![Sample our Area Studies journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5922/TOC-Subject-Sample-2021-Area-Studies.jpg"})
Abstract
Income inequality is one of the most serious issues globally and China is the representative examples of this issue. Income inequality remains high in China and may negatively affects subjective well-being. This study clarifies whether income inequality affects subjective well-being in China. Using five waves of the 2010-2018 data from China Family Panel Studies, a panel data analysis reveals the following: First, general income inequality measured by provincial Gini coefficients has a significant U-shaped impact. Second, between-group income inequality, measured as income ratio between urban hukou residents and migrants with rural hukou, has a significant U-shaped impact. Third, urban-rural income inequality measured by provincial urban to rural household per capita income ratio has an inverted-U-shaped impact. To address endogeneity problems of income inequality, this study adopts instrumental variable approach. For the further robustness checks of the validity of the instrumental variable used, this study adopts the recent Conley et al. (Citation2012) bounds approach. Our results are robust after addressing endogeneity problem. One important policy implication stemming from our results is the need to adopt strategies that ensure a more inclusive society without hukou-related and urban–rural discrimination.
Acknowledgements
We deeply appreciate Professor Xinxin Ma at Hosei University for her valuable comments. We are also grateful to anonymous peer reviewers, the editor of the journal, participants at annual meeting 2020 at Japanese Association for Chinese Economy and Management Studies, The Agricultural Economics Society of Japan Annual Conference 2021, and Annual Conference 2021 at The British Association for Chinese Studies, for valuable comments.
Disclosure statement
No potential conflict of interest was reported by the authors.
Data availability statement
This research used publicly available datasets. Data can be collected from the website of Peking University Open Research Data (https://opendata.pku.edu.cn/dataverse/CFPS).
Notes
1 We also considered an alternative measure of median income. However, there are many observations with zero income, and the average BI using median income is 35.51, a value that does not seem to better reflect the actual situation. Thus, following existing studies which examined the impact of BI on SWB (e.g., Jiang et al., Citation2012; Zhang & Churchill, Citation2020), we defined BI as mean income ratio between urban hukou residents and migrants with rural hukou within the same province.
2 Stewart (2002) proposed the concept of “horizontal inequality,” defined as the existence of severe inequalities between culturally defined groups. She differentiated horizontal inequality from the normal definition of inequality, vertically ranking individuals or households and measuring inequality over ranges of individuals. She named the latter type of inequality “vertical inequality.” Horizontal inequalities are multidimensional, having political, economic, and social elements.
3 According to official statistics of the National Bureau of Statistics of China, Gini coefficients and the urban to rural household per capita income ratio have started to decline since 2008 and 2009, respectively.
4 See Ngamaba et al. (2018) for a detailed survey on the association between income inequality and SWB.
5 Schneider (2016) gives an overview of the results of existing studies. The results are classified into an overall positive effect, negative effects, or no significant outcomes.
6 Jiang et al. (Citation2012) stated that the different conclusions were possibly mostly related with the fact that the literature does not make a fine distinction between identity-related income gaps and general inequality.
7 The variable definitions are in in the Appendix.
Table A1. Definition of variables.
8 We also adopt the Lewbel (2012) two-stage least square (2SLS) approach, which is used to address endogeneity problems in the absence of appropriate external instruments. However, it must be noted that this is the next best strategy.
9 There are three possible interpretations for the result. First, official unemployment rate obtained from China Statistical Yearbook is registered unemployment rate in urban areas. This does not include rural hukou residents or unemployed persons who are not registered. Some existing studies argue that since the mid-1990s, the number of unemployed persons who are not registered as unemployed has increased rapidly, and the official unemployment rate no longer reflects the reality of unemployment very well (e.g., Marukawa, 2014; Zhang, 2003). As this study focuses on both urban and rural hukou residents, the official unemployment rate does not matter for SWB of the whole observations. Second, the unemployment rate fluctuates little. The lowest unemployment rates were 1.4 in 2010, 1.3 in 2012, 1.3 in 2014, 1.4 in 2016, and 1.4 in 2018. The highest unemployment rates were 4.4 in 2010, 4.2 in 2012, 4.5 in 2014, 4.2 in 2016, and 4.0 in 2018. The lowest mean unemployment rate is 3.18 in 2018 and the highest mean unemployment rate is 3.58. Third, unemployment rate is relatively highly correlated with income inequality. In a regression of income inequality measured by provincial Gini coefficients, BI, and URIR on state unemployment rates, controlling for country and year fixed effects, the coefficients on the unemployment rate are significant at the 1% level at 0.019, 0.273, and 0.149, respectively.
10 It is generally argued that the warning level may cause social unrest.