Income Redistribution in Urban China by Social Security System-An Empirical Analysis Based on Annual and Lifetime Income

Article excerpt

1. INTRODUCTION

Following 30 years of economic growth since 1978, China has made significant progress in raising the living standards of urban and rural residents and decreasing the number of people living below the poverty line. The poverty rate declined from 53% in 1981 to 8% in 2001 (Ravallion and Chen 2007). (1) On the other hand, a dramatic widening of the income gap occurred over this period. According to the World Bank (2003), the Gini coefficient in China rose from 0.3 in the 1980s to 0.42 in 1993, which was the most rapid among developing countries. (2) Moreover, several studies estimated measures of long-term income inequality in China and found an upward trend (Ravallion and Chen 2007; Meng et al. 2010). Li and Luo (2011) used the latest household survey data to correct the potential biases caused by problems such as the difference in living costs between urban and rural areas, under-representation by high-income households in the sample, and a Gini coefficient as high as 0.485. Chen, Ravallion, and Wang (2006) found that certain subgroups have been adversely affected or have been unable to participate in the new economic opportunities because of their lack of skills, long-term illness, or disability. Some of the "left behind" households were poor at the start of this study and some became poor, even though aggregate poverty rates have tended to fall over time. As a redistributive policy, the effect of the social security system in China today is an important topic for scholars and policymakers.

Social security systems generally include a social insurance system, a social assistance system, and a social welfare system. Social insurance requires the insured persons to pay certain insurance premiums or taxes to be eligible for benefit which is generally not for the purpose of income redistribution. For some social insurance plans, however, benefits received have little to do with contribution and the benefits may not depend completely on the amount paid. Thus social insurance plans also contribute to income redistribution to a certain extent. Funded by public finance, social assistance, and social welfare, on the other hand, are direct income redistribution plans. If the social security system in one country mainly provides social insurance, it has smaller income distribution effects; if social assistance and social welfare, funded by public finance, constitute the major part of the social security system, the system has relatively strong income distribution effects. In addition, the proportion of the population covered by social security and the extent of protection provided by social security also have direct impacts on its income distribution effects. Therefore, the role that the social security system plays in income distribution depends on the composition and specific designs of the system. Its effect on the income distribution of social security is an empirical issue.

Generally speaking, there are two perspectives to test regarding the redistributive role of social security. One is to investigate the effect of certain programs in the system. The other is to estimate the overall effects of income transfers made by all kinds of programs in the system. Annual household or individual data is always used. However, annual data might not reflect the real economic situation of the household or the individual (Rosen and Gayer 2007). When evaluating the redistributive effect of public pension plans in particular, life cycle data are needed to compare the total contribution and benefit. During the working age period, one contributes to the pension system and hence the net benefit is negative in each year in this period of the life cycle. In contrast, one acquires positive net benefits after retirement if based on annual income. Therefore, we have to predict and estimate the contribution and benefit and obtain the net benefit over the life cycle. Nelissen (1998) compared the differences between contribution and benefit using detailed annual income and life cycle income data. …