Measurement Of Income Bipolarization In China

With the rapid growth of the economic aggregate in our country, personal income disequilibrium causes so many contradictions and conflicts. However, some phenomena, like group conflicts and a hollowed out middle class, are can’t explained by poverty or inequality indexes. In this case, western scholars begin to bring the notion of polarization to explain these problems. The most serious polarization is bipolarization. Income bipolarization means a uniform society gradually divides into two groups, the income gap between these groups is large and within groups is small. When the population of these two groups are equal and each member of one group has a zero income while each member of the other group shares the whole income equally, which is considered as complete bipolarization, this kind of polarization is maximum.This article uses CHNS data, chooses proper indicators according to boundary characteristics and depicts the income distribution to compute bipolarization. Empirical analysis indicates that, when boundary is endogenic, income bipolarization grouped by median or average increases from1989to2004, then decreases after2004; whereas the degree of bipolarization grouped by poverty line presents the strong rising trend, which means income antagonism between the poor and the other becomes more serious. When boundary is exogenous, urban-rural bipolarization increases from1989to2004, then decreases after2004; coastland-inland bipolarization increases from1989-2000, then decreases after2000; finally, bipolarization between man and woman presents a rising-falling-rising tendency.On the decomposition of bipolarization, we can realize how the factors changes affect the bipolarization. This article introduces four polarization indexes—Bossert and Schworm(2006), Duclos、Esteban and Ray (2004), Deutsch、Silber and Hanoka (2007) and Wolfson (1994), then analyzes time-decomposition of BS and W measures. The result shows that:Time decomposition of BS measure can be divided into "mean effect" and "inequality effect", while two changes in opposite direction. The contribution of these factors brings the changes in the total bipolarization. Then we decompose W measure into four components to look at the contribution of various factors in bring the changes in bipolarization. Before2004, skewness、antagonism and aggregation from group1(whose units are lower to the median) keep the same direction with the total change in polarization, while aggregation from group2(whose units are higher to the median) effects in opposite direction. After2004, the change effect of skewness、antagonism and aggregation from group1contrary to the total change in polarization, while aggregation from group2keep the same direction with the total changes.Therefore, we need to take a series of steps to reduce the degree of polarization, such as further perfect the income distribution system, extend the proportion of middle-income group, and balance urban and rural economic and social development, improve the income level of the poor group, and reduce the difference in gender.