Are people debt averse? Our theory and experiment show debt aversion is a preference in its own right, distinct from risk, time, and loss preferences. We find debt aversion in 89% of lab participants, and it predicts debt holding in a large representative German sample.

Why is the racial wealth gap still as large today as in 1962? We show the earnings gap is the largest factor. Standard models make the two gaps equal; with non-homothetic preferences they need not be. One consequence: transfers quickly unwind unless the earnings gap closes.

Women are more likely to work in low-hours jobs, which pay less and are more exposed to minimum wages. Using an equilibrium search model, we find that the minimum wage shifts women to higher-hours jobs and reduces non-employment, narrowing gender gaps in wages and hours.

Women get more out of college than men. Not just money. In 1960s US men & women graduated at similar rates; now women lead in every racial & ethnic group, thanks to wages, marriage market, mothers’ education. Women’s chance of getting marriage offers after college increased a lot

Discussion was part of a 2025 conference organized by Peking University HSBC Business School and IER. Will be part of special issue on Demographic Change and Migration: Facts, Models and Policy Implications that the IER will publish in December 2026.

The discussion was part of a 2025 conference organized by Peking University HSBC Business School and IER. Will be part of special issue on Demographic Change and Migration: Facts, Models and Policy Implications that the IER will publish in December 2026.

Are firms risk averse? We develop a dynamic model of firm investment using quantile preferences to capture risk attitudes. Investment depends on downside risk attitudes through the quantile Euler equation. Empirical estimates provide evidence of firm downside risk aversion.

Capital is easier to build than to shed. This discussion of Suzuki's "Local Shocks and Regional Dynamics in an Aging Economy" adds asymmetric capital adjustment, shows it stays tractable in dynamic hat algebra, and illustrates numerically how it changes regional dynamics.

This discussion highlights how digitalization affects inequality through both firms’ labor-demand decisions and workers’ skill choices, pointing to new research on heterogeneous technology effects and policy design.