Dynamic Causal Effects in a Nonlinear World: the Good, the Bad, and the Ugly

A-Tier
Journal: Journal of Business & Economic Statistics
Year: 2025
Volume: 43
Issue: 4
Pages: 737-754

Authors (2)

Michal Kolesár (not in RePEc) Mikkel Plagborg-Møller (not in RePEc)

Score contribution per author:

2.011 = (α=2.01 / 2 authors) × 2.0x A-tier

α: calibrated so average coauthorship-adjusted count equals average raw count

Abstract

Applied macroeconomists frequently use impulse response estimators motivated by linear models. We study whether the estimands of such procedures have a causal interpretation when the data generating process is in fact nonlinear. We show that vector autoregressions and linear local projections onto observed shocks or proxies identify weighted averages of causal effects regardless of the extent of nonlinearities. By contrast, identification approaches that exploit heteroscedasticity or non-Gaussianity of latent shocks are highly sensitive to departures from linearity. Our analysis is based on new results on the identification of marginal treatment effects through weighted regressions, which may also be of interest to researchers outside macroeconomics.

Technical Details

RePEc Handle
repec:taf:jnlbes:v:43:y:2025:i:4:p:737-754
Journal Field
Econometrics
Author Count
2
Added to Database
2026-01-29