Macroeconomic Effects of Government Spending in Japan

Abstract

It is often recognized that government spending shocks are well anticipated before they actually materialize. This paper uses a new approach that is possible to identify a government spending shock as both an unanticipated and an anticipated (i.e. news) shock to study the macroeconomic effects of government spending in Japan. We find that, in the whole sample 1968Q1- 2010Q1, there is a clear difference between an unanticipated government spending shock and an anticipated one: the former significantly increases both GDP and consumption while the latter does not. We also find that government spending is effective in terms of simulating the economy in the pre-bubble period, but in the post-bubble period it is not and even aggravates the business cycle fluctuations at some horizons.