Inefficiency and congestion in Chinese production before and after the 1978 economic reforms

Abstract

Congestion is identified with input increments that result in output decrements. A new model in DEA (Data Envelopment Analysis) developed for identifying congesting inputs and estimating their impact is here applied to Chinese data over the period 1966–1988 for 3 industries: Textiles (labour intensive), Chemicals (capital intensive) and Metallurgical (in between these two extremes)—to identify sources and estimate amounts—where this model is used. Congestion is identified as occurring only in labour and primarily in years preceding the 1978 economic reforms. China's large labour employment problem and associated worker welfare programs suggest that some tradeoff with output reductions may have been desired in these periods. However, it is also shown how these DEA congestion models can be used to identify tradeoffs that could have been used to improve either, or both, output and labour employment. Although restricted to identifying possibilities for augmenting employment in the present paper, this approach can also be used in DEA applications where other types of input as well as output augmentations are desirable. Appendix A compares the approach in this paper with previously available approaches to congestion.