earticle

영어

Decomposition of labor productivity growth through shift share analysis is frequently used to analyze the effects of structural changes in industries on aggregate productivity growth. The recent adoption of chained index method in the system of national accounts in Korea, however, causes additivity of industry productivities as well as real value added to break down. This study compares recently proposed decomposition methods which preserve the additivity, and apply them to the long-run labor productivity data from 1963 to 2014. The results show that the differences between additive decomposition methods are mostly negligible, while traditional non-additive decomposition method by Denison overstates the contribution of industry productivity growth and understates the effect of labor input reallocation among industries. Results from the decompositions also show that labor reallocation effects among four broad industries, namely agriculture, manufacturing, construction, and services, peaked between late 1970s and early 1980s but nearly disappeared by mid 1990s. Contribution of manufacturing industry to aggregate labor productivity growth was largest in 1990s but dropped subsequently. While the contribution of service industry dropped as well, the gap between the two sectors reduced to 0.3% due to faster decline in manufacturing industry. For the last 10 years, manufacturing as well as wholesale and retail trade sectors contributed most to the growth of aggregate labor productivity, while the employment in those sectors has declined. In contrast, social welfare services as well as professional, scientific and technical activities sector has been rapidly growing in terms of employment, which means that the employment structure in Korea's service industry is changing both toward improving quality of life and helping manufacturing industry to enhance its productivity.