What causes changes in the effects of fiscal policy? A case study of Japan

In the past two decades, the Japanese government has spent a considerable amount of money to counteract the severe recessions that have recurred since the early 1990s. Numerous studies have pointed out that the effects of these expenditures have diminished since around the 1990s. However, none of these studies has statistically explored the reasons for this diminution, which they implicitly or explicitly mention. The purpose of this study is to statistically investigate these reasons, using a threshold vector autoregression (VAR) in which the causes pointed out in the literature are adopted as the threshold. If the null hypothesis that the estimated parameters are equal under each regime is rejected, we can conclude that a given cause does affect the macroeconomic structure and, in turn, the fiscal policy effects. We then estimate the impulse response functions in both sample periods, as constructed on the basis of threshold estimates, and compare the effects of fiscal policy in each period. The following are the main results of the study. First, we found that the diffusion index of the attitudes of financial institutions toward lending and the yearly change in the annual average of the quarterly ratios of the structural primary budget balance to potential GDP significantly reject the null hypothesis; therefore, we concluded that these variables have a definite impact on fiscal expansion effects. Second, the resulting impulse response functions show that the effects are traditional, although there are some notable differences. In particular, when banks' attitude toward lending is tight and the financial condition of the government is bad, the demand-enhancing effects of government expenditure should be considered weak. In this regard, the traditional accelerator effects of private investment, the existence of liquidity-constrained households, and non-Keynesian effects are key operative concepts. (C) 2014 Elsevier B.V. All rights reserved.