اثر نابرابری درآمد بر رشد اقتصادی چین

A pilot empirical study is carried out on how income inequality affects growth through incorporating panel data information into a quarterly macro-econometric model of China. Provincial urban and rural household data are used to construct income inequality measures, which are then used to augment household consumption equations in the model. Model simulations test the inequality effect on GDP growth and its components. Results show that income inequality forms robust explanatory variables of consumption and that the way inequality develops carries negative consequences on GDP and sectoral growth.

Since undertaking market reforms, the Chinese economy has achieved sustained high growth and rapid progress in poverty reduction. The World Bank estimates that in the more than two decades since reforms started, average income per capita in China has quadrupled while more than 270 million people have been lifted out of poverty (Chen & Wang, 2001). GDP growth in China averaged nearly 10% annually for the past decade, and it has not yet shown signs of slowing down. If there seems to be a dark lining to these extraordinary achievements, it is that income inequality in the country – seen as a whole, within/between urban and rural areas, and across provinces – has also risen quite rapidly in the period (see, for example, Chen & Wang, 2001; Zhang & Kanbur, 2005; World Bank, 1997). Income inequality is generally seen to affect long-term economic growth, although there is no consensus on the direction of the effect. If income inequality affects growth positively, it is possible that the poverty-reducing impact of this growth offsets the direct adverse effect of inequality on welfare, and thus reason to tolerate relatively high inequality. On the other hand, if inequality affects growth negatively, then addressing it immediately should be an important concern. This paper investigates empirically how much and in what ways income inequality affects China’s economic growth by means of incorporating income disparity measures derived from provincial panel data of urban and rural household income into a macro-econometric model, and simulating the effects of changes in income inequality on growth. The rest of the paper is structured as follows. Section 2 gives a summary of the inequality situation in China. Section 3 briefly surveys the literature on the transmission mechanism between income inequality and economic growth. Section 4 describes our modeling approach and discusses available inequality measures which might be pertinent to our investigation. Section 5 describes the estimation results of incorporating income inequality into the macro-econometric model. Section 6 presents the results of model simulations showing the effects of inequality changes on other economic variables. The last section concludes with a summary of the findings and a brief discussion of relevant policy implications.

Macroeconometric models are incapable of tackling the issue of how income inequality affects growth empirically for lack of explicit channels relating aggregate income and consumption to income distribution. The present study runs a pilot experiment to incorporate panel data information into a macro-econometric model of China so that the issue of how income inequality affects growth can be studied through model simulations. A panel of provincial urban and rural household income data is used to construct income inequality measures, which are used to augment the urban and rural consumption equations of the China model. Simulations are then carried out on the modified model to see how future changes in income inequality would affect the macro-economy. Through model augmentation, the rapidly changing income inequality is found to exert significant impact on consumption of both urban and rural households. While rising urban income inequality holds back urban consumption in the short run, increase in the relative income level between rural and urban areas is found to stimulate household savings in the long run. Through model simulations, we observe several interesting results. We find that significant changes in income inequality – whether within-urban, within-rural, or urban–rural – carry negative effects on macro-economic stability as they cause consumption and then investment to undulate. Comparing the effects of shocking each of the urban and rural inequality measures, we find that increases in urban inequality carry more favorable (or less negative) effects than increases in rural inequality. In simulating the impact of changing urban–rural average income disparity, we see that long-run GDP growth is highest when urban–rural income gap is narrowed (i.e. rural-favorable growth), as compared with the scenario where it is widened, and that the urban-favorable growth scenario (i.e. widening urban–rural gap) would only benefit the industrial sector in the long run. These results carry significant policy implications. Specifically, our findings indicate that anti-inequality policies should prioritize narrowing the urban–rural income gap and alleviating within-rural area inequality. Policies fostering economic growth in the rural areas augmented by rural social welfare provision, such as on education and health care, are examples of such policies. Policies facilitating greater labour mobility should also be considered to further utilize the agricultural labour surplus, see also (Besley & Burgess, 2003) for more discussion on relevant policy measures. More fundamentally, our findings show the imperative of having anti-inequality policies high on the agenda, not just for economic growth, but for social stability, as widening income inequality in any respect is found to have economically destabilizing effect. Several extensions of the present study are desirable. First, it is desirable to extend the fiscal block of the model to establish explicit links between income inequality and income re-distribution policies. Secondly, it is desirable to extend the consumption block to base it on panel data entirely so as to achieve data consistency between aggregate income levels and income inequality measures. Thirdly, it is desirable to explore explicit links between income inequality and employment distribution among the three sectors of GDP. More data would be needed for this extension. Whichever direction of extension, a wider mix of time-series and panel data in macroeconometric modeling seems the desirable way forward.