SO₂ emissions and the environmental Kuznets curve: the case of Chinese provinces

Abstract

This paper aims at estimating the Environmental Kuznets Curve (EKC) for the sulphur dioxide (SO₂) emissions in a panel of 28 Chinese provinces. First, using a fixed effects model, econometric findings reveal an N shape EKC with a turning point of 4500 yuans (index 1990). However, a Chow test reveals a break in 1995, so that the estimation of the model indicates an increasing linear relationship between GDP per capita and SO₂ emissions. The previous results imply that the decrease of the Chinese sulphur dioxide emissions during 1996–1999 did not result from the ECK but from an exogenous public action.

Keywords:

• sulphur dioxide emissions
• environmental Kuznets curve
• Chinese provinces
• fixed effect models

JEL Classifications:

• C23
• O53
• Q53

Acknowledgements

The authors wish to thank the referee for comments and the participants of the 18th Chinese Economics Association Annual Conference ‘Integrating China into the World Economy’, at the University of Nottingham, 16–17 April 2007, and of the Seventh European Society for Ecological Economics (ESEE) International Conference ‘Integrating Natural and Social Sciences for Sustainability’, Leipzig, 5–8 June 2007, who made helpful remarks.

Notes

Notes

1. See Jiang (2003), Nolan and Rui (2004) or Woo (2007) for an overview of China's environmental problems.

2. The term ‘air pollution’ covers a wide range of problems including emissions of particulates, sulphur dioxide, nitrous oxides and carbon dioxide.

3. Note that coal dominates the picture. Today, 96% of the thermal electric production comes from coal (versus 75% in 1980).

4. Acid rain is formed when SO₂ emitted from coal burning, combines with NO_x (nitric oxide), present in the atmosphere from motor vehicles and home furnace emissions, and with water vapour, to form sulphuric acid (H₂SO₄) and nitric acid (HNO₃).

5. A pH of 7.0 is the absolute neutral. Below this amount, it is acidic; so that any rain falling with pH below 5.6 is considered acid rain.

6. During the 1990s, the WHO estimates that excessive concentrations had a deep impact on mortality with 376,000 persons dying each year.

7. Employed by Grossman and Krueger (1991, 1994), Lucas, Wheeler, and Hettige (1992), Shafik and Bandyopadhyay (1992), Selden and Song (1994), Holtz-Eakin and Selden (1995), Suri and Chapman (1998).

8. This result is obtained from the cubic polynomial function derivative, a + bx + cx ². The values of the turning points are then: with a = 3β ₃; b = 2β ₂ and c = β ₁.

9. We refer only to this type of low polluting energy because the other types, such as wind and solar powers, are not very developed, and nuclear energy, at the time of writing, only started production 10 years ago. Moreover, we do not integrate the hydroelectric production of electricity owing to a geographical contingency bias. Indeed, South-West provinces are the largest hydroelectric producer, but the coal used in these provinces has the highest sulphur content. In these conditions, the hydroelectricity coefficient is significant and positive, but it is a biased result.

10. Even if we don’t use SO₂ emissions per capita, the econometric results are similar. We prefer SO₂ emissions in absolute terms because, what is at stake, is the real impact on the Chinese environment. The sources of SO₂ emissions locate quasi-systematically near big cities so that their effects are condensed on a relatively small area. The amount of rural population in the hinterland only has a relatively small impact. For instance, most acid rain in Xingjian province appears near its capital Urumqi. If we used SO₂ emissions per capita, it would then tend to hide the reality of ecological impacts.

11. The FDI are no more significant, confirming our presumption that it is an over-aggregated indicator for our purpose.

12. Fujian is very close to the GDP value.

13. We use a Chow test by separating the sample in two sub-periods: 1985–1995 and 1996–2003. We found (see Table 1) that C = 13.92 > = 1.45 so that, the coefficient stability hypothesis must be rejected.

14. In the Ninth Five-Year Plan Period (1996–2000), the Chinese State Environmental Protection Administration (SEPA) began to promote a policy of total emission control (TEC). National SO₂ TEC targets were established. SEPA assigned individual TEC targets to provinces, autonomous regions and municipalities. The provincial governments subsequently assigned TEC targets to local governments and/or emission sources.

15. The 10th five-year plan (2001–2005) projected to contain emissions 20% under the 2000 level. Instead, the real amount of SO₂ discharged into atmosphere grew by 28%.