社会网络、收入稳定性与农户储蓄率:基于工具变量面板分位数回归模型的研究

Econometric analyses of social network, income stability and saving rate of household based on the instrumental quantiles regression model of panel data

本文基于2003-2011农户微观面板数据,利用固定效应工具变量分位数回归模型,分析了农户的社会网络与收入稳定性对其储蓄行为的影响。研究发现:农户社会网络、收入稳定性、家庭人口规模、未成年人口性别比例、保险参与、家庭财富对农户的储蓄率有显著的影响;随着农户储蓄率分位点的上升,社会网络对储蓄行为的影响逐渐减弱;农户储蓄行为表现出显著的U型生命周期特征。

The economic phenomenon of high saving rate and low spending rate has existed in China for a long time. The situation is not conducive to expanding domestic demand and developing economy healthily. The saving rate is seriously high in rural areas because of the duality development status between urban and rural districts in China. Thus, the study on rural households＇ savings, especially on the influence factors of high saving rate makes good sense. There are abundant literatures in influence factors of saving. However, there is very little research on the effect of social network on savings. The social network effects on household economic strategy are very important. Yi Xingjian（2012） studies the effect of social networks on rural household savings behavior, but there are some shortcomings in the research. The endogenous problem of social networks is not taken into account in the paper. In addition, the effect of income stability is not considered. Thus, it is concluded that there is an inverted u-shaped life cycle curve showing in household savings behavior which does not accord with Chinese actual situation. Based on perusing extensively referenced literature, this paper has some important contributions. First, the endogenous of the social network is taken into account. In contrast, a proper social network IV is chosen and its effectiveness is tested in this paper. Second, income stability of the rural household and its endogenous are considered in the analysis of the saving behavior. An income stability variable is deduced from the household income equation. Third, the regression model controls the variable of minors＇ sex ratio based on China customs. Finally, the quantile regression model based on panel data can eliminate household unobservable heterogeneity and analyze the influence factors for savings quantiles. The paper is organized as follows. We begin in Section 0 by summarizing the introduction. In Section 1, we deduce the saving rate equation from the consumption utility maximization assumption. In Section 2, the data and its main characteristics are presented, also we describe the econometric regression model in this section. Econometric analysis and results are discussed in detail in Section 3. The results reveals that social network, income stability, the sex ratio of the minor, insurance participation, and family wealth have statistically significant impact on saving rate. As the quantiles of the saving rate rise, the impact of social network on saving rate gradually decreased. For lower saving rate household, the income stability has a significant negative effect on its savings rate. The U-pattern life cycle characteristic is shown in saving behavior of the lower saving household. Furthermore, the paper extends the study scope and depth for the effect of social network on household saving strategy at the last part of Section 3. The results show that social network has a stronger influence on saving rate of high-income and low-income level household than on that of middle-income level household. Social network has a greater impact on saving strategy of lower wealth household. The main results are concluded and the policy suggestion is given in Section4.