Working Papers and Works in Progress
March 2026
This paper studies the trade-offs associated with income redistribution in an overlapping generations model in which marginal propensities to save increase with permanent income. By transferring permanent income from high savers to low savers, redistribution lowers aggregate savings and depresses investment in capital. If more capital is welfare improving, the government faces a trade-off between redistribution and investment whose size depends on the distribution of marginal propensities to save across the income distribution. I characterize this trade-off theoretically and quantify its magnitude. In a simple model with endogenous labor, I show that the welfare costs of a standard labor income redistribution policy can be decomposed into the familiar equity-efficiency trade-off and a new term corresponding to this permanent income redistribution-investment channel. I calibrate a quantitative overlapping generations model to match empirical estimates of marginal propensities to save from U.S. household panel data. In a back-of-envelope calculation using the simple model, the permanent income redistribution channel accounts for about 30 percent of the welfare cost of the labor income distortion channel. In the quantitative model, this ratio rises to between 44 and 47 percent.
March 2026
We derive a new criterion for evaluating over- or under-saving in over-lapping generations economies with both ex-ante and ex-post household heterogeneity. Our criterion, which we call the Savings Wedge, nests the traditional `r vs. g' (Golden Rule) criterion as a special case, but accounts for how aggregate savings affects households' exposure to uninsurable risk and shifts the income distribution, and can be thought of as a risk and redistribution robust Golden Rule criterion. A second-order approximation of the Savings Wedge yields a tractable formula of measurable statistics. Using PSID data from 1978 to 2024, we measure the Savings Wedge in the United States and compare our measure to the traditional Golden Rule. The latter implies mild under-saving for most of our sample as r>g. Relative to the Golden Rule, our criterion suggests substantially more under-saving, driven by both high levels of capital income risk and capital income inequality being greater than labor income inequality.
Publications and Papers Under Review
Journal of Political Economy: Macroeconomics (September 2025)
Revise and Resubmit: Quarterly Journal of Economics