Éléments de comptabilité de la croissance chinoise

This paper investigates the current state of growth accounting in China. The growing empirical literature on Chinese economic growth is based on several stylised facts which stress high but erratic long-term economic growth, strong accumulation of capital and rapid population growth. From these preliminary observations, growth accounting exercises found in the existing empirical literature suggest that production factors and total factor productivity have had different contributions to growth since the beginning of the People's Republic of China. However, a closer look at this literature reveals some weaknesses which question its validity. Two constraints appear particularly important: one is of a statistical nature, because Chinese data are more or less reliable; the other is of a methodological nature, because existing growth decomposition is not based on econometric estimations of the underlying production function. Considering those weaknesses, we propose an exercise in growth accounting on a long-term basis and an estimation of the aggregate production function for the period from 1978 to 1995 in order to check conclusions which can be considered as robust. Our results tend to confirm the major role of capital accumulation in accounting for growth during the pre-reform period and the relatively important contribution of total factor productivity growth in the current economic growth of China. They also show that, in spite of the improvements made by the Chinese State Statistical Bureau in measuring economic change, provided data remain dubious and may consequently lead to biased estimation of the amplitude of growth and its sources. This uncertainty constitutes a serious handicap in the search for a relevant analysis of Chinese growth both in the long run and for recent years and suggests that statistical measurement of Chinese growth performance needs to be improved in order to be able to draw valid inference about its sources. Our results finally show that during the period from 1978 to 1995, the productivity gains can be partly, but not fully explained by a process of labour reallocation from low-productivity to higher-productivity activities.