Identification, estimation and inference in Panel Vector Autoregressions using external instruments

TL;DR

This paper introduces a new method for identifying and estimating Panel Vector Autoregressions using external instruments, enabling more accurate inference of dynamic relationships in panel data.

Contribution

It proposes a novel identification approach inspired by SVAR-IV, introduces the $-ATE, and demonstrates reliable inference methods through simulations and an application to fiscal multipliers.

Findings

Confidence sets based on Anderson-Rubin statistics are reliable.

State-level military spending has a fiscal multiplier above one.

Effects of military spending persist into the following year.

Abstract

This paper proposes an identification inspired from the SVAR-IV literature that uses external instruments to identify PVARs, and discusses associated issues of identification, estimation, and inference. I introduce a form of local average treatment effect - the $\mu$-LATE - which arises when a continuous instrument targets a binary treatment. Under standard assumptions of independence, exclusion, and monotonicity, I show that externally instrumented PVARs estimate the $\mu$-LATE. Monte Carlo simulations illustrate that confidence sets based on the Anderson-Rubin statistics deliver reliable convergence for impulse responses. As an application, I instrument state-level military spending with the state's share of national spending to estimate the dynamic fiscal multiplier. I find multipliers above unity, with effects concentrated in the contemporaneous year and persisting into the following year.