026; University of Groningen
Population Aging, Public Finances, and Alternatives for Retirement Reform
joint with
Frederik Bjørn ChristensenWe study retirement reforms that ensure sustainable public finances in the face of population aging. We build a structural life-cycle model with a pension scheme that includes a public pay-as-you-go pillar and a mandatory fully-funded pillar. The two pillars interact through a means-testing mechanism. The higher the fully-funded benefit, the lower the public pay-as-you-go benefit. The interaction allows us to assess a reform in which increases in fully-funded contributions and benefits reduce public pension benefits through means testing. We compare this reform to three alternatives: Increasing the retirement age, cutting public benefits, and increasing taxes to finance growing public pension expenditures. We estimate the model to Danish micro data and find that expanding fully-funded pensions to indirectly lower public pensions yields the highest welfare. Among the remaining reforms, we show that directly lowering public benefits outperforms hiking taxes and increasing the retirement age.
Presentations: Statistics Norway (SSB); Oslo Macro Conf. 2024; ISI Delhi; University of Barcelona; IIM Bangalore; Netspar Workshop 2025; Danish Central Bank; OECD Econ Department; 2nd Arne Ryde Workshop in Lund*; Annual Congress of the Swiss Society of Economics and Statistics 2024; 45th Annual Meeting of the Norwegian Association of Economists; End-Of-Year Conference of Swiss Economists Abroad 2022; Iowa State University* — (*) Presentation by co-author
Raising the Bar: Pension Age Reforms, Labor Supply, Welfare Dependence, and Taxes
joint with
Frederik Bjørn Christensen,
Svend Erik Hougaard Jensen,
and
Miguel Sousa DuarteWe study the effects of a six-month increase in the early and full pension eligibility ages in Denmark using a regression discontinuity design. We track individuals for 12 years post-reform announcement and estimate comprehensive behavioral responses in labor supply, pension and welfare claims and and both private pension and non-pension savings. We find no upstream labor effects in the years between the reform announcement and the affected eligibility ages. In contrast, employment increases and retirement is delayed at each postponed threshold, including the full retirement age, where—unlike at the early retirement ages—no discrete financial incentives can account for the response. We also document midstream and downstream labor supply adjustments, indicating that responses extend beyond narrow claiming windows. Despite evidence of both active and passive substitution into alternative welfare programs, these spillovers remain limited, and the reform yields a substantial positive net fiscal effect. Cumulative savings rise substantially, with a non-trivial large effect around the full retirement age. The overall responses are disproportionately driven by salaried workers with low liquid and pension wealth, whereas the self-employed and pension-wealth-rich individuals adjust significantly less.
Presentations: PeRCent Annual Conference 2024; Colloque Retraite et Vieillissement, Paris*; Workshop on Consumption and Saving over the Life Cycle, Iceland* — (*) Presentation by co-author
Willingness to pay for job characteristics in later life: Evidence from a stated choice experiment
joint with
Max Groneck,
Tunga Kantarci,
and
Johanna WalleniusThis paper studies older workers’ willingness to pay (WTP) for changes in working hours, flexibility, physical demands, and stress using a tailored survey module fielded in the Dutch LISS panel. Respondents aged 45–75 completed a sequence of strategic survey questions that elicited reservation wages for jobs identical to their reference job as well as for jobs differing in a single attribute. This design allows a clean separation between attachment to existing job bundles and valuations of marginal changes in job characteristics. We find substantial and asymmetric valuations: workers demand sizeable compensation to accept more demanding conditions, while willingness to pay for improvements is more modest and highly heterogeneous. Gender differences are pronounced for working hours, whereas valuations of flexibility, physical demands, and stress are broadly similar for men and women. Age gradients are limited overall, with increasing aversion to strenuous and stressful work at older ages.