Explaining central intervention in local extra-budgetary practices in China

Asian Survey

Volumn 51, Issue 3, 2011, Pages 497-519

  Zhan, Jing Vivian ^a  

^a CHINESE UNIVERSITY OF HONG KONG   (Hong Kong)

Author keywords

Central local relations; China; Data analysis; Extra budgetary finance; Principal agent problem

Indexed keywords

CENTRAL-LOCAL RELATIONS; LOCAL GOVERNMENT; STATISTICAL ANALYSIS;

CHINA;

EID: 79959333759     PISSN: 00044687     EISSN: 1533838X     Source Type: Journal    
DOI: 10.1525/as.2011.51.3.497     Document Type: Article

References (53)

1. Roderick Kiewiet and Mathew McCubbins, The Logic of Delegation: Congressional Parties and the Appropriations Process (Chicago: University of Chicago Press, 1991).
2. Unless otherwise indicated, all the fiscal statistics in this paper are derived from the database I constructed based on Ministry of Finance, ed., Zhongguo Caizheng Nianjian [Finance year book of China], 1992-2007 (Beijing: China Finance Journals Press, 1992-2008)
3. and idem, Difang Caizheng Tongji Ziliao [Local fiscal statistics], 1994-2006 (Beijing: China Finance and Economy Press, 1995-2007).
4. Fan Liming and Li Qiyun, Zhongguo Difang Caizheng Yunxing Fenxi [Analysis of the operation of local finance in China] (Beijing: Economic Science Press, 2001)
5. Zhou Feizhou and Zhao Yang, "Pouxi Nongcun Gonggong Caizheng: Xiangzhen Caizheng de Kunjing he Chengyin-dui Zhongxibu Diqu Xiangzhen Caizheng de Anli Yanjiu" [Analyzing rural public finance: Township fiscal predicaments and causes-case studies of township finance in central and western regions], Zhongguo Nongcun Guancha [Rural China Observations] 4 (August 2003), pp. 25-37.
6. Christine Wong and Deepak Bhattasali, China: National Development and Sub-National Finance: A Review of Provincial Expenditures (Chinese version), trans. Wu Suping and Wang Guijuan (Beijing: CITIC Publishing House, 2003).
7. Chen Zhimei, Zhongguo Xianxiang Caizheng Fengxian Wenti Yanjiu [Research on county and township fiscal risks in China] (Beijing: Renmin Publishing House, 2007)
8. Jean C. Oi and Zhao Shukai, "Fiscal Crisis in China's Township: Causes and Consequences," in Grassroots Political Reform in Contemporary China, eds. Elizabeth Perry and Merle Goldman (Cambridge, Mass.: Harvard University Press, 2007), pp. 75-96
9. Albert Park, Scott Rozelle, Christine Wong, and Changqing Ren, "Distributional Consequences of Reforming Local Public Finance in China," China Quarterly 147 (September 1996), pp. 751-78.
10. EBR is legally recognized and officially reported government revenue. It differs from the illicit revenue collected by government departments and agencies, which some scholars refer to as offbudget revenue. However, the management and documentation of EBR are more chaotic than those of tax revenue. Although there are aggregate statistics and occasional publications of approved EBRs or announcements of abolished ones, there has not been a comprehensive list of all existing EBRs in China.
11. Administrative charges are charges on targeted citizens or legal persons by government departments, agencies, and social groups when conducting public administration or providing certain public services, such as fees for granting professional certificates.
12. Earmarked government funds are charges by governments and relevant government departments on citizens, legal persons, or other organizations in support of certain causes and for specified uses, such as an educational surcharge.
13. Revenues of administrative departments are management fees or other funds collected by administrative departments from their subordinate enterprises, agencies, and social organizations.
14. These are funds raised by township governments and their agencies from township and village enterprises, land users, and individuals, etc.
15. These comprise other fiscal incomes of administrative departments and agencies not included in budgetary management, such as donations to governments and profit return or interest income of extra-budgetary funds.
16. Author's interviews in Hubei Province in 2004 and Guizhou Province in 2009; Zhou and Zhao, "Pouxi Nongcun Gonggong Caizheng," pp. 25-37.
17. Wang Shaoguang, "The Rise of the Regions: Fiscal Reform and the Decline of Central State Capacity in China," in The Waning of the Communist State: Economic Origins of Political Decline in China and Hungary, Andrew G. Walder, ed. (Berkeley: University of California Press, 1995), pp. 87-113.
18. Carsten Herrmann-Pillath and Feng Xingyuan, "Competitive Governments, Fiscal Arrangements, and the Provision of Local Public Infrastructure in China," China Information 18:3 (November 2004), pp. 373-428
19. Wu Licai and Linda Chelan Li, "Xiangzhen Caizhen jiqi Gaige Chutan: Hongzhen Diaocha" [Township finance and its reform: A study on Hong Town), Zhongguo Nongcun Guancha 4 (2003), pp. 13-24.
20. For example, see Chen Bo, Zhongguo Fan Fubai Ershi Nian [Twenty years of anti-corruption in China] (Beijing: People's Publishing House, 2000).
21. Thomas Bernstein and Xiaobo Lu, Taxation without Representation in Contemporary Rural China (Cambridge: Cambridge University Press, 2003)
22. Andrew Wedeman, "Budgets, Extra-budgets, and Small Treasuries: Illegal Monies and Local Autonomy in China," Journal of Contemporary China 9:25 (2000), pp. 489-511.
23. See detailed discussion in Garrett Hardin, "The Tragedy of the Commons," Science 162:3859 (December 13, 1968), pp. 1243-48.
24. Daniel Berkowitz and Wei Li, "Tax Rights in Transition Economies: A Tragedy of the Commons?" Journal of Public Economics 76:3 (June 2000), pp. 369-97.
25. For details, see Jing Vivian Zhan, "Undermining State Capacity: Vertical and Horizontaln Diffusions of Fiscal Power in China," Asian Politics & Policy 1:3 (July/September 2009), pp. 390-408.
26. Jean C. Oi, "Fiscal Reform and the Economic Foundations of Local State Corporatism in China," World Politics 45 (October 1992), pp. 99-126
27. Hehui Jin, Yingyi Qian, and Barry Weingast, "Regional Decentralization and Fiscal Incentives: Federalism, Chinese Style," Journal of Public Economics 89:9-10 (September 2005), pp. 1719-42. (Pubitemid 41093633)
28. Shaoguang Wang, "For National Unity: The Political Logic of Fiscal Transfer in China," in Nationalism, Democracy, and National Integration in China, eds. Leong H. Liew and Shaoguang Wang (London: Taylor and Francis, 2004), pp. 221-46.
29. Moreover, if a province ceases collecting certain EBR for fear of central punishment, it should be regarded as the result of central intervention, or the anticipation of it.
30. R. S. Eckaus, "Some Consequences of Fiscal Reliance on Extrabudgetary Revenues in China," China Economic Review 14:1 (2003), pp. 72-88. (Pubitemid 36649593)
31. The "tax-for-fee" reform aimed to replace some extra-budgetary charges with taxes. Anhui Province was selected in 2000 as the pilot area (see State Council, "Guanyu Jinyibu Zuohao Nongcun Shuifei Gaige Shidian Gongzuo de Tongzhi" [Notification on furthering the experiment of taxfor-fee reform], http://www.mof.gov.cn, accessed March 29 2009), and Jiangsu was added in 2001
32. see Feng Xingyuan, Liu Huisun, and Wei Zhimei, "Nongcun Shuifei Gaige Jiqi Peitao Gaige Wenti Yanjiu" [Study of rural tax and fee reform and related reforms], Shuiwu Yanjiu [Taxation Research Journal] 10 [October 2002], pp. 8-14). Then the reform was extended to 20 provinces in 2002
33. (see Zhu Shouyin, Zhang Haiyang, and Yan Hui, "Nongcun shuifei gaige shidian he xiangcun guanli tizhi gaige genzong yanjiu baogao" [Follow-up report on pilot reform of rural taxes and fees and village governance reform], Jingji Yanjiu Cankao [Review of Economic Research] 40 [2003], pp. 2-24).
34. For example, if the central government had not intervened in a year, a province could project a 15% growth rate of its EBR. But due to the central intervention, the province only saw a growth rate of 5%. Therefore, the observed central intervention does not capture this type of intervention.
35. These include 22 provinces, five autonomous regions, and four centrally administered municipalities, and exclude Taiwan, Hong Kong, and Macao.
36. Per capita budgetary gap = (budgetary expenditure - budgetary revenue) / population. Budgetary expenditure and revenue include outgoing and incoming budgetary transfers, respectively.
37. Deficit ratio = (budgetary expenditure - budgetary revenue) / budgetary revenue. Budgetary expenditure and revenue do not include budgetary transfers.
38. As state employees are paid by budgetary funds, not EBR, the changes in EBR are unlikely to affect the number of state employees. Thus, this measurement avoids the endogeneity problem.
39. I use the constructive budgetary expenditure instead of extra-budgetary expenditure to measure local productivity for two reasons: first, there is very limited data on the itemized extra-budgetary expenditure; second, using budgetary expenditure helps avoid the endogeneity problem because the changes in EBR, especially in the following year, should not affect the pattern of budgetary expenditure in the current year.
40. Starting 1998, simple commercial construction was merged into capital construction and no longer existed.
41. Land and sea zone development expenses were established in 1998.
42. Some may argue that economic capacity increases local bargaining power against the center as well, but I would like to distinguish this bargaining power based on economic means from the political bargaining power discussed earlier.
43. Meirong H.,Zhongguo Dang Zheng Jun Gaoji Lingdaoren Cidian [Who's Who: Current Chinese Leaders] Hong Kong: Wen Wei Publishing Co., Ltd. 2003
44. and Li Gucheng, ed., Zhongguo Dalu Lingdaoren Cidian [Dictionary of leaders in Mainland China] (Hong Kong: Ming Pao Enterprise Co., Ltd., 1997). The other members are found on the Internet.
45. For variables that are highly skewed, I transform them by taking a natural log.
46. In MLE Models 1 & 2, per capita EBR is put on a special status, and its coefficient is fixed as 1, which assumes that when locally collected EBR grows it makes central intervention more likely. But in the standard probit and panel models, it is treated in the same way as the other independent variables, and hence its coefficient is estimated and reported, which serves as a validation of my model.
47. It is worth noting that because of the way the intervention probability function is formulated, the coefficients of the independent variables in MLE Models 1 & 2 should be interpreted negatively, i.e., a positive sign means negative effect on the dependent variable, and vice versa. By contrast, the interpretation of the coefficients in the probit and panel models is straightforward: a positive sign means the variable increases the probability of central intervention and vice versa.
48. It positively influences the likelihood of central intervention except in Model Panel 2 that includes fixed year effect. This change of sign and decrease of statistical significance may be due to the larger number of parameters in the estimation, which increases the variance of the estimates. We may explore the effect of deficit ratio more accurately when more data become available.
49. Victor Shih and Qi Zhang, "Who Receives Subsidies? A Look at the County Level in Two Time Periods," in Paying for Progress in China: Public Finance, Human Welfare, and Changing Patterns of Inequality, eds. Vivienne Shue and Christine Wong (New York: Routledge, 2007), pp. 145-65.
50. The effect of per capita GDP is statistically significant in most of the models and specifications (including those not presented in Table 2). Its effect only becomes statistically insignificant when fixed year factor is included in the estimation, which may be because the larger number of parameters increases the variance of the estimates. But still we see that per capita GDP negatively affects the likelihood of central intervention, although with a lower degree of confidence. The relationship may be more accurately assessed when more data become available.
51. There cannot be a reverse causality that the decrease of EBR increases local GDP because the dependent variable is the decrease of EBR in the following year, which cannot affect the economy in the current year. The same logic applies to other independent variables as well. One may argue that the expectation of lower fees next year can boost the economy now. To deal with this challenge, I used FDI inflow as an instrumental variable for per capita GDP, because EBR drop has no significant effect on FDI inflow while FDI is highly correlated with per capita GDP. The estimation yielded very similar results to the presented models. Therefore, per capita GDP should have a real effect on the dependent variable instead of the other way around.
52. The percentage of rural economy in GDP is statistically significant in most of the models and specifications (including those not presented in Table 2). Its effect only becomes statistically insignificant when fixed region factor is included in the estimation. But still we see that it positively affects the likelihood of central intervention, although with a lower degree of confidence. The relationship may be more accurately assessed when more data become available.
53. Table 2 does not present the coefficients of the region and year factors, but it is worth reporting some interesting findings. In Panel 1 that includes fixed region effect, Fujian, Jiangsu, Zhejiang, and Jiangxi appear statistically significant, and they suffer less central intervention than other provinces. The center may have given favorable considerations to the three coastal provinces and Jiangxi, the traditional revolutionary base of the Chinese Communist Party. By contrast, the inland provinces of Guizhou, Inner Mongolia, Ningxia, Qinghai, Yunnan, and particularly Tibet are subject to significantly more central intervention, which suggests that the central government may have paid extra attention to curbing the extra-budgetary exaction in these underdeveloped and largely rural western provinces. But in Panel 2 when the year factors are also included, only Jiangxi and Zhejiang remain statistically significant, both suffering less central intervention, but all other provinces are statistically insignificant. On the other hand, none of the year factors appear statistically significant. Being in a particular year does not seem to affect the central likelihood of intervention.