Building Bridges or Walls? Understanding Urban-Rural Price Convergence in China

Abstract

Market liberalization reforms in China have led to greater domestic market integration. However, significant urban-rural price differences suggest that segregation between urban and rural markets persists in China. We study the evolution of price differences between urban and rural areas across 25 Chinese provinces over the period 1985-2020. Firstly, we record a substantial and persistent gap between urban and rural price levels within each province. Secondly, we find that this gap narrowed after China implemented the Price Law, which is suggestive of deeper urban-rural integration associated with policy changes. Additionally, we observe notable differences in urban-rural price gaps between provinces, however, using the log-t test, we find evidence of regional convergence in these price gaps. Finally, we investigate the impact of transport infrastructure development on urban-rural price differences. Using spatial econometric analysis and instrumental regressions, we find that road and railway construction has a significant negative effect on the urban-rural price divide, suggesting that infrastructure development facilitates market integration between urban and rural areas in China.

Appendices

Appendix

Urban-Rural Price Differences

In this appendix, we provide details of how the Eq. (6) presented in the paper to capture urban-rural price difference was derived. The living expenditure basket consists of a set of products (goods and services), and their weights calculated based on the quantities purchased in the base year. For simplicity, we can group all goods included in the basket into two generalized categories - traded products and non-traded products. So the price of the basket in urban and rural area, $P^U_{it}$ and $P^R_{it}$, respectively, could be presented as the cost of traded goods and their share as well as the cost of non-traded goods with their share in living expenditures

$$\begin{aligned} P^U_{it}=(1-\alpha ^U_{i})P^U_{TRit}+\alpha ^U_{i}P^U_{NTRit}, \end{aligned}$$

(G1)

$$\begin{aligned} P^R_{it}=(1-\alpha ^R_{i})P^R_{TRit}+\alpha ^R_{i}P^R_{NTRit}, \end{aligned}$$

(G2)

where $\alpha ^U_{i}$ and $\alpha ^R_{i}$ are the shares of non-traded goods in living expenditures in urban and rural areas, respectively, in province i; $P^U_{TRit}$ ($P^R_{TRit}$) and $P^U_{NTRit}$ ($P^R_{NTRit}$) are expenditures on traded and non-traded goods in urban (rural) area, respectively. Using Eqs. (G1) and (G2) and assuming that the structure of living expenditures does not vary across provinces, i.e. $\alpha ^U_{i}=\alpha ^U$ and $\alpha ^R_{i}=\alpha ^R$, we can present the urban-rural price differences as

$$\begin{aligned} \frac{P^U_{it}}{P^R_{it}}=\frac{(1-\alpha ^U)}{(1-\alpha ^R)}\frac{P^U_{TRit}}{P^R_{TRit}}+\frac{\alpha ^U}{\alpha ^R}\frac{P^U_{NTRit}}{P^R_{NTRit}}, \end{aligned}$$

(G3)

where $\frac{(1-\alpha ^U)}{(1-\alpha ^R)}$ and $\frac{\alpha ^U}{\alpha ^R}$ could be understood as relative shares of traded and nontraded goods in the basket, respectively. Since in our data, $\alpha ^U$ and $\alpha ^R$ do not change over time and across provinces, we can denote relative shares of traded and non-traded goods as $\alpha$ and $\gamma$, respectively. Then, urban-rural price differences ($q_{it}=P^U_{it}/P^R_{it}$) can be presented as

$$\begin{aligned} q_{it}=\alpha p_{TRit}+ \gamma p_{NTRit}, \end{aligned}$$

(G4)

where $p_{TRit}= P^U_{TRit}/P^R_{TRit}$ and $p_{NTRit}=P^U_{NTRit}/P^R_{NTRit}$ are relative prices of traded and non-traded goods in province i at time t, respectively.

Table 8 Correlation Coefficients

Table 9 Determinants of Price Differences: Local Railway