3 Empirical Analysis
3.1 Model and empirical results
Holtz-Eakin et al. (1988) extended the classic vector autoregressive model (VAR) to panel vector autoregressive model (PVAR), making it more effectively to analyze the internal relation between panel data more effectively. A classic PVAR model can be expressed as follows:
Where yi,t is M×1 variable vector; Гj is M×M the coefficient matrix of the different lag variable; αi is M×1 vector represents the fixed effect; γt is M×1 vector represents time effect; μi,t is random error term, obey the standard Gaussian process. We estimate the urban and rural models respectively, to compare with the results of their analysis, the model is as follows:
In the urban model, yi,t=[URit, UCit, UEit]′, in the rural model, yi,t=[URit, RCit, REit]′. For more details about PVAR model and its estimation method see Holtz-Eakin et al.(1988) and Arellano and Bover (1995).
We estimate respectively two major categories of urban and rural residents consumption with two lag levels. The purpose of this study is to explore the impact of development of urbanization on consumption; We list the estimate results of equations which take the total consumption level and consumer Engel coefficient as endogenous variabls.
As can be seen from Table 2, the majority of the variables were significant at the 10% level, indicating that the equation estimation effect is better. These results are more intuitive, but it does not reflect the dynamic effects between PVAR system variables, so we will further analyze the impulse response function, more accurately assess the dynamic trend of urbanization’s impact on urban and rural consumption.
Table 2 Empirical Results
3.2 Impulse response analysis
As can be seen from Figure 3, when urbanization changes by a unit standard deviation, the level of urban residents’ consumption shows steady upward trend. The effect factor remains stationary between 0.05-0.1, which indicates that urbanization can promote the upgrading of urban residents consumption level effectively by promoting urban economic development, accumulating factors of production, enhancing income level and other ways. Similarly, from Figure 4, when urbanization changes by a unit standard deviation, the level of rural residents’ consumption has a positive impact. This effect shows weakening trend and its factor is from 0 to 0.04. After 5th period, the impact of urbanization on consumption has almost completely disappeared. From the above results, the development of China’s urbanization, promoted the growth of urban and rural consumption, but the boost in not its leading role in urban consumption is relatively stationary, while the leading role in the rural consumption is gradually reduced.
From the dynamic effects of urbanization on urban consumption structure (Figure 5), a unit of urbanization standard deviation can continue to reduce the Engel coefficient of urban residents’ consumption, which improves the consumption structure of urban residents, and its impact factor is from -0.005 to 0. The same result can be seen from Figure 6, the impact of urbanization also effectively reduces the Engel coefficient of rural areas, which improve the consumption structure of rural residents, and the impact factor remains to about -0.004. From the results above we can know that, while improving the development of China’s urbanization, the consumption structure of the urban and rural areas is improved and the intensity to the two regions are basically the same, which means that there is no “distortions” effect of urbanization on improvement of urban and rural consumption structure.
Figure 3 The Dynamic Effect of Urbanization on the Urban Consumption
Figure 4 The Dynamic Effect of Urbanization on the Rural Consumption
Figure 5 The Dynamic Effect of Urbanization on Urban Consumption Structure
Figure 6 The Dynamic Effect of Urbanization on Rural Consumption Structure