An Investigation of the Stagnation in Productivity of the Japanese Economy in the 90s Using the Hicks-Moorsteen Productivity Index

本研究基于希克斯-莫尔斯汀生产率指数法，对生产率进行了新的分解分析，并将其应用于90年代日本经济的生产率停滞。与基于索洛残差或托恩奎斯特指数的标准分析不同，所提出的生产率分解分析具有允许存在生产效率低下的优点。就我们的目的而言，这是有吸引力的，因为它使效率低下对生产力的影响可以衡量。在形式上，我们表明，希克斯-莫尔斯汀生产率指数可以分解为因素的技术进步，效率，规模和投入产出的混合效应。实证分析使用了包括日本在内的OECD 17国（1965-1990年）、日本47个都道府县（1981-2000年）、日本宏观经济和产业的时间序列（1981-1998年）三个数据集。这些结果揭示了日本经济中生产率变化的一些有趣方面。第一，80年代的生产率仍处于追赶发达国家的过程中，在整个80年代，效率的提高使生产率接近发达国家的水平。第二，90年代的生产力停滞主要是由于技术进步不活跃，而1996年的生产力暂时恢复是由于效率提高，反映了需求的增加。第三，信息技术的进步可能会提高技术进步因素，从而在90年代后期提高制造业的几个部门的生产率。

This study develops a new decomposition analysis of productivity based on the Hicks-Moorsteen productivity index approach, and applied it to investigating the stagnation in productivity of the Japanese economy in the 90s. Unlike the standard analysis based on the Solow residuals or The Tornqvist index, the proposed decomposition analysis of productivity has an advantage that the existence of productive inefficiencies is allowed. For our purpose, this is appealing because it makes the impacts of inefficiencies on productivity measurable. Formally, we show that the Hicks-Moorsteen productivity index can be decomposed into the factors of technical advance, efficiency, scale and mix effects of inputs and outputs. The empirical analysis is conducted using three data sets : a panel on OECD 17 countries (1965-1990) including Japan, a panel on 47 Japanese prefectures (1981-2000), and time series of the Japanese macro economy and industries (1981-1998). Those results reveal some interesting aspects of productivity change in the Japanese economy. First, productivity in the 80s was still on the process of catching up the developed countries in the sense that improvement in efficiency raised the productivity toward the level of the developed countries throughout the 80s. Second, the stagnant productivity in the 90s is mainly attributable to inactive technical advance while a temporary recovery of productivity in 1996 was driven by the improvement in efficiency reflecting an increase in demand. Third, an advance in IT might raise the productivity of several sectors in the manufacturing industry in the late 90s by boosting the technical advance factor.