The Impact of Subsidies for Small and Medium-sized Firms in Japan: A Regression Discontinuity Design Approach

The paper analyzes the causal effects of the subsidies for small and medium-sized firms in Japan (collectively called the Monozukuri Subsidy), provided by the Japanese Small and Medium Enterprise Agency. Given the uniqueness of its selection process, where a firm's acceptance is determined discontinuously by scores granted by selection committees, we utilize a fuzzy regression discontinuity design (RDD) approach in order to derive causal effects of the subsidies. In addition, our RDD estimates for different years are aggregated using a meta-analysis framework. We do not find a significant difference between treated and control firms on their value-added, value-added per worker, employment, or capital. The results suggest that its fuzzy selection process should be improved to derive robust and statistically significant estimates.