Spatial Econometric Analysis Of The Influence Factors Of Provincial Consumer Price Index

The consumer price index(CPI) are relative numbers that measure the prices of a set of representative consumer goods and services over time. It can be used to show the changes of price level of consumer goods and services. The index is closely related to people’s daily lives, and it is also an important index for economic analysis,policy-making, general price level detection and national economic accounting. Study on factors affecting the consumer price index and their interaction can help provinces regulate prices and has guiding significance for the provinces to formulate macro policies.Previous study of the consumer price index take the different provincial indexes as independent, ignoring the spatial interactions between the various provinces and cities.Yet regional economic behavior has certain connection on geographical space, which may deviate the established models. This dissertation adopts the spatial econometric approach, introducing interaction of different provinces by spatial weight matrix. It first analyzes the distribution of the average consumer price index from year 2006 to year2013 by the four points map. Through calculation of index of Moran’s I of 30 provinces and cities, it studies the spatial correlation of CPI of different provinces and cities.Results show that the spatial distribution of CPI of different provinces is not random but related. The correlation lies in that the provinces and cities with higher CPI tend to be adjacent with each other, the same is true for provinces and cities with lower CPI. It then use local Moran index, Moran scatter plot to further analyze the correlation. It conducts multicollinearity test on urban residents ’disposable income, money supply,fixed asset investment, agricultural production price index, industrial production price index. Results show that there is no multicollinearity between the five indicators. So they can all be included in the model as factors influencing CPI. Then, we construct empirical analysis through the general panel model, the spatial lag model, the spatial error model and the the geographical weighted regression model.The empirical results show that:(1) Spatial lag model model is more suitable for the study of our country’s consumer price index and its influencing factors, and the geographic weighted regression model is better to reveal the heterogeneity of consumer price index in the province of China.(2) Urban residents’ disposable income, money supply, fixed asset investment, agricultural products price index and industrial production price index all passed 5% significant levels test.The effect of urban residents’ disposable income, agricultural production price index and industrial productprice index is positive, while money supply and fixed asset investment is negative.(3)From the local point of view, the impact of the consumer price index of different province has a big difference. The impact of different factors on the consumer price index varies with the provinces.Finally, based on the above theoretical and empirical research, it puts forward the recommendation that the government should stabilize prices and take precautions against inflation.