1 Introduction

When it comes to the development strategies for poor provinces in China, the first set of policy options usually involve investment in infrastructure and education. The central and local governments in China are often the major entities to implement such strategies, which are reflected in the official economic development plans of all levels of governments. One argument supporting these strategies is that the less-developed provinces suffer from the lack of natural endowments that are indispensable for business startups. And it is the responsibility of governments to provide necessary public goods, such as road and education, to lift them out of the constraints of nature.

Despite the potentially beneficial impact of a natural resource on economic prosperity, natural-resource abundant economies tend to grow at a slower pace, e.g., Angola, Nigeria, Sudan, and the Congo. In contrast, countries that had only limited access to natural resources, such as Japan, Hong-Kong, Korea, Singapore, and Switzerland, experienced remarkably high economic growth rates. This is the phenomenon known as the Resource Curse, which is referred first by Sachs and Warner (1995). And then it leads to hot research on the relationship between a resource and economic growth, such as Leite and Weidmann (1999), Gylfason (2001), etc. Summing up the literature, Frankel (2010) concludes that there are six channels that some have suggested could lead to sub-standard economic performance, which are long-term trends in world commodity prices, volatility, crowding out of manufacturing, civil war, poor institutions, and the Dutch Disease.

The difference in natural resource and economic development of different regions inside a country can be substantial. Natural resource and economic development also often exhibit a negative relationship. Partridge et al. (2012) investigate the causes of county poverty rates in Appalachia America from the respective of the natural resource curse that explains such a negative relationship. James and Aadland (2011) make an empirical investigation of U.S. counties to test the hypotheses of natural resources curse, finding supportive evidence. Following the same reasoning of Sachs and Warner (2001), Zhang et al. (2008) find that the curse of natural resource also exists in China by constructing a resource intensity variable that is a single standard coal index composed of coal, oil and natural gas.

However, there is no consistent view on the natural resource curses because it is difficult to distinguish the effects of a natural resource on economic growth among many other factors that also affect economic growth, such as geography, institution, etc. Bao et al. (2002) investigate the effects of geography on regional economic development in China. They find China’s regional development disparities can be explained mainly by the inflow of FDI and the movement of labor from the inland areas to coastal areas due to the differential of factor returns which is in turn accounted by the geographic factors, such as the distance to main ports, the proportion of population with 100km of coastline and ocean navigable rivers. Rodrik et al. (2004) further demonstrate that it is institutions that matter more than geography. Once institutions are controlled for, conventional measures of geography have at best weak direct effects on incomes. Mehlum et al. (2006) resort the divergent economic performance to differences in the quality of institutions. Natural resources can exert a positive influence on economic growth when institutions are producer-friendly, and, on the contrary, the natural resource curse becomes true when institutions are grabber-friendly.

In this paper, we aim to disentangle the effects of natural resources on economic growth, controlling for other possible factors. Especially, economic reformation and openness of China for the last three decades are considered as the most essential ingredients for boosting fast economic growth. However, even though we add variables to reflect reformation and openness, the effects of natural endowments are still significantly positive for regional development in China.

The structure of the paper is organized as follows: section 2 gives the model specification and explains the possible effects of each variable on economic development level and growth; section 3 provides data description and simple comparative analysis of different province of China, and section 4 reports the empirical results; section 5 is the conclusion.