Empirical Analysis on Urban Retail Business Spatial Distribution Influence Factors ZHOU Chunhua, CHEN Jiandong School of Economics, University of Jinan, Shandong, 250022 Abstract: The development of urban retail business is the most active factor to promote the development of the city. Because of the backward of China urban retail business spatial distribution and lack of theory research and methods exploration, the layoff and plan of retail business is comparatively difficult. This paper classifies 35 capital cities (vice-provincial cities) according to the development degree of urban retail business and find out the key factors that affect the retail business spatial distribution by using factor analysis, regression analysis and cluster analysis. Hope to provide some useful advice for the whole development of urban retail business spatial distribution. Keywords: retail business, spatial distribution, multivariate statistical 1 Introduction Retail business is one of the fastest growing and the highest degree of market opening industry in China. As the important space of retail business, city has the close relationship with it. Scholars abroad start to study the retail business spatial distribution earlier. In 1920s R.E.Park, L.Wirth and E.W.Burgess from Chicago University USA, did lots of research of urban residential area, industrial area and central business area's formation and change and created concentric ring model. In 1960s, with the development of economy and technology, economists study the spatial distribution of retail business from a more rational point of view. D.L.Huff proposed his retail gravitation model based on probability in 1963.At the end of the 20th century, Badcock developed the commercial location theory by further studying the central place theory and other traditional business location theories. The studies of domestic scholars mainly focus on the scale and level distribution of retail network. Professor Ning Yuemin (1984) first created a set of indicators to define the business center and analyzed the factors that influence Shanghai business center and give some practical suggestions. Xue ling and Yang Kaizhong made the quantitative estimation of population potential and business attractiveness in Haidian District to study the urban business activity spatial framework by using the spatial interaction theory and model. lai Zhibin(2009)analyzed population, economy and market competition, the three main factors that affect the location of retail business network and proposed a retail business network location model based GIS and illustrated its practice procedures. Based on the studies before, this paper lists more factors that could influence the spatial distribution of urban retail business and give a much more deeply analysis by collecting more data, using factor analysis, regression model and cluster analysis. 2 Selecting Variables and Data Resources In this paper, 15 factors that may influence the development of urban retail business are chosen. They are : x(geographical 1 area of the city construction,unit:square kilometers), x(urban 2 population, unit:million), x(population 3 density,unit:people / km), x(disposable 4 income of residents per capita,unit:yuan), x(consumption 5 expenditure of urban residents per capita,unit:yuan), x6 (urban GDP per capita,unit:yuan), x(urban 7 fixed asset investment,unit:million), x(urbanization, 8 unit:%), x(commercial 9 business premises selling prices,unit:yuan / square meter), x(the 10 number of wholesale and retail enterprises limit above (legal representativenumber),unit:number), x11 292
(year-end area of urban road,unit: square meters), x(urban 12 road area per capita,unit: square meters), x(number 13 of public vehicles,unit: vehicles), x(year-end 14 number of taxi,unit: vehicles), x(annual 15 public transportation (electric) car passenger volume,unit: million).data of this paper is form City Statistical Yearbook2009,Urban Life and Price Yearbook of China 2009,China City Statistical Yearbook 2009,Statistical Yearbook of China Real Estate 2009. 3 Multivariate Statistical Analysis on Factors Affecting Urban Retail Business 3.1 Factor Analysis According to the data that influence urban retail business test each variables with KMO and Bartlett's Test of Sphericity. The test results are as follows: KMO and Bartlett's Test of Sphericity Kaiser-Meyer-Olkin Measure of Sampling Adequacy..722 Bartlett's Test of Sphericity Approx. Chi-Square 523.277 df 105 Sig..000 According to the result above, KMO is 0.722, P is close to 0, so it is suitable to conduct factor analysis. Use SPSS to analyze, the results are as follows: Factor loading matrix Component Factor score coefficient matrix Component 1 2 3 4 1 2 3 4 x1 0.847-0.394 0.069-0.11 x1 0.046 0.152-0.026-0.176 x2 0.735-0.419-0.146 0.072 x2 0.146 0.014-0.174-0.053 x3 0.092 0.58 0.231 0.615 x3 0.017-0.024-0.257 0.739 x4 0.816-0.144 0.486-0.066 x4-0.11 0.326 0.032-0.008 x5 0.789-0.205 0.499-0.114 x5-0.125 0.343 0.047-0.064 x6 0.635-0.035 0.681-0.007 x6-0.193 0.39 0.024 0.103 x7 0.884-0.279 0.029 0.074 x7 0.094 0.097-0.124 0.013 x8 0.578 0.662 0.128 0.073 x8 0.013 0.026 0.172 0.253 x9 0.425 0.547-0.129-0.578 x9-0.012-0.012 0.601-0.411 x10 0.904 0.051-0.272-0.023 x10 0.188-0.073 0.045-0.045 x11 0.756 0.367-0.104 0.153 x11 0.136-0.058 0.018 0.217 x12 0.009 0.821 0.186-0.214 x12-0.115 0.037 0.423 0.034 x13 0.905 0.147-0.323 0.075 x13 0.223-0.123 0.003 0.059 x14 0.881-0.105-0.335 0.093 x14 0.231-0.106-0.092 0.012 x15 0.864 0.173-0.383 0.031 x15 0.234-0.152 0.041 0.015 F = 0.046x + 0.146x + 0.017x 0.110x 0.125x 0.193x + 0.094x + 0.013x 0.012x 1 1 2 3 4 5 6 7 8 9 + 0.188x + 0.136x 0.115x + 0.223x + 0.231x + 0.234x 293
F = 0.152x + 0.014 x 0.024x + 0.326x + 0.343x + 0.390x + 0.097 x + 0.026x -0.012x 2 1 2 3 4 5 6 7 8 9-0.073x 0.058x + 0.037x -0.123x 0.106 x -0.152x F = 0.026x 0.174x 0.257x + 0.032x + 0.047x + 0.024x 0.124x + 0.172x + 0.601x 3 1 2 3 4 5 6 7 8 9 + 0.045x + 0.018x + 0.423x + 0.003x 0.092x + 0.041x F = 0.176x 0.053x + 0.739x 0.008x 0.064x + 0.103x + 0.013x + 0.253x 0.411x 4 1 2 3 4 5 6 7 8 9 0.045x + 0.217 x + 0.034x + 0.059x + 0.012x + 0.015x According to the extracted factors, considering retail sales of the corresponding factor on the dependent variable of the situation, Four factors are used as independent variables in the regression analysis. 3.2 Regression Analysis of Factors According to the analysis of SPSS, F1, F2, F3, F4 get their scores respectively. Use F1, F2, F3, F4 as the independent variable, y as the dependent variable to construct the regression equation. Use Eview to do regression analysis on main components. F3 is insignificant so can be deleted and reconstruct the regression equation. The result is: Variable Coefficient Std. Error t-statistic Prob. C 29250482 2790304. 10.48290 0.0000 F1 34709270 2838005. 12.23016 0.0000 F2 14160610 2838004. 4.989637 0.0000 F4 5786043. 2838005. 2.038771 0.0518 According to the second regression result, F4is insignificant at the significant level 0.05, but the difference is very small, it can be considered significant and can pass Parameter estimation by significant test. form the overall results from the model fitting effect, modified R 2 reaches 0.858, which shows that the whole fitting effect is very good. Serial correlation LM test and Heteroscedasticity White test. The result shows that Serial correlation and Heteroscedasticity don't exist. LM test results F-statistic 0.954618 Probability 0.399098 Obs*R-squared 2.210682 Probability 0.331098 White test results F-statistic 1.823232 Probability 0.126218 Obs*R-squared 13.52060 Probability 0.140430 After a principal component regression, the final regression equation is Y = 29250482+ 34709270F + 14160610F + 5786043F 1 2 4 Put F1, F2, F4into the above equation, The corresponding impact of the variable x on Y can be got. It can be simplified as follows: Y=24109014+33261349.77 x 1 +50767837.74 x 2-1432237.695 x 3 +47820405.16 x 4 +49143382.86 x 5 +3 9332316.73 x 6 +45121298.63 x 7 +40019967.16 x 8 +34345280.46 x 9 +43954376.92 x 10 +51062615.04 x 11 +24685549.14 x 12 +49376340.48 x 13 +40341597.29 x 14 +47519914.87 x 15 294
From the result we can see that urban population x 2,disposable income of urban residents per capita x 4, consumption expenditure of urban residents per capita x 5, urban fixed asset investment x 7, year-end area of urban road x 11,number of public vehicles x 13,annual public transportation (electric) car passenger volume x 15, have much larger effect on Y than other factors. 3.3 Cluster Analysis Conduct Q cluster analysis on the relative indicators of retail business of 35 capital cities( vice-provincial cities) ( The data of Lasa is missing in the statistic year book) From the Q cluster analysis above, 35 capital cities ( vice-provincial cities) can be classified as follows: The first: Beijing, Shanghai; The second: Guangzhou, Shenzhen; The third: Chongqing; The forth: Tianjin, Nanjing, Wuhan, Chengdu, Shenyang, Xian, Jinan, Qingdao, Dalian, Hangzhou, Fuzhou, Xiamen, Ningbo. According to Pedigree Chart. It can be classified in detail: Wuhan, Chengdu, Shenyang, Xian; Jinan, Qingdao, Dalian; Fuzhou, Xiamen; four cities left. The fifth: the rest 17 cities. From the classification of Pedigree Chart, it is very clear that on the condition that given the number of wholesale and retail enterprises above city limits and the whole sales of them, disposable income of urban residents per capita,consumption expenditure of urban residents per capita, urban fixed asset investment,year-end area of urban road,number of public vehicles,annual public transportation (electric) car passenger volume are all comparatively lower than expected, which is the main restrictive factors of urban retail business. 295
4 Conclusion The factors that can influence urban retail business spatial distribution are very complicated and some cannot be measured. Besides all the factors above, the local government and planning department have large influence on it too. With the development of the diversity trend of residents consumption mode, psychological factors are the field that needs more attention to discuss. From the factors we study above, we can improve rational planning and distribution, accelerate urban economy development and enlarge the circulation infrastructure construction to promote the development of urban retail business. Author in brief: Zhou Chunhua(1977-)female, lecturer, mainly engaged in regional economy. Address: 106 Jiwei road school of economics of University of Jinan, Jinan, Shandong Post code: 250022 Telephone: 18660158256 E-mail: se_zhouch@ujn.edu.cn References [1]. Yang Ying. the process of western commercial space theory since 1920s. tropical geography, 2000, (20):11 (in Chinese) [2]. Chen Xiangping. the development of urban commercial location theory abroad. Shanghai University of Technology Journal, 2003, 25, (1): 14 (in Chinese) [3]. Ning Yuemin. Study on Shanghai Central Location. Geography Journal, 1984, (2): 163-171 (in Chinese) [4]. Yang Wuyang. The past, present and future of Beijing retail business network. Geography Journal, 1994, (1): 35-36 (in Chinese) [5]. Xue Ling, Yang Kaizhong. Business Distribution Based on Space Interaction Model. Geography Study, 2005, (2): 35-37 (in Chinese) [6]. Lai Zhibin. Retail Business Location Model Based on GIS. Geography Information World, 2009, (2): 22-26 (in Chinese) [7]. GeJintian. Aggregation and differentiationof Urban Retail Business. Dongyue Forum, 2009, (6): 34-41 (in Chinese) 296