Lecture Title: Micro Effects of Macro Shocks: Causal State-Dependent Local Projections

Speaker: Wang Weining, University of Bristol, UK

Lecture Time: April 7, 2026, 14:00

Lecture Location: Room 206 of the School

Abstract:

State-dependent local projections (LPs) are widely used to estimate how macroeconomic shocks affect heterogeneous households and firms, yet their causal interpretation has recently been questioned. Researchers show that the validity of state-dependent LPs hinges on a high-level restriction on the horizon-specific conditional mean and characterize the classes of micro-macro data-generating processes under which this restriction holds and fails. Researchers arguethat this restriction is broadly consistent with first-order perturbation solutions of heterogeneous-agent models in macroeconomics and heterogeneous-exposure models in macro-finance, in which case state-dependent LPs recover causal impulse responses. However, the parametric state-dependence specifications that dominate empirical practice generally destroy this causal interpretation, so recovering the causal impulse response requires nonparametric estimation of state dependence. Researchers develop a nonparametric sieve estimator that restores the causal interpretation and delivers valid uniform inference in macro–micro panels. In an application to monetary policy, allowing nonparametric state dependence materially changes conclusions about heterogeneous transmission.

Speaker Introduction:Dr. Weining Wang is currently Professor of Economics at University of Bristol. Previously held Chair Professorship in Econometrics at University of Groningen Faculty of Economics and Business, and Chair Professorship in Financial Econometrics at Department of Economics and Related Studies at University of York, UK. Received Ph.D. in Economics from Humboldt University of Berlin. Research interests lie in econometrics, particularly in nonparametric and semiparametric methods, high-dimensional econometrics, network models, and time series analysis. Published in leading academic journals including Annals of Statistics, Journal of the American Statistical Association, Journal of Econometrics, Journal of Business & Economic Statistics, and Econometric Theory. Research focuses on panel data

models, high-dimensional time series, and other advanced econometric methods, with applications to important economic and financial problems such as systemic risk analysis, asset pricing of financial derivatives, and social network analysis.