Incorporating undesirable outputs into Malmquist TFP indices with an unbalanced data panel of Chinese power plants

Abstract

In this article, we examine the effects of undesirable outputs on Malmquist Total Factor Productivity (TFP) indices. Our empirical work uses an unbalanced panel of 1626 real observations on coal-fired power plants in China during 1996–2002. In order to meet the requirement of a balanced panel for calculating the Malmquist indices, an innovative fake unit approach has been introduced. To the authors' knowledge, no other published papers on TFP change have used this method before. Our final results show that (1) the original growth path of the Chinese electricity till 2002 heavily depended upon an increase of resource input and (2) the environmental TFP change of Chinese coal-fired power plants was about 2–3% (at least with respect to SO₂), which mainly comes from technical change rather than efficiency change during the report period.

Keywords:

• Malmquist index
• data envelopment analysis
• environmental TFP change
• fake unit approach
• Chinese electricity

JEL Classification:

• D24
• L94

Acknowledgements

The authors have benefited from comments received at previous presentations in Cambridge University. The views expressed in this article belong to the authors and do not necessarily reflect the views and policies of the Asian Development Bank (ADB) and its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this article and accepts no responsibility for any consequence of their use.

All remaining errors are the authors’.

Notes

¹Strong disposability of outputs implies that given an input vector x, if an output vector y can be produced, then y* can also be produced as long as y* ≤ y. Weak disposability of outputs means that if y can be produced, then can also be produced (Fare et al., 1989).

²Data for total generation capacity are from China Energy Databook (CED; 2004).

³Because the focus of this article is on environmental performance analysis of power plants, not on the effects of labour quality, therefore, the labour quality is not differentiated in this article. Similar treatment can be seen in the literature, such as Fare et al. (1985, 1989, 1996), Korhonen and Luptacik (2004), Yaisawarng and Klein (1994) and Yang and Pollitt (2009).

⁴Data for calculation are from China Energy Databook (2004).