In macroeconomics, the effectiveness of fiscal policy is an old, but new controversial question that has been debated for many years. The controversy naturally arises if analytical frameworks, sample periods, or targeted countries are different. However, it sometimes arises, even when using similar analytical frameworks and sample periods. The purpose of this study is to explore whether revisions to the GDP data series are a source of the controversy in Japan. Since the mid-1990s, the Japanese economy has remained stuck in a liquidity trap where interest rates have fallen to zero. However, not only has the cumulative fiscal deficit increased to an unprecedented level, but many structural problems have also emerged. Under these circumstances, it is unclear whether the fiscal policy has worked effectively in the Japanese economy. When a reduced-form equation or VAR model was estimated, the empirical findings were quite different depending on GDP benchmark year we used in the analysis. When benchmark year 2011 was used, fiscal expenditure was effective under ultralow interest rates. In contrast, when benchmark year 2015 was used, it was not effective, especially since around 2010, even under ultra-low interest rates. This implies that in Japan, the effectiveness of fiscal policy must be carefully interpreted, noting which GDP benchmark year is used in the analysis.