Well-Being in Transition: Life Satisfaction in Urban China from 2002 to 2012

Abstract

The improving strength of the labor market is chiefly responsible for the overall increase in life satisfaction in urban China from 2002 to 2012. This is especially true for the segment of the population most vulnerable to the negative effects of the on-going transition to a free-market based economy—people with less than a college education. Income comparison and habituation effects offset any positive effect of increased personal income during this time. The result is that increases in income are not significantly related to the increase in life satisfaction during this time. In the interest of protecting the life satisfaction of those most vulnerable, attention must be paid to maintaining a strong labor market as internal migration restrictions are loosened and the labor market is further liberalized in China. These findings are based on repeated cross-sectional data spanning from surveys used in the annual economic reports published by the Chinese Academy of Social Sciences. A modified version of the Oaxaca decomposition method is developed to take advantage of annual data and also control for adaptation to income effects. The change in life satisfaction from 2002 to 2012 is then divided into segments associated with changes in various life domains.

Appendices

Appendix 1

See Table 6.

Table 6 OLS regression results of life satisfaction equation, by level of education

Appendix 2: Equivalence of Oaxaca Decomposition and This Paper’s Approach

The specification of the pooled cross-sectional regression used in the paper is

$$\widehat{LS}_{i,t} = \varvec{x}_{i,t} \widehat{\varvec{\beta}} = \varvec{z}_{i,t} \widehat{\varvec{\gamma}} + \sum_{t = 2001}^{2012} d_{t} \widehat{\delta }_{t} ,$$

(4)

where $d_{t}$ is a dummy for year t.

To implement Oaxaca decomposition, we run regressions for the years of 2002 and 2012, respectively, based on the same specification as in Eq. (4). Then, we have

$$\begin{aligned} \widehat{LS}_{i,2002} & = \varvec{z}_{i,2002} \widehat{\varvec{\gamma}}_{2002} + \widehat{\alpha }_{2002} , \\ \widehat{LS}_{i,2012} & = \varvec{z}_{i,2012} \widehat{\varvec{\gamma}}_{2012} + \widehat{\alpha }_{2012} , \\ \end{aligned}$$

(5)

where $\widehat{\alpha }$ is the estimate of constant.

Then, the Oaxaca decomposition can be expressed as

$$\begin{aligned} \Delta \overline{LS} & = \overline{LS}_{2012} - \overline{LS}_{2002} = \bar{\varvec{z}}_{2012} \widehat{\varvec{\gamma}}_{2012} - \bar{\varvec{z}}_{2002} \widehat{\varvec{\gamma}}_{2002} + \left( {\widehat{\alpha }_{2012} - \widehat{\alpha }_{2002} } \right) \\ & = \left( {\bar{\varvec{z}}_{2012} - \bar{\varvec{z}}_{2002} } \right)\widehat{\varvec{\gamma}}_{2012} + \bar{\varvec{z}}_{2002} \left( {\widehat{\varvec{\gamma}}_{2012} - \widehat{\varvec{\gamma}}_{2002} } \right) + \left( {\widehat{\alpha }_{2012} - \widehat{\alpha }_{2002} } \right) \\ & = \left( {\bar{\varvec{z}}_{2012} - \bar{\varvec{z}}_{2002} } \right)\widehat{\varvec{\gamma}}_{2002} + \bar{\varvec{z}}_{2012} \left( {\widehat{\varvec{\gamma}}_{2012} - \widehat{\varvec{\gamma}}_{2002} } \right) + \left( {\widehat{\alpha }_{2012} - \widehat{\alpha }_{2002} } \right) \\ & = \left( {\bar{\varvec{z}}_{2012} - \bar{\varvec{z}}_{2002} } \right)\widehat{\varvec{\gamma}} + \bar{\varvec{z}}_{2002} \left( {\widehat{\varvec{\gamma}} - \widehat{\varvec{\gamma}}_{2002} } \right) + \bar{\varvec{z}}_{2012} \left( {\widehat{\varvec{\gamma}}_{2012} - \widehat{\varvec{\gamma}}} \right) + \left( {\widehat{\alpha }_{2012} - \widehat{\alpha }_{2002} } \right) \\ \end{aligned}$$

(6)

where the second and third lines of Eq. (6) represent the Oaxaca decomposition with different base years, and the last line is an improved version of Oaxaca decomposition which avoids the issue of double base years. We assume $\widehat{\varvec{\gamma}}$ is obtained from Eq. (4).

According to Eq. (4),

$$\Delta \overline{LS} = \overline{LS}_{2012} - \overline{LS}_{2002} = \left( {\bar{\varvec{z}}_{2012} - \bar{\varvec{z}}_{2002} } \right)\widehat{\varvec{\gamma}} + \left( {\widehat{\delta }_{2012} - \widehat{\delta }_{2002} } \right),$$

(7)

By comparing Eqs. (6) and (7), we can find that, the contribution of survey year dummies in (7), $\hat{\delta }_{2012} - \hat{\delta }_{2002}$, is equivalent to the contribution of the regression coefficients in (6), and the rest parts, the contribution of the change in the values of variables are identical between (6) and (7).

The shortcoming of the approach in the paper compared to Oaxaca decomposition is that the former cannot distinguish the contribution of the change in each regression coefficient. However, the former could be acceptable if the total contribution of regression coefficients is relatively small, which is the case in the paper, or the study cares more about the contribution of the change in variable values.

Appendix 3: Calculation of p Values Reported in Tables 3 and 4

The p value of the percent contribution of variable $x_{k}$ is derived from the following test.

• H0 percent contribution of $x_{k}$ = 0.

• H1 percent contribution of $x_{k}$ is not 0.

As the contribution of $x_{k}$ is calculated from $c_{k} = \frac{{\left( {\bar{x}_{k,2012} - \bar{x}_{k,2002} } \right)\hat{\beta }_{k} }}{{\Delta \overline{LS} }} \times 100\%$, it is equivalent to test if $\beta_{k}$ is 0, conditional on the changes in life satisfaction ($\Delta \overline{LS}$) and in explanatory variables ($\bar{x}_{k,2012} - \bar{x}_{k,2002}$), or treating the two as constant.

If the percent contribution is calculated from the summation of many $c_{k}$ (e.g., the percent contribution of age is equal to the summation of contributions of each age dummy), i.e., $\sum c_{k} = \frac{{\sum \left( {\bar{x}_{k,2012} - \bar{x}_{k,2002} } \right)\hat{\beta }_{k} }}{{\Delta \overline{LS} }} \times 100\%$, then the p value is from the test whether $\sum \left( {\bar{x}_{k,2012} - \bar{x}_{k,2002} } \right)\beta_{k}$ is 0, assuming $\Delta \overline{LS}$ and each $\left( {\bar{x}_{k,2012} - \bar{x}_{k,2002} } \right)$ are constant (or conditional on them).