Anticipated Fiscal Adjustment and Identification of Vector Autoregressions

Due to the complex nature of fiscal policy decisions and implementations, fiscal vector autoregressive models have suffered from an underidentification problem that has caused substantial debate about measuring the efficacy of fiscal stimulus. To help resolve this issue this paper incorporates additional information regarding how private agents’ anticipation of fiscal adjustment in the short and medium run to finance the debt innovation induced by an exogenous government spending shocks. I employ robust identification strategies, including pure sign restrictions, elasticity bound restrictions, and frequency domain analyses, to trim the set of admissible model solutions to qualitatively homogenous estimates. Under the reasonable a priori assumptions on fiscal news perceived by economic agents, I arrive at the following results: (i) government spending shocks accompanying with anticipation of transfer adjustments nest the multiplier values documented by traditional VAR literature on fiscal policy (ii) government spending shocks associated with anticipation of distortionary tax adjustments substantially hinder stimulus effects of the fiscal intervention (iii) spending reversals, dynamic fiscal financing achieved by further reduction of government spending in the future to the initial spending hike, are rare options for fiscal financing in the postwar U.S. data, and exogenous increases in government spending are rather financed either by tax hikes or by transfer cuts.