Website CV Paper
Fields
Macroeconomics, Labor
Job market paper
Distributional Dynamics
We develop a method to estimate joint distributions at quarterly (business-cycle) frequency by combining heterogeneous microdata sources. The method recovers the full joint distribution even when surveys observing all variables jointly are infrequent and other sources observe only subsets. We treat distributions as functional time series, summarize their dynamics through latent factors, and estimate a Bayesian state-space model accommodating differences in survey frequency, coverage, and measurement. Combining U.S. microdata and macroeconomic time series, we construct the first quarterly estimates of the joint distribution of household consumption, income, and wealth from 1962 to 2024. Validation exercises demonstrate reliability despite sparse observations.