چشم انداز استانی بر نابرابری درآمدی در جامعه شهری چین و نقش املاک و درآمد کسب و کار

This paper tries to investigate the role of non-wage income in explaining the income inequality in urban China. Our findings show that the contribution of income sources to inequality is different between the provinces with different extent of inequality. We find that in the coastal provinces, the contribution of wage income to inequality is decreasing, while the contribution of business and property income is increasing and getting more important; in contrast, in the western provinces, the role of wage income is larger than the other provinces, while the role of business and property income is smaller and remains unchanged. Our empirical results also suggest that the provinces with higher share of business and property income have high income inequality.

In previous studies, the data used to calculate income inequality in China mainly came from household survey data and grouped data. The China Household Income Project (CHIP) surveys provide micro (unit record) data, whereas the National Bureau of Statistics (NBS) (2006, 2007, 2008, 2009, 2010,2011-a) surveys are available as grouped data. The CHIP has conducted four waves of household surveys in the years of 1988, 1995, 2002 and lastly, in 2007. Eichen and Zhang (1993) describe the 1988 survey and Li, Luo, Wei, and Yue (2008) describe the 1995 and 2002 surveys. Further, a study by Luo, Li, Sicular, Deng, and Yue (2013) provides information for the 2007 survey. Major findings based on the 1988, 1995 and 2002 CHIP surveys can be found in Démurger, Fournier, and Li (2006), Gao and Riskin (2006), Gustafsson et al., 2008a and Gustafsson et al., 2008b, and Khan and Riskin, 1998, Khan and Riskin, 2005 and Khan and Riskin, 2008. Key findings based on the recent 2007 CHIP survey are presented in Deng and Gustafsson (2011), Ding, Li, and Myers (2011), Gustafsson and Ding (2011), Knight, Sicular, and Yue (2011), Li, Luo, and Sicular (2011), Liu and Li (2011), Sato, Sicular, and Yue (2011), and Yang, Démurger, and Li (2011). Needless to say, it is better to use micro data in order to analyze the income distribution. If one has access to the relevant household level unit record data, then there are accurate computational methods for estimating inequality measures directly from the data. Despite its merits, a shortcoming of the micro data, such as the CHIP survey, is that it includes a relatively smaller number of provinces than the NBS survey. For example, the CHIP survey covered 12 and 16 out of 31 Chinese provinces in the 2002 and 2007 surveys, respectively. In addition to this limitation, the CHIP survey has been conducted for specific years. In contrast, the NBS survey includes national coverage and has been carried out every year since the mid-1980s. Above all, unit record data from household surveys, particularly the NBS surveys, are rarely accessible to researchers, while grouped data from NBS surveys are available. Consequently, grouped data from NBS surveys are widely used to estimate the poverty and inequality measures in China. We use POVCAL to estimate the Gini coefficients for urban households in China at a provincial level using grouped data. POVCAL has been developed by the World Bank to estimate the poverty and inequality measures with grouped data. Previous studies using POVCAL to estimate China's poverty and inequality can be found in Ravallion and Chen (2007) and Chen and Wang (2001). Other techniques have also been proposed to analyze grouped data. For example, Wu (2003) and Wu and Perloff (2003) developed a maximum entropy density method for grouped data. Wu and Perloff (2004) applied the maximum entropy density method to grouped data from the NBS survey in order to calculate the Gini coefficients in China. Similarly, Chotikapanich, Giffiths, and Rao (2007) developed the generalized Beta distribution for grouped data, and Chotikapanich, Rao, and Tang (2007) applied the method to grouped data from the NBS survey in order to show that the generalized Beta distribution is a good parametric choice in estimating the inequality measures of China. A recent OECD working paper by Herd (2010) applied the method in Chotikapanich, Rao, and Tang (2007) to grouped data from the NBS survey in order to estimate the income inequality in China. Previous studies using grouped data from the NBS survey focused on only the methodology they developed, and attempted to examine the trends of income inequality in China. However, previous studies have not made an effort to estimate the extent of income inequality in China at the provincial level. It is not easy to estimate the inequality measures for all Chinese provinces for several years due to the difficulties in collecting grouped data for households in 31 Chinese provinces. In addition to the difficulties in collecting data, it is to some extent very much labor and time-intensive. Nonetheless, the important reason to estimate inequality measures for China at the provincial level is that many of its provinces by themselves would be equivalent to the country. According to the The Economist (2011, February 26th), for example, Guangdong's GDP is almost as big as that of Indonesia; the output of both Jiangsu and Shandong exceeds that of Switzerland's. Shanghai's GDP per person is as high as that of Saudi Arabia. At the other extreme, poor Guizhou has an income per head close to that of India. Besides the heterogeneities in size, openness, geographical location and level of economic development, policies for economic growth sought by local governments are also diverse across the provinces of China. For instance, the economic growth model of Guangdong is in sharp contrast to that of Chongqing. The ‘Guangdong Model’ places a strong emphasis on the role of private enterprises regarding economic growth, whereas ‘the Chongqing Model’, pursued by the expelled leader Bo Xilai, stresses the dominant role of the state sector and a fair distribution of income. As a result, the extent and causes of income inequality in China are different between the provinces. The objective of this paper is to investigate the role of non-wage incomes in inequality in urban China. Our paper is related to Deng and Gustafsson (2011) and Song, Storesletten, and Zilibotti (2011). Utilizing CHIP data, Deng and Gustafsson (2011) investigate the source of income inequality in urban China and find that one of the major channels for the increase in inequality between 2002 and 2007 was the rapid increase in business income. A theoretical two-sector model, developed by Song et al. (2011), suggests that the rising inequality within urban areas may be due in part to the slow growth of wages relative to entrepreneurial income. They also suggest a positive correlation between the Gini coefficient at the provincial level in 2006 and the employment share of domestic private enterprises. However, the positive correlation between the Gini coefficient and the employment share of the private sector is not robust because they simply used the provincial Gini coefficients for a single year. Our findings suggest that provinces with greater exposure to non-wage activities have higher income inequality. On the contrary to the argument of Song et al. (2011), we find that the employment share of the private sector is not related to income inequality. We also show that the contributions of income sources to inequality are different between the provinces: in the coastal provinces, the role of wage income to urban income inequality in China is decreasing, while the role of business and property income is increasing and getting more important; in contrast, in the western provinces of China, the role of wage income is larger than that of the other provinces, while the role of business and property income is smaller and remains unchanged. The rest of the paper is organized as follows. In Section 2, we describe the method in estimating the provincial Gini coefficients with grouped data for urban households in China. In this section, we also provide information related to the data employed in the analysis. In Section 3, we compare our estimated measures for urban income inequality in China as a whole with other estimates. We then examine the extent of income inequality in each province and show that the sources of urban income inequality are different between the provinces with different extents of income inequality. Finally, we try to show that the provinces with higher property and business income have high income inequality. Conclusions are drawn in Section 4.

We have tried to investigate the causes of income inequality in urban China with a provincial perspective. In doing so, we constructed a panel data for the provincial Gini coefficients for urban households. We found that inequality in urban areas is higher within the coastal provinces than within the interior provinces. We also showed that the contribution of income sources to inequality in urban China is different between the provinces with different extents of inequality. In general, we found that in the coastal provinces, the contribution of wage income to income inequality is decreased, while the contribution of business and property income is increased and getting more important. In contrast, in the western provinces, the role of wage income is larger than the other provinces, while the role of business and property income is smaller and remains unchanged. Finally, we showed that the provinces with higher share of business and property income have high urban income inequality. Previous studies have extensively tried to investigate issues, such as the relationship between inequality and growth, as well as openness to trade and inequality. However, researchers have faced a limitation to apply these issues to a country like China due to a lack of appropriate income inequality measures at the provincial level. China is a vast country and many of its provinces by themselves are equivalent to a country. Moreover, the extents and causes of income inequality, as well as the level of economic development are different between provinces. The findings of this paper could shed light on the empirical researches over old but still important issues.