Total factor productivity growth in China: a review

Abstract

The debate on the role of total factor productivity (TFP) in China's rapid economic growth has led to the emergence of a large pool of papers on this topic. There is however hardly any consensus in the literature. This paper surveys 74 studies published from the 1990s onwards and employs meta-analysis to investigate whether the empirical findings are systematically affected by the choice of methods, selection of samples, and objectives of individual studies. Insights gained are used to draw implications for further studies.

Keywords:

• meta-analysis
• total factor productivity growth
• Chinese economy

JEL Classifications::

• O11
• O47
• O53

Acknowledgements

I thank an anonymous referee, Tim Coelli, Jinghai Zheng and participants of APPC2008 and ACESA2009 for very helpful comments. Work on this paper benefited from the financial support of the University of Western Australia (UWA Research Grant scheme) and Australian Research Council (DP1092913).

Notes

Notes

1. For a review, see Griliches (1994).

2. For comprehensive reviews, refer to Kumbhakar and Lovell (2000), Coelli et al. (2005) and Greene (2008).

3. See, for example, Baumol (1986), De Long (1988), Dowrick and Nguyen (1989) and Wolff (1991, 1996).

4. For earlier literature reviews, see Wu (1993), Wu and Yang (1999) and Fan and Zhang (2002).

5. There are also many studies that are related to the pre-reform period and hence ignored in this review. Examples include Tang (1980), Wiens (1982) and Perkins and Yusuf (1984).

6. There are also papers that investigate technical efficiency performance only, such as Kong, Marks, and Wan (1999), Zheng, Liu, and Bigsten (1998) and Shiu (2002). These are excluded from this review.

7. The number of papers increases over time. By the time of finalizing this version, for instance, several papers had appeared but could not be incorporated into the quantitative analysis. These papers include Chen, Huang, and Yang (2009), Chen et al. (2008) and Zheng, Bigsten, and Hu (2009), to cite a few.

8. See Stanley (2001) for a review.

9. These outliers are samples with extreme small or large TFP growth estimates, which are located on the two tails of the histogram in Figure 1.

10. The reliability or quality of Chinese data has, for a long time, been questioned by economists. For example, Maddison (1998), H. Wu (1993) and Holz (2004) raised the question about the accuracy of Chinese GDP statistics and hence re-estimate their own series. Zheng, Hu, and Bigsten (2009) presented a study comparing the officially recorded income with the estimated ‘potential’ income in China. Other economists have also made efforts to check the quality of employment statistics (Rawski 2000; Liu and Chan 2001; and Giles, Park, and Zhan 2005). More recently Ozyurt (2009) and Cao et al. (2009) considered the quality of both labor and capital data and examined productivity growth in the industrial sector.

11. Examples include Zhang (1991), He (1992), Chow (1993), Wang and Yao (2003), Holz (2006) and Zhang (2008).

12. The choice of techniques and their impact on efficiency and productivity estimates is still debated in the literature. For more discussions, see Hjalmarsson et al. (1996), Coelli et al. (2005) and Odeck (2007).