No.21 Synthesizing the Empirical Evidence on the Productivity of Public Capital in Japan: A Meta-Analysis

Over the past few decades many attempts have been made to measure the output elasticity of public capital or infrastructure, with varying results. This paper focuses on the accumulation of empirical analyses in Japan that have been overlooked in previous meta-analyses and quantifies the impact of public capital on output in the country through a meta-analysis of 1,038 estimates from 44 studies. No significant evidence of publication bias was found. The output elasticity of infrastructure in Japan is positive at 0.146, and this value is strongly influenced by heterogeneity of infrastructure such as project entity, time frame, region, and sector. Projects undertaken solely by local governments or in non-metropolitan areas have relatively higher elasticities, but their value declines over time. Social infrastructure such as housing, sanitation, and education have relatively low elasticities compared to other sectors. These results add new insights to the debate on the productivity of public capital and provide a benchmark as the first meta-analysis of infrastructure output elasticities in Japan.

Keyword: Public capital, Infrastructure, Japan, Productivity, Output elasticity, Meta-analysis, Publication bias