Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper considers estimation and inference of time-varying generalized impulse response functions (TV-GIRFs) for time-varying vector autoregressive (VAR) models. We use the local linear kernel method to estimate time-varying model coefficients, propose an easy-to-implement estimator for TV-GIRFs, and then establish its asymptotic properties for inferential purposes. Extensive simulation experiments show that our estimation method works well in finite samples. To demonstrate the empirical relevance, we apply the proposed TV-GIRFs to estimate the time-variation in U.S. government spending multipliers and the time-varying volatility spillovers among five major Asian stock markets, respectively.