The effects of government spending under anticipation: the noncausal VAR approach

Abstract

Fiscal foresight, economic agents receiving information about future fiscal policy, affects the consistency of results about the causal effects of government spending. This study explores the propagation of government spending shocks using a noncausal VAR model that allows for anticipation of exogenous fiscal policy changes. Overcoming the issue of insufficient information, the government spending shock is extracted from an anticipated error term by using institutional information about the conduct of fiscal policy. In addition, the approach nests the conventional causal structural VAR as a special case. In the U.S. economy, the identified spending shock comoves with defence expenditures. The shock increases consumption, employment and output one and a half years prior to its materialisation in government spending. After the shock arrives, real wages respond positively while investment turns negative. The estimated fiscal multiplier is close to unity.