US fiscal policy shocks: Proxy‐SVAR overidentification via GMM
提出利用广义矩方法（GMM）对向量自回归（VAR）模型进行完全或几乎完全的工具变量化，从而实现过度识别。以1948-2019年美国财政VAR为例，该方法提高了脉冲响应函数和乘数的估计精度，并能在缺乏工具变量时估计非财政冲击的影响。
Summary Using external instruments, one can recover the effects of individual shocks without fully identifying a vector autoregression (VAR). We show that fully or almost fully instrumenting a VAR—that is, using an instrument for each shock—allows one to overidentify the model by incorporating the condition that the structural shocks are uncorrelated, via the generalized method of moments (GMM). We apply our approach to a fiscal VAR for the United States over 1948–2019, where the overidentifying restrictions are not rejected. The overidentified structural vector autoregression (SVAR) yields (a) greater precision in estimating impulse response functions and multipliers and (b) measures of the effects of nonfiscal shocks even when there is no instrument for them.