Working Papers

Abstract: Technological progress in the form of automation boosts productivity, but can cause adverse labor market outcomes for transitional generations. I study the role of unions, and labor adjustment costs that firms face more broadly, in shaping the evolution of wages and employment of workers exposed to labor replacement during the automation transition. Using variation across local labor markets in the U.S. since 1980, I first document that unionization gives rise to intergenerational redistribution by shifting the incidence of wage and employment decline from older, incumbent to young, incoming cohorts. Moreover, unions accelerate the overall employment decline in automating occupations, thereby inducing faster labor reallocation. I develop a quantitative equilibrium model of technological change and unionization that jointly rationalizes the two empirical observations through the impact of union-imposed firing costs on firms' intertemporal choice how to optimally adjust their workforce over time when gradually adopting automation. Within automating occupations, unions reduce the welfare cost of automation of older workers along the transition by up to 4\% of permanent consumption by lowering their layoff risk and wage decline. The impact is shifted to young workers, raising the welfare costs of cohorts entering during the transition by up to 2\%. Incoming workers endogenously respond to automation by entering non-adopting occupations which limits the welfare impact on them. The impact of high unionization spills over into non-adopting occupations as the accelerated reallocation of labor depresses wages there.
[pdf]

The Medical Expansion, Life-Expectancy and Endogenous Directed Technical Change,
with Dirk Krueger and Alexander Ludwig.

Abstract: We build a quantitative theory of income growth, the increase in life expectancy in the last two centuries, and the emergence and expansion of a modern health sector in the 20th century. To do so, we develop a two-sector overlapping generations model with endogenous and directed technical change in which income growth, life expectancy, and technological progress in the health sector and the final goods sector, as well as relative price of health goods are jointly determined in general equilibrium. The model interprets the facts as three phases of a dynamic equilibrium in which households are initially poor and the quality-adjusted price of health goods is prohibitively high so that demand for health goods is zero, life is short and life expectancy stagnant. As income grows, fueled by technological progress, households start consuming basic health goods, life expectancy starts to rise, and directed technological progress eventually, with a delay of ca. 100 years, leads to the emergence and expansion of a modern health sector. We find that approximately 30% of the increase in life expectancy from 1940 onward can be attributed to the modern health sector, suggesting that the increased expenditures on basic health goods (e.g., better food and hygiene) played a major role in the expansion of life expectancy even in the 20th century. We decompose the increase in the relative price of health goods into two components: rising household demand for health goods relative to final goods; and productivity growth in the modern health sector relative to the final goods sector, driven by endogenous technological progress. We show that between 1940 and 1980, both components contribute roughly half of the overall increase in the health price, but after 1980, technological progress in the modern health sector accelerates and becomes the dominant force.
[CEPR WP] [pdf]

Dynamics of the Wealth Distribution in the Presence of Higher-Order Earnings Risk.

Abstract: This paper introduces higher-order earnings risk consistent with recent empirical findings into a benchmark heterogeneous-agent macro model to examine its implication for the distribution of wealth. I find that higher-order earnings dynamics induce higher earnings inequality driven primarily by persistent earnings losses at the bottom. Poor households respond by strongly cutting consumption leading to more consumption and less wealth inequality which reinforces the known issue of generating the empirically observed wealth dispersion in this class of models. In addition to lower overall consumption, the higher-order earnings moments, particularly excess kurtosis, are passed through to consumption dynamics of the poor. Both effects combined mean that those households are willing to pay up to 1.7% of permanent consumption to avoid higher-order earnings risk. Moreover, the latter effect induces consumption dynamics of the poor to be predominantly driven by idiosyncratic earnings changes which significantly reduces the correlation between their consumption and aggregate output. Since wealthier households are not affected strongly the implications for the aggregate dynamics of the economy are negligible. Methodologically, I develop a new General Polynomial Chaos Expansion approach to solve for the aggregate dynamics of this class of models, and contrast its efficiency with previous methods.
[pdf]

Publications

Work in Progress