218 WHY LAW ENFORCEMENT IS WEAK IN CHINA: THE MINDSET OF THE FRONTLINE TAX OFFICIALS Huina Xiao This article examines the decision-making process of the frontline tax bureaus in China. It suggests that the pressures of the revenue target, political and social networks, and a flawed tax legal system have shaped the frontline tax bureaucrats’ calculations on costs and benefits. Drawing on the cost-benefit analysis, this study finds four types of decisions in these situations: symbolic enforcement, creative enforcement, weak enforcement, and negotiated enforcement. The study suggests that to discipline frontline tax officials is never easy since most are constrained in a vicious circle. Only when the institutional environment is changed can the outcome of law enforcement be substantially enhanced. The typology is shown to be more applicable and comprehensive to understand the problems of law enforcement in developed and developing countries. INTRODUCTION .......................................................................................................................219 STUDYING EVERYDAY PRACTICES IN THE CHINESE FRONTLINE TAX OFFICES ........................221 THE INSTITUTIONAL ARRANGEMENTS OF THE FRONTLINE TAX OFFICES IN CHINA ................226 Fiscal Relations between the Center and the Localities ....................................................227 Tax Bureaus Embedded in the Political and Social Network ............................................229 Too Many Discretions: Tax Legislations, Procedures and Administrative Litigations ......232 Evaluating and Disciplining the Frontline Tax Officials ...................................................235 THE APPROACH OF COST-BENEFIT ANALYSIS .........................................................................239 The Costs of Enforcement ..................................................................................................239 The Benefits of Enforcement ..............................................................................................240  Postdoctoral Fellow, Centre for Chinese & Comparative Law, School of Law, City University of Hong Kong. Email: huinaxiao2@cityu.edu.hk. This article is partly based on the author’s unpublished doctoral dissertation. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 219 FOUR TYPES OF ENFORCEMENT PRACTICES ............................................................................242 Symbolic Enforcement ........................................................................................................242 Creative Enforcement .........................................................................................................243 Weak Enforcement ..............................................................................................................250 Negotiated Enforcement .....................................................................................................252 THE PATTERN AS A TOOL TO ORGANIZE PREVIOUS STUDIES ...................................................256 CONCLUSIONS .........................................................................................................................258 APPENDIX ...............................................................................................................................260 INTRODUCTION China has long been labeled a weak and lax enforcement country due to its limited organizational capacity, 1 corruption and local protectionism. 2 To curb the discretion of enforcement agencies and overcome implementation issues, the central government launched various political legal campaigns to enhance enforcement (also called ad hoc enforcement,3 selective enforcement,4 or strike hard5), known as “a pattern of long periods of weak and lax enforcement, followed by short periods of strong and coercive enforcement.”6 1 See Sean Cooney, Making Chinese Labor Law Work: The Prospects for Regulatory Innovation in the People’s Republic of China, 30 FORDHAM INT’L L.J. 1050, 1050-51 (2006). 2 See ANDREW WEDEMAN, FROM MAO TO MARKET: RENT SEEKING, LOCAL PROTECTIONISM, AND MARKETIZATION IN CHINA 2-3 (2003); MELANIE MANION, CORRUPTION BY DESIGN, BUILDING CLEAN GOVERNMENT IN MAINLAND CHINA AND HONG KONG 85 (2004); BENJAMIN VAN ROOIJ, REGULATING LAND AND POLLUTION IN CHINA: LAWMAKING, COMPLIANCE, AND ENFORCEMENT; THEORY AND CASES 264, 271, 283, 288 (2006). 3 BENJAMIN VAN ROOIJ, REGULATING LAND AND POLLUTION IN CHINA: LAWMAKING, COMPLIANCE, AND ENFORCEMENT; THEORY AND CASES 352 (2006). 4 Zhiyong Dai (戴治勇), Xuanzexing Zhifa (选择性执法) [Selective Enforcement of Laws], 4 FAXUE YANJIU (法学研究) [CHINESE JOURNAL OF LAW] 28, 28 (2008) 5 HAROLD TANNER, STRIKE HARD: ANTI-CRIME CAMPAIGNS AND CHINESE CRIMINAL JUSTICE 1979–1985 (1999). 6 Benjamin Van Rooij, The Campaign Enforcement Style: Chinese Practice in Context and Comparison, in COMPARATIVE LAW AND REGULATION: UNDERSTANDING THE GLOBAL REGULATORY PROCESS 217, 217 (Francesca Bignami & David Zaring ed., 2016). See also, Xin He, Regulating Rural-Urban Migrants in Beijing: Institutional Conflict and Ineffective Campaigns, 39 STAN. J. INT’L L. 177, 177-78 (2003); Nicole Ning Liu, Carlos Wing-Hung Lo, Xueyong Zhan, and Wei Wang, Campaign-Style Enforcement and Regulatory Compliance, 75 PUBL. ADMIN. R. 85, 85 (2014). 220 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 Another strand of studies has found a mixed enforcement style, which is based on bureaucratic targets (mandates, quotas, top-down personnel assessment, and cadre evaluation) system.7 That is, local agencies both “work” very hard and “shirk” very ably. O’Brien and Li defined this as “selective policy implementation” that implements unpopular policies (birth control, revenue collection, funeral reform) and sets popular policies (fee limits, village election, corruption regulation) aside.8 Cai found that under the pressures and incentives of the higher level government, local agents tended to engage in statistical inflation.9 Birney defined it as the “rule of mandates” distinguished from the “rule of law”, wherein higher priority mandates regarding economic growth and revenue collection undermine lower priority mandates like combatting corruption and enforcing election law. 10 Instead of examining the relationships between different types of target, recent studies have focused on a single type of target. For instance, in his insightful study, Wang showed the role of cadre evaluation in implementing the targets of environment protection, as required by the Eleventh and Twelfth Five-Year Plans. Though the top-down target approach achieved some positive outcomes, several problems like goal specification, goal displacement, information asymmetry and collusion have undermined environmental implementation.11 Kostka further examined the impacts of a targets system on law environmental policy implementation and 7 See Alex Wang, The Search for Sustainable Legitimacy: Environmental Law and Bureaucracy in China 37 HARV. ENVTL. L. REV. 365, 365 (2013); Genia Kostka, Command without Control: The Case of China’s Environmental Target System, 10 REGUL. & GOV. 58 (2016). 8 Kevin O’Brien & Lianjiang Li, Selective Policy Implementation in Rural China, 31 COMP. POLIT. 167, 168 (1999). 9 Yongshun Cai, Between State and Peasants: Local Cadres and Statistical Reporting in Rural China, 163 THE CHINA Q.,783, 803 (2000). 10 Mayling Birney, Decentralization and Veiled Corruption under China’s Rule of Mandates, 53 WORLD DEV. 55, 55 (2014). 11 Wang, supra note 7, at 374. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 221 found four outcomes based on two variables: intended or unintended outcome, and desirable or undesirable outcome.12 To what extent are these findings true in tax authorities? What are the factors or motivations that have shaped the behavior patterns of tax officials? Drawing on a cost-benefit analysis, this article examines how the frontline tax authorities make decisions and in what circumstances a particular decision is preferred. The remainder of this article begins with the descriptions of the tax authorities’ structure in China and data collection methods of the study. It then presents the institutional environment of the frontline tax authorities. The article continues with definitions of cost and benefit; then the four different reactions and outcomes based on the cost and benefit analysis will be presented. It will end with some discussions. STUDYING EVERYDAY PRACTICES IN THE CHINESE FRONTLINE TAX OFFICES At the end of 1993, China set up two administrative tax systems: state and local taxation offices (国家税务局和地方税务局) (Figure 1). Besides the central taxes, the state bureau also administered shared taxes such as VAT (Value-added Tax) and part of corporate income taxes. The local bureau managed local taxes, individual income tax and part of corporate income taxes.13 There were four levels in the state taxation structure: the state administration of taxation (SAT), the provincial office, the municipal office (including directly governed 12 Kostka, supra note 7, at 62. 13 Taxpayers who pay VAT are required to report corporate income tax to state tax bureaus, and taxpayers who pay business tax are required to report corporate income tax to local tax bureaus. See State Administration Of Taxation (hereafter SAT), Guanyu Tiaozheng Qiye Suodeshui Zhengguan Fanwei Wenti de Tongzhi (关于调 整新增企业所得税征管范围问题的通知) [Notice on Adjusting the Collection Scope of Corporate Income Tax Law], available at http://www.chinatax.gov.cn/n810341/n810765/n812171/n812675/c1190618/content.html 222 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 city and autonomous regions) and the county office.14 The county-level taxation office was permitted to build sub-offices or stations.15 The administrative dimensions, with respect to institution, cadre, fund and leader’s position, were vertically managed. To illustrate, the upper state bureaus appointed and approved the president and vice president of the lower state tax bureaus. Local administration of taxation was established based on the administrative hierarchy of China. There were three levels: the provincial local taxation bureau (including directly governed city regions and autonomous regions), the municipal local taxation bureau and the county local taxation bureau. Their administrative systems, institutional arrangements and recruitment were expected to follow the regulation of local governments. Both local governments and the SAT led local administration of taxation, but the former played the major role. The SAT leadership focused on the instruction and coordination of tax policy and business, supervision of the tax system and policy, and communication of good practice. Therefore, the relationship between SAT and local administration of taxation seemed more cooperative than supervisory. One small example makes the point: the candidate for the president of the provincial local administration of taxation was approved and appointed by local government after seeking opinions from the SAT. 14 Id. 15 Local taxation bureaus at the county level can also build sub-offices or stations. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 223 Figure 1. Organizational Chart of China’s Tax Administration16 The data of this study was complemented by interviews with 34 tax officials in a state county-level tax bureau (Bureau P) and a local county-level tax bureau (Bureau R) in County J, southeast China. In 2015, the county’s gross domestic product per capita reached 25.7 billion RMB, compared with 162 billion in the strongest county of the province (ranked No. 5 in the country)17 and 4.2 billion in the poorest county of the province.18 It covers an area of 2,100 square kilometers, with a population of 0.86 million and 16 towns.19 The county is noted for its chemical, agriculture, and tourism industries. In Bureau P, there are eight departments, including departments of personnel and education, supervision, financial management, tax policy, administration of tax collection, tax payment services, information management, income accounting, an audit bureau, a general 16 See SAT, Organizational Chart of China’s Tax Administration, http://www.chinatax.gov.cn/eng/n2367721/c2390734/content.html. 17 China Top 100 Counties Yearbook 2015. 18 Provincial Statistical Yearbooks 2016. 19 The County Government Annual Working Report 2016. 224 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 County-level Tax Bureau Department of Audit* Department of Policy and Regulation* Department of Supervision Department of Administration of Tax Collection* Department of Information Management Department of Taxpayer Services* Department of Financial Affairs Department of Administrative Managment* Tax Branches* office, and six tax branches (Figure 2). The six branches administer different types of taxpayers. Regarding the sectors, the first branch takes charge of larger businesses (including the financial and insurance sector); the second branch collects taxes from export enterprises and overseas-funded enterprises; the third branch manages the taxes from the stone materials industry and gas stations; and the fourth, fifth, and sixth branches deal with smaller-sized enterprises that pay less than 1 million RMB annually. With respect to the size of the businesses, the first and second branches administer large businesses (which pay more than 1 million per year), while the remaining two branches manage all small and medium sized enterprises and individually owned businesses. In total, 102 staff members were working for Bureau P in 2016. In order to make paying tax easier, the sixth and seventh branches are located in different places to Bureau P. Figure 2. The Organizations of the County-Level Tax Bureau in China 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 225 Notes: The departments marked with an asterisk (*) were interviewed; the three dotted boxes were the departments held by Bureau R. Bureau R belongs to the local taxation system. It was established in July 2012, to collect taxes from a new harbor development zone (海港開發區). Bureau R’s administrative level is the same as Bureau P’s. Bureau R directly reports to the local city-level tax bureau. Bureau R has three major departments, including a general office, a tax service hall, and tax administration office, with a total of 16 tax officials. Table 1. The Profile of Informants in County J, City Z (2015-2016) Gender Age Service years CPC member Home origin Degree Bureau Male 29 ≤30 7 ≤5 7 Yes 32 Local 32 High school 0 State 21 Female 5 31-40 10 6-10 6 No 2 Non- local 2 College 6 Local 13 41-50 9 11-15 1 Bachelor’s 25 >51 8 16-20 3 Master’s 3 >21 17 Total 34 34 34 34 34 34 34 The author was officially approved to examine Bureau P and Bureau R. Living in the dormitories of the two tax offices for one month enabled me to conduct detailed interviews with and acquire comprehensive data from tax officials to local corporate taxpayers, and from frontline tax branches to city-level offices. My method was to develop a rapport that allowed 226 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 in-depth interviews to be conducted with those who had knowledge of the tax administration. My questions focused on what their major concerns in their everyday work were. I explored the resources they controlled, the internal and external restrictions they had, and how they solved problems. Table 1 shows the number of interviews conducted in Bureau P and Bureau R. Some interviews were conducted in their offices. In a couple of instances, I asked the interviewees questions when they were dealing with particular cases outside the tax bureaus or when we were having a meal or tea together. I also collected specific cases and statistical information that could lend insight into the logic of their behavior. None of the interviews were taped. Instead, I took detailed notes immediately after conducting each interview. THE INSTITUTIONAL ARRANGEMENTS OF THE FRONTLINE TAX OFFICES IN CHINA The central and local governments motivate tax authorities to maximize revenues by establishing a binding target system consisting of the tournament system and the cadre responsibility system (cadre evaluation). 20 The semi-decentralized fiscal system has empowered local governments to develop their economy and use partial revenues without constraints; tax bureaus zealously pursue revenue maximization and partially enforce law due to the political and social embeddedness; inconsistent laws, ambivalent procedures and abused discretions enhance the target system. 20 See, e.g., Kevin O’Brien & Lianjiang Li, Selective Policy Implementation in Rural China, 31 COMP. POLIT. 167, 172 (1999); Alex Wang, The Search for Sustainable Legitimacy: Environmental Law and Bureaucracy in China, 37 HARV. ENVTL. L. REV. 365, 417 (2013); Graeme Smith, Measurement, Promotions and Patterns of Behavior in Chinese Local Government, in RURAL POLITICS IN CONTEMPORARY CHINA 111, 117 (Kevin O’Brien, Emily Yeh, and Jingzhong Ye ed., 2016). 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 227 Fiscal Relations between the Center and the Localities After China initiated its socialist transformation modeled on the Soviet Union in 1953, many enterprises became nationalized or collectivized, and there was little resistance to tax collection. From 1950 to the Great Leap Forward, the profits contributed by state enterprises jumped from 13 per cent to 62 per cent of annual revenues.21 During the Cultural Revolution (1966–1976), local tax authorities were cancelled or merged, tax officials were demoted, and the taxation system was demolished. After the 1978 plenum,22 China embraced a “reform and opening up” policy and embarked on a nationwide market transformation. Meanwhile, the center experimented with numerous revenue-sharing arrangements to grapple with fiscal relations between central and local governments. After 1983, the contract system, known as “eating from separate pots” and in which provinces and the center shared total tax revenues, came to dominate. 23 The new system encouraged local governments to develop a local economy, but also led to fiscal deficits for the central government.24 In order to improve fiscal capacity, in 1994 the center came up with the division of taxes system in which “the central and provincial governments [had] their own respective fixed 21 ELIZABETH REMICK, BUILDING LOCAL STATES: CHINA DURING THE REPUBLICAN AND POST-MAO ERAS 140 (2004). 22 The Third Plenum of the Eleventh Central Committee of the Chinese Communist Party was the decisive meeting held in Beijing, China, from 18–22 December 1978. See Zhongguo Gongchandang Dishiyijie Zhongyang Weiyuanhui Disanci Quanti Huiyi Gongbao (中国共产党第十一届中央委员会第三次全体会议公 报) [Report on the Third Plenum of the Eleventh Central Committee of the Chinese Communist Party], ZHONGGUO GONGCHANDANG LICI QUANGUO RENMIN DAIBIAO DAHUI SHUJUKU (中国共产党历次全国代表大 会数据库) [CENTRAL COMMITTEE OF THE CHINESE COMMUNIST PARTY DATABASE] (Dec. 22, 1978), http://cpc.people.com.cn/GB/64162/64168/64563/65371/4441902.html. 23 Remick, supra note 21, at 142. 24 See Jean Oi, Fiscal Reform and the Economic Foundations of Local State Corporatism in China, 45 WORLD POLIT. 99, 123 (1992); SHAOGUANG WANG & ANGANG HU (王绍光 & 胡鞍钢), ZHONGHUO GUOJIA NENGLI BAOGAO (中国国家能力报告) [REPORT OF CHINESE STATE CAPACITY] 26 (1993); Shaoguang Wang, China’s 1994 Fiscal Reform: An Initial Assessment, ASIAN SURV. 801, 802 (1997). 228 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 expenditures and revenue sources” and “shared receipts from several locally-collected taxes”.25 Thus, central tax, local tax and shared tax are three major types of tax. VAT, the pillar of the tax system, is shared by the center (75 per cent) and localities with (25 per cent). Consumption tax belongs to the center and business tax belongs to the localities. The center shares 60 per cent of corporate income tax, while localities share 40 per cent. The SAT administers the central tax and the shared tax, while the local administration of taxation administers the local tax. Finally, to encourage local governments to develop their economy, the central government has also initiated a transfer payment system. On the one hand, the center returns a certain amount of tax to localities based on growth regarding the revenue of VAT and consumption tax. When VAT and consumption tax increase by one per cent, the refund increases by 0.3 per cent.26 On the other hand, the center subsidizes the localities based on local needs and their efforts in increasing tax revenues.27 Under the system of division of taxes, local governments are motivated to boost their economy by introducing investments and businesses so that they can collect sufficient tax revenues, tax refunds and subsidies to afford enormous expenditures. Not only should local governments ensure fiscal balance, but they also need to achieve certain revenue growth corresponding to GDP growth every year. The mechanisms of the intergovernmental tournament and the cadre responsibility enable the higher-level governments to obtain 25 Michel Oksenberg & James Tong, The Evolution of Central-Provincial Fiscal Relations in China, 1971– 1984 The Formal System, 125 THE CHINA Q. 1, 5 (1991). 26 The State Council, Guowuyuan Guanyu Shixing Fenshuizhi Caizheng Guangli Tizhi de Jueding (国务院关 于实行分税制财政管理的决定) [The Decision on Implementing the Financial Manageable System of Division of Taxes], available at http://cpc.people.com.cn/GB/64184/64186/66685/4494212.html. 27 Feizhou Zhou (周飞舟), Fenshuizhi Shinian: Zhidu ji qi Yingxiang (分税制十年：制度及其影响) [A Decade of Division of Taxes: The System and Its Evolution], 6 ZHONGGUO SHEHUI KEXUE (中国社会科学) [SOCIAL SCIENCES IN CHINA] 100, 102 (2006). 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 229 information, give rewards and impose sanctions on the head of the lower level governments.28 Tax Bureaus Embedded in the Political and Social Network Establishing an autonomous or semi-autonomous tax authority with an ability to implement tax laws and policies amid resistance among society is a common practice around the world.29 In 1988, China removed the General Taxation Bureau from the Ministry of Finance (MOF) and started the SAT to replace its functions. The SAT is a quasi-Ministry whose president reports directly to the State Council, but acting under the guidance of the MOF.30 Though state tax bureaus at local levels are directly governed by the SAT, there are several constraints from local governments. The amount of tax revenues is the first concern of local governments. In order to ensure fiscal balance and improve performance evaluations, local governments also allocated revenue targets to tax authorities. Despite the county level state tax bureau being governed by the SAT and independent from the local government, they were constrained by the targets that were required by the local government. Take County J as 28 Katharina Pistor & Chenggang Xu, Governing Stock Markets in Transition Economics: Lessons from China, 7 AMERICAN LAW AND ECONOMICS REVIEW 184, 197-200, 203-206 (2005). 29 Ole Therkildsen, Autonomous Tax Administration in Sub-Saharan Africa: The Case of the Uganda Revenue Authority, 31 FORUM FOR DEVELOPMENT STUDIES 59, 59 (2004). 30 See WORLD BANK, CHINA: REVENUE MOBILIZATION AND TAX POLICY 117 (1990) (The major responsibilities of the SAT include implementation of the taxation objectives of the State Council, administration of national taxes, tax policy, the drafting of national tax laws, the drafting and issue of administrative regulations, the collection of (national) tax revenues, the use of taxation as a macroeconomic level, the management of tax official within provincial governments, training and coordination within the tax system, negotiation with foreign government and representation of China in international bodies, and administration of taxation on offshore oil reserves). 230 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 one example: in 2015, the municipal state tax office assigned County J state tax bureau a target of collecting 10.86 billion RMB, while the county government assigned 11.29 billion. Local governments rewarded or punished based on the performance of the tax bureaus to secure reliable implementation from the tax authorities. The chief of the fifth tax branch of Bureau P (governed by the SAT) (S1) explained to me: The local government will give a bonus if we meet or exceed the target. They also decide to reward role models in tax collection or moral and cultural progress (文明单位). They also control party affairs and labor unions. The better the county-level state tax bureaus fulfil the target, the happier the head of the local government, and the more they win bonuses and rewards. Every leader favors quantified achievements. If the county level state tax bureaus fail to meet the target, both formal and informal sanctions are a source of concern for the heads. For example, they might be demoted from their position. The local government and municipal state tax office will scold the leaders of the county level state tax bureau. They should submit a report to explain why they have failed to meet targets. They lose face. The higher level officials will get tough on them by initiating more audits and investigations. As Cai put it:31 In China’s case, since the tenure and promotion of lower level officials are decided by their higher level counterparts, they are held more accountable to higher level officials than to the public and thus are more likely to suffer pressure from above. Tax bureaus also face social constraints in a relational (social connections or social ties, guanxi or renqing in Chinese) community. In his classic From the Soil, Fei suggested that:32 31 Yongshun Cai, Between State and Peasants: Local Cadres and Statistical Reporting in Rural China, 163 THE CHINA Q. 783, 784 (2000). 32 XIAOTONG FEI, GARY HAMILTON, AND WANG ZHENG, FROM THE SOIL: THE FOUNDATIONS OF CHINESE SOCIETY 62 (1992). 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 231 Our pattern is not like distinct bundles of straws. Rather, it is like the circles that appear on the surface of a lake when a rock is thrown into it. Everyone stands at the center of the circles produced by his or her own social influence. There are 102 tax officials in Bureau P: 90 of them (88 per cent) are from County J, and the rest are from neighboring counties or cities. In Bureau R, 14 of 16 (87.5 per cent) are from County J and two are from other counties of the city. The old saying “everyone has a cousin three thousand miles away” implies the vastness of kinship networks in Chinese society.33 In the county, people establish relationships not only through blood relations, but also through friends, classmates, comrades-in-arms and townsmen.34 Moreover, the internal network of bureaucrats enhances the embeddedness. Feng studied the political network in Zhong County of Henan Province. The managers contended that “youth is precious, capability is for reference, guanxi is the most important (年龄是个宝，能力做参考，关系 最重要)”.35 They built up the network through drinking, playing cards and gift-giving.36 Social ties also played a critical role in the internal and external circumstances of tax authorities. Building relationships with various governmental agencies and tax officials is imperative for firm survival.37 Relational businesses are capable of using these political and social connections to avoid audits or sanctions or to receive help to comply creatively with the tax law. 33 Id. at 63. 34 Junqi Feng (冯军旗), Zhongxian Ganbu (中县干部) [Zhong County’s Cadres], 161 (2010). 35 Id. at 171. 36 Id. at 164. 37 Shibin Sheng et al, The Effects of Business and Political Ties on Firm Performance: Evidence from China, 75 J. MARKETING 1, 1 (2011). 232 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 Too Many Discretions: Tax Legislations, Procedures and Administrative Litigations Street-level bureaucrats normally have wide discretion in deciding the nature, amount and quality of sanctions,38 but that is not to say that they are unrestrained by rules, regulations and directives. The dilemma is that simply eliminating discretion makes the job more difficult due to complicated situations, responses to unique circumstances and the need to promote clients’ self-regard.39 The solution could be a “correct balance between compassion and flexibility on one hand, and impartiality and rigid rule application on the other”.40 The problem with discretion in Chinese grassroots tax bureaus is that there can be too much compassion and flexibility, which often leads to discretion abuse due to inadequate tax laws and limited procedures. China, as a transitional country, is still on the road towards a consistent and complete tax law system. Most taxes are regulated by administrative regulations made by the State Council instead of laws 41 legislated by the National People’s Congress (the highest legislative institution in China) in 1980, 1995, and 2007, respectively,42 the legal effects of which are stronger than administrative and local regulations.43 It is reasonable to make laws based on long-term experimentation since legalization is costly. However, illegalization makes the law 38 MICHAEL LIPSKY, STREET-LEVEL BUREAUCRACY: DILEMMAS OF THE INDIVIDUAL IN PUBLIC SERVICES 13- 14 (2010). 39 Id. at 15. 40 Id.at 15-16. 41 Except for the Corporate Income Tax Law, Individual Income Tax Law and Law of PRC on the Administration of Tax Collection. 42 Individual Tax Law was amended in 2005, 2007 and 2011; Law of PRC on the Administration of Tax Collection was amended in 2001, 2013 and 2015. 43 Lifa fa (立法法) [Legislation Law] (promulgated by Standing Comm. Nat’l People’s Cong., March 15, 2000, rev'd March 15, 2015), art. 88, available at http://www.gov.cn/zhengce/2015-03/18/content_2834713.htm. Administrative regulations are made by the State Council; local regulations are issued by the local People’s Congress; rules are formulated by local governments (province, municipality and county) 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 233 enforcer (the State Council or the SAT) become the rule makers, which leave opportunities for them to protect their interests. For instance, the SAT made a rule called the strict liability of buyers, meaning that whether or not the buyer knows an invoice is illegally issued, they cannot deduct input taxes based on those invoices. This rule gave rise to several disputes among tax authorities because the SAT rule had no legal basis. In 2008, the State Council amended Article 9 of the Interim Regulation of PRC on VAT by introducing the SAT’s regulations as a legal basis for verifying illegal VAT input invoices.44 Moreover, the procedures of law enforcement offer enormous discretion to the local tax authorities. There were three major organizational reforms within tax authorities. In the 1980s, grassroots tax cadres, also called special managers (專管員), enjoyed the authorities of tax collection, administration, and investigation. The nanny tax bureaus visited taxpayers and collected taxes one by one.45 But there was little resistance because most taxes were collected from collective and state-owned enterprises.46 With the privatization and the rapid growth of private enterprises in the 1990s, tax authorities started to separate the powers of collection, administration and investigation—that is, the specific department of tax collection was set up for collecting taxes only. The tax branches took charge of routine administration with respect to tax consultants, business registrations, calling in arrears, monitoring, tax assessment and 44 See Wei Cui (崔威), Zhongguo Shuiwu Susong Shizheng Yanjiu (中国税务行政诉讼实证研究) [The Empirical Study on Tax Administrative Litigations in China] 9 QINGHUA FAXUE (清华法学) [TSINGHUA U. L. J.] 135 (2015). 45 SAT (国家税务总局), 2002–2006 Nian Zhongguo Shuishou Zhengshou Guanli Zhanlue Guihua Gangyao (2002 年–2006 年中国税收征收管理战略规划纲要) [The Outline of Strategic Planning on Tax Collection and Administration from 2002–2006] (Mar. 10, 2003), available at http://www.chinatax.gov.cn/n810341/n810765/n812198/n813086/c1207101/content.html. 46 Remick, supra note 19, at 140. 234 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 inspection; the audit bureaus investigated cases of violation of tax laws.47 In 1997, a further reform “centralizing collection, emphasizing checks, intensifying administration on the basis of tax returns and optimized services and by depending on computer networks, and correctly dealing with the relationships among collection, administration and checking” was initiated to further enhance the separation of the powers.48 Three reforms improved procedural justice and efficiency but ambivalent procedures that often led to discretion abuse remained. Cui suggested that the existing tax legislation and regulations have empowered tax authorities with too much discretion.49 Article 63 of the Law of PRC on the Administration of Tax Collection provides that if a taxpayer evades taxes, the tax authority shall pursue the unpaid or underpaid taxes, the late fee, and a fine of an amount from half to five times that of the taxes unpaid or underpaid. Article 37 of the Provisional Regulations of National Invoice Management prescribes that when someone illegally issues an invoice, the tax authority shall confiscate the illegal gains. If the amount is less than 10,000 RMB, the tax authority may fine no more than 50,000; if the amount is more than 10,000, the fine shall be more than 50,000 and less than 500,000. Articles 38 and 39 also provide a broad scale of fines for illegal invoice behaviors. The 2012 SAT guidance suggested that provincial tax authorities (autonomous regions and municipalities) should formulate the discretion benchmark for regulating administrative discretion based on the local 47 SAT (国家税务总局), Guanyu Jingyibu Jiaqiang Shuishou Zhengguan Jichu Gongzuo Ruogan Wenti De Yijian (关于进一步加强税收征管基础工作若干问题的意见) [Concerning Some Issues on Further Strengthening the Basic Work of Tax Administration] (Oct. 22, 2003), available at http://www.chinatax.gov.cn/n810341/n810765/n812198/n813046/c1204889/content.html. 48 Id. 49 Cui, supra note 42. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 235 circumstances of law enforcement.50 However, this caused other problems. The provincial discretion benchmarks included illegal type, illegal behavior, legal basis, severity and sanctions, but failed to consider taxpayers’ intentions. The benchmarks varied between provinces, bringing regional unfairness.51 The tax officials were hesitant to apply the criteria because they might incur an enforcement risk when taxpayers complain. If taxpayers can exhaust legal remedies for injustice, the discretion of tax authorities can be monitored and regulated. Cui’s empirical study showed that administration taxation litigation reached 2,000 RMB annually at the end of the 1990s, fell to 1,000 in 2004, and has remained at 400 in recent years. Three factors are attributed to the low litigation rate for tax disputes in China. The cost of litigation is high: taxpayers should request administrative reconsideration first and pay all their arrears before going to court. Moreover, the tax legislation is inconsistent and incomplete, so that grassroots tax offices enjoy too much discretion. Finally, most taxpayers know their cases are not strong enough because they have failed to pay taxes honestly.52 Evaluating and Disciplining the Frontline Tax Officials Tax authorities exercised an annual top to bottom performance appraisal. The upper tax authorities rewarded the lower level tax authorities that ranked highest based on overall 50 SAT (国家税务总局), Guanyu Guifan Shuiwu Xingzheng Cailiangquan Gongzuo De Zhidao Yijian (裁量 权工作的指导意见) [Regulating the Work of Administrative Discretion for Taxation] (Mar. 7, 2012), available athttp://www.gov.cn/gongbao/content/2012/content_2266923. 51 Longlin Qin (秦隆伶), Shuiwu Xingzheng Chufa Ziyou Cailiangquan: Cunzai Wenti Yu GuifanJianyi (税 务行政处罚自由裁量权:存在问题与规范建议) [Discretionary Power of Tax Administration Penalty: Issues and Suggestions for Standardization] GUOJI SHUISHOU (国际税收) [INT’L TAX’N IN CHINA] 58–60 (2014). 52 Cui, supra note 42. 236 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 performance or on a specific indicator (e.g. organizational revenue performance). The indicators varied every year.53 In 2012, there were seven indicators involving twenty-four sub-indicators with a total of 100 points in the province of County J (Table 2). Performance assessment motivated tax authorities to increase revenue by the way of audit, inspection or investigation. For instance, the sub-indicators of investigated arrears read that tax authorities gained two points if the unpaid tax accounted for one per cent of annual revenue and 0.2 points if it was in excess of 0.1 per cent, with a 0.2 point deduction if it was below 0.1 percent. 53 For example, in County J, before 2016, there are 100 points and only seven major indicators; after 2016, there are 1,000 points and more than 20 indicators. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 237 Table 2. The 2015 Indicators of Performance Evaluations in the State Tax Office of County J Indicators Sub-indicators Points Percentage Total Organizational revenue Revenue income 6 6% Accuracy rate of prediction 3 3% Revenue growth and contribution 4 4% Tax assessment 9 9% 22% Tax administration International corporate tax administration 3 3% Anti-tax avoidance 2 2% Export tax refund 5 5% Investigated arrears 4 4% Quality of tax administration 3 3% Corporate income tax administration 3 3% Risk management 2 2% 22% Service of tax payment Establishment of tax service hall 3 3% Establishment of tax website 2 2% Response rate of 12366 hotline 1 1% 6% Regulation of law enforcement Legal administration 3 3% Administrative reconsideration 2 2% Supervision of law enforcement 5 5% 10% Technological tax administration GTP 3 3% Information application 3 3% Model application 3 3% Business creation 2 2% Accuracy rate of data quality 3 3% 14% Development of administrative team Advanced work unit 5 5% Illegal behavior 12 12% Individual assessment 3 3% 20% 238 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 Budget administration The budget progress 3 3% Security management 3 3% 6% Total 100 100% 100% The upper level tax office not only monitors the task fulfillment progress, but also disciplines irresponsible or illegal acts by the lower level tax office. Managing agents, however, is never easy in most societies, especially in authoritarian states where the rule of law is often lacking.54 Cai found that the upper level tax office adopted a “selective or differentiated discipline” based on “the severity of the consequences of the official’s failure, and his or her role in the failure.”55 However, if officials are able to block the flow of information, their malfeasance will also be hard to find.56 Embedded in the political and social network, tax officials are able to cover up their malfeasance. The upper level tax office audited bureau R in August 2015. A work team consisting of three officials from either the municipal tax office or another same level tax office investigated the cases that Bureau R had handled in the previous week. The head of Bureau R had worked in the municipal tax office for seven years and knew all three officials well. The investigation result showed no serious offences but some small violations. These internal investigations were conducted annually. According to the head of Bureau R (R1), none had 54 YONGSHUN CAI, STATE AND AGENTS IN CHINA: DISCIPLINING GOVERNMENT OFFICIALS 181 (2014). 55 Yongshun Cai & Lin Zhu, Disciplining Local Officials in China: The Case of Conflict Management, 70 THE CHINA J. 98, 98 (2013). 56 Id. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 239 been sanctioned before. Without external monitoring and a sound rule of law, simply relying on “the monitoring of the central or provincial authorities is essentially impossible.”57 THE APPROACH OF COST-BENEFIT ANALYSIS Economic theory builds on the assumption that actors are naturally self-regarding and oriented towards maximization of their own interests. It assumes that all parties involved in enforcement and regulation seek to maximize their utility;58 it also assumes that enforcement is costly.59 The bureaucratic study observed that government officials were “purely self- interested” 60 or “primarily self-interested and highly opportunistic in pursuit of that self- interest,”61 and were motivated by goals that benefited themselves rather than society.62 This assumption especially fits with the grassroots officials in China, the largest authoritarian state in the world where a free press and the rule of law are often lacking. A cost-benefit analysis helps us to understand how frontline tax bureaus make decisions on their routine activities. The Costs of Enforcement Tax officials’ consideration of cost includes organizational cost and personal cost. In his influential study, Stigler suggested that the extent of law enforcement relies on the amount of organizational resources devoted to the task63—for instance, professional and technical staff, funds, assessable information, etc. With severer punishments and higher enforcement costs, 57 Cai, supra note 52, at 190. 58 George Stigler, The Theory of Economic Regulation, 2 BELL J. ECON., no. 1, Spring 1971, at 3; BALDWIN ROBERT ET AL., UNDERSTANDING REGULATION: THEORY, STRATEGY, AND PRACTICE 43 (2012). 59 George Stigler, The Optimum Enforcement of Laws, in ESSAYS IN THE ECONOMICS OF CRIME AND PUNISHMENT 55 (1974). 60 ANTHONY DOWNS, INSIDE BUREAUCRACY 88 (1967). 61 STEVEN SOLNICK, STEALING THE STATE: CONTROL AND COLLAPSE IN SOVIET INSTITUTIONS 5 (1998). 62 Cai, supra note 52, at 181. 63 Stigler, supra note 57. 240 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 “the marginal deterrence of heavy punishment could be very small or even negative.”64 Thus, the optimal and rational enforcement is to pursue major offences with heavy penalties instead of minor offences. The optimal tax audit strategy aims to maximize tax and penalty revenues as well as minimize audit costs.65 Not only has the enforcement cost (or organizational cost) swayed the decision-making of frontline tax officials, but personal cost has also played a role in China. As discussed before, tax authorities care more about their career safety and are risk averse. Grassroots tax bureaus have paid more attention to meeting the target tax revenue to avoid unnecessary trouble. They have also dealt carefully with the political and social ties between senior members of local government and taxpayers. The Benefits of Enforcement Rational actors attempt to maximize utilities. The benefits of enforcement include an organizational benefit and a personal benefit. The goal of organizational enforcement is to achieve a certain degree of compliance.66 When the benefits of enforcement are trivial, tax authorities might deviate from enforcement. Career prospects (e.g. promotion), rewards, and social resources are all benefits for frontline tax officials. Tax officials—especially the principals of tax agencies—would like to meet tax revenue targets to show their eligibility, avoid unnecessary troubles and maintain good relations with the local government. Tax officials also benefit from bribery (rent 64 Id.at 57. 65 James Andreoni et al, Tax Compliance, 36 J. ECON. LIT. 818 (1998). 66 Stigler, supra note 56. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 241 seeking) owing to political or social ties. Discipline from the higher-level government or tax offices can reduce the potential benefits of enforcement. It depends on how the tax officials estimate the risks of detection and the severity of sanctions. According to Cai, higher-level governments or departments often tolerate officials’ malfeasance.67 The effects of top-down monitoring and discipline, therefore, might be trivial on the mental accounting of benefits among tax officials. Understanding the logic of decision-making in enforcement has both theoretical and practical significance. The upper left quadrant in Table 3 refers to symbolic enforcement, when both benefits and costs are low. In the bottom left quadrant, weak enforcement happens when the costs are high and the benefits low. The upper right quadrant, creative enforcement, refers to low costs and high benefits, while the bottom right quadrant, negotiated enforcement, refers to high benefits that are costly. Table 3. The Cost and Benefit Decisions of Frontline Tax Officials Benefit Low High Cost Low Symbolic enforcement (Tax promulgation) Creative enforcement (Collecting tax revenues) High Weak enforcement (Auditing small offences) Negotiated enforcement (Investigating the case of larger businesses initiated by an upper-level tax office) 67 Cai & Zhu, supra note 53. 242 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 FOUR TYPES OF ENFORCEMENT PRACTICES Symbolic Enforcement When the costs and benefits were lower, tax officials tended to enforce the task symbolically regarding research papers, tax propaganda, party work, and the rest. The following were some tasks of the fifth tax branch of Bureau P in 2016: Tax promulgation. Gain one point when a propaganda report is adopted by the county tax bureau (maximum four points); deduct one point each when tax branches or departments fail to attend the propaganda project organized by the county tax bureau. Research work. Gain six points when every tax branch submits at least one good research article or report before 10 May and every department and audit bureau submits at least one before 31 October. Gain another four points when the work is chosen as a significant social and economic issue by leaders or published; deduct two points when tax authorities fail to complete the above activities. Study reports of the collection and administration reforms. Based on the arrangements and requirements of the SAT and the provincial tax office, every tax branch, department, and audit bureau submits a study report on the legal system and reform of tax collection and administration, internet-based administration and modernization to the city tax office on time. Party work. In order to promote President Xi Jinping’s talk on family traits, every tax branch should submit an article based on a local family tradition. Tax promulgation aimed to encourage tax agencies to bring the new tax laws and policies to the public to improve their awareness of paying taxes. Publication of reports was in internal tax periodicals. Research work or project studies aimed to motivate tax officials to study the latest tax policies, laws or comparative experiences, as well as the problems found 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 243 in their everyday work. Party work targeted the disciplining of tax cadres and enhancing the party ideology. All the above tasks were easier for tax bureaus to implement and irrelevant to external political and social norms. Implementing these tasks gained credit for performance assessment, but enforcing them made no difference to the result of the performance evaluation because almost every tax branch and department was able to implement them. Hence, their strategy was to enforce these tasks symbolically. To implement tax promulgation, the fifth branch of Bureau P asked two tax officials to put a table at the door with a banner reading, “Everyone is obliged to pay tax,” then delivered a brochure to the taxpayers. Another official took photos during the activity. They wrote an activity report with illustrations. Bureau R came up with a new solution to promote tax policies. They gathered the taxpayers, who just happened to deal with tax business, to their conference room, showed them a video, delivered a tax brochure, and took photos. Regarding writing a research paper or report, their normal strategy was to copy and paste from online materials. Symbolic enforcement was commonly used to implement policies, laws, or organizational assignments that brought limited benefits for frontline tax branches or departments. In order to fulfill the assignments, tax agencies either forged evidence or indifferently enforced them. Creative Enforcement When the costs of a task were low and its benefits were high, tax officials were keen to implement it. They might be motivated to enforce the task creatively to maximize future 244 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 benefits. Take the collection of tax revenues as an example. In the 1980s, collecting revenues was costly because tax cadres would visit taxpayers and collect taxes one by one. The two decades of administrative reforms, including computerization, GTP, the hotline service and the like, had significantly reduced tax authorities’ costs for collecting tax revenues. On the other hand, collecting sufficient tax revenues brought high benefits. It could improve their performance evaluations at the end of the year. Tax bureaus that were able to meet or exceed their tax revenue target had more opportunity to gain promotions and win bonuses. The tax revenue target was set based on local economic growth, prior tax revenue targets and the amount of tax revenue in the last year. The municipal tax office predicted the target by estimating the local development plan with respect to new investments and new projects. To illustrate, the harbor developmental zone of County J invested 7.4 billion RMB to run nineteen new projects regarding the chemical industry, infrastructure, harbor facilities and real estate that could contribute around 50 million to the annual revenue in 2013.68 In 2014, there were 11 new significant projects in which a total of around 11 billion were invested; these were predicted to contribute 150 million to annual revenue.69 With this information, the municipal tax office was able to predict rough tax revenue for the coming year. The tax revenue target for the next fiscal year was often higher than the one for the last year. Bureau R, for example, was assigned tax revenue targets of 200 million,70 230 million, 500 million and 990 million from 2012 to 2015, respectively, by the municipal tax office. The tax revenue 68 The Analysis of Tax Revenue of Bureau R in January 2013. 69 See Economic Report made by The Government of Harbour Developmental Zone of County J in December 2014. 70 It was revised in December 2012. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 245 targets of Bureau P increased by 33.66 per cent on average every year from 2006 to 2012.71 There were two types of target: the municipal tax office and the county government. The municipal tax office emphasized the previous tax revenue more, while the county government focused more on expenditure and the fiscal balance. Therefore, the tax revenue target that was required by the county government tended to be higher. Table 9 and Table 10 show the tax revenue target and tax revenue in Bureau R (county-level local tax bureau) and Bureau P (county-level state tax bureau). There were three possible scenarios when implementing the tax revenue target: meeting the target, being unable to meet the target, and finding the target somewhat difficult to meet. 71 Its tax revenue target increased by 45.3 per cent in 2007, 32.79 per cent in 2008, 38.18 per cent in 2009, 18.42 per cent in 2010, 35.8 per cent in 2011 and 32.45 per cent in 2012. I calculated the revised tax revenue targets. 246 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 Table 4. Tax Revenue Target and Tax Revenue in Bureau P (2006–2015) (Millions, RMB) Year Tax revenue The municipal tax office The county government Original target Task completion ratio (%) (before revision) Revised target Task completion ratio (%) (after revision) Original target Task completion ratio (%) (before revision) Revised target Task completion ratio (%) (after revision) 2006 172.43 142.5 121.00% N/A 121.00% 153.3 112.48% N/A 112.48% 2007 236.79 207.1 114.34% N/A 114.34% 214.52 110.38% N/A 110.38% 2008 311.1 275 113.13% N/A 113.13% 310 100.35% N/A 100.35% 2009 390.1 335 116.45% 380 102.66% 365 106.88% N/A 106.88% 2010 494.87 405 122.19% N/A 122.19% 456.7 108.36% N/A 108.36% 2011 637.32 550 115.88% N/A 115.88% 625.8 101.84% N/A 101.84% 2012 764.04 723 105.68% N/A 105.68% 831 91.94% 750 101.87% 2013 865.07 1020.4 84.78% 860 100.59% 1019.8 84.83% 850 101.77% 2014 889.27 1411 63.02% 883 100.71% 1400 63.52% 865 102.81% 2015 895.68 1086 82.48% 830 107.91% 1129 79.33% 820 109.23% 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 247 Table 5. Tax Revenue Target and Tax Revenue in Bureau R (2012–2015) (Millions, RMB)72 Year Tax revenue The municipal tax office The county government Original target Task completion ratio (%) (before revision) Revised target Task completion ratio (%) (after revision) Original target Task completion ratio (%) (before revision) Revised target Task completion ratio (%) (after revision) 2012 200.00 300.00 66.67% 200 100.00% Missing Missing Missing Missing 2013 370.00 230.00 160.87% N/A 160.87% 300.00 123.33% N/A 123.33% 2014 980.00 500.00 196.00% N/A 196.00% 550.00 178.18% N/A 178.18% 2015 990.00 990.00 100.00% N/A 100.00% 950.00 104.21% N/A 104.21% Fulfilling the Tax Revenue Targets and Controlling the Amount Bureau P was able to meet the tax revenue target required by the municipal tax office and the county government without any revisions from 2006 to 2008 and from 2010 to 2012. Tax authorities monitored the progress of target implementation monthly. In June, they examined the outcome for the half-year. The analysis of tax revenues in June of 2006 and 2010 read:73 In the first half-year, due to the excellent economic performance, we collected 77.64 million RMB. It accounted for 54.48 per cent of the tax revenue target assigned by the municipal tax office. We achieved the goal of “double half” (双过半) (that target was half accomplished when the time passed half). We collected 250 million RMB, 61.14 per cent of a total of 410 million annual target. It 72 Bureau R was established in 2012. 73 Similar statements were written in the analyses of tax revenues for Bureau P from 2007 to 2009. 248 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 was completed 21 days earlier than the plan. The tax revenue target mechanisms, however, brought undesired outcomes. When the county tax bureau and its tax branches met the tax revenue target before December, they tried to control the amount of tax revenue by asking some larger companies to stop issuing invoice, or by paying part of the tax in the coming January. This strategy was used to keep a lower tax revenue target for the next year. For instance, the accountant for a local larger company explained: Tax officials called our financial manager to stop issuing invoice until the end of the December because they had met the tax revenue target. This is very common. Larger companies liked to cooperate with tax agencies. They tried to control the amount to obtain a smaller tax revenue target in the next year. The lower the target, the less the pressure. Unable to Fulfil the Tax Revenue Targets Sometimes, tax agencies were unable to collect sufficient tax revenues as a result of mistaken prediction or unanticipated incidents. Bureau R was allocated a tax revenue target of 300 million RMB at the beginning of 2012. However, it had only collected a total of 210 million in November because some significant projects were not accomplished on schedule. Obviously, it was impossible for Bureau R to collect 90 million during the remainder of the year. The municipal local tax office then changed the target to 0.2 billion so that Bureau R could just meet the target (Table 5). The targets for Bureau P were revised from 2012 to 2015. Figure 3 shows that Bureau P exceeded the target slightly after it was adjusted. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 249 Figure 3. Before and After Revision of Tax Revenue Target in Bureau P (2006–2015) (Millions, RMB) Why did the municipal tax office approve the revision? It was because the county government tried to beautify the annual performance. The leader of the county government mobilized the municipal tax office to reduce the tax revenue target. Informal political and social networks played roles during the negotiation. The municipal tax office would examine whether the target assigned by the provincial tax office could be completed. If it could, they would increase the targets for the developed counties and reduce the targets for the underdeveloped counties. All parties favored this win-win outcome. That pattern was duplicated at different levels of government and tax offices. Somewhat Difficult to Fulfil the Tax Revenue Targets When it was a bit hard to meet the tax revenue target on time, tax bureaus might either seek help from the local larger businesses or audit a company that was suspected of evading 250 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 taxes. To reduce the enforcement cost, tax officials required the taxpayers to self-check. If taxpayers found something wrong with their tax return, they would pay arrears. This was so- called self-correction. If taxpayers failed to check themselves properly, tax officials would investigate on the spot. Taxpayers were afraid of detection and sanctions so that they normally paid a certain amount of taxes extra to the local tax authorities. The solution was effective to ensure the tax authorities to complete the revenue targets. Weak Enforcement When the costs of enforcement were high while its benefits were low, tax agencies had no motivation to enforce it. To illustrate, tax agencies avoided detecting small offences, which were difficult to find. According to Stigler’s optimum enforcement of laws, “The state will pursue more tenaciously the offender who commits a larger crime (or repetitive crimes) and thus increase also the probability of apprehending him.”74 Combatting tax evasions by small and medium sized companies required more manpower, financial resources, information and so forth. It was also risky for tax agents as the law enforcement could break their political and social network, which might bring high personal cost. On the other hand, detecting small or minor offences achieved little compliance. The vice-chief of the fifth tax branch of Bureau P (P2) said: There are six tax branches with approximately forty tax officials in Bureau P. Some of them lack specialized tax knowledge and some of them are too young to have experience. Except for them, there are only twenty experienced auditors. Some tax evasions are very difficult to find. The cash economy is everywhere in China. It is 74 Stigler, supra note 56. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 251 useless to detect those evasions if everyone disobeys the law. You stop one, others continue. The chief of the first tax branch of Bureau P (P3) further explained: The larger the size of the companies, the better compliance they have. Firstly, larger companies should build up a transparent information system, so they should comply with the law voluntarily. Otherwise, they are easy to to be caught… Every company experiences from the youth to the maturity. At the beginning, they are struggling to survive so they are often unable to pay tax; when they become stronger, they care more about their social status and reputation. They also enjoy more political and economic benefits—for example, tax relief, project support and funding. Compared with larger companies, small and medium-sized companies contributed a small amount of revenue to tax bureaus. The major revenue resource for Bureau R was the property industry. Except for Business Tax (43.96 per cent), land use tax, property tax, VAT on land, tax on farmland occupation and deed tax (16.81 per cent) were all related to the real estate industry. The taxes levied from the architecture industry and real estate accounted for 88.22 per cent of the annual tax revenue from 2012 to 2015. Larger businesses contributed 83.27 per cent to the tax revenue. Heavy reliance on the real estate industry in local government was not new in China. Zhou suggested that grassroots governments uphold urbanization because they earn revenues through developing the real estate industry.75 In Bureau P, ten larger companies whose annual amount of tax payment exceeded 10 million RMB paid 38.7 per cent of the annual tax revenue; in addition, around 130 medium and larger companies whose annual amount of tax payment exceeded 1 million provided 75.96 per cent of the annual revenue from 2010 to 2015. In all, Chinese tax revenue was largely 75 Zhou, supra note 25. 252 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 levied from larger businesses and concentrated on one or two significant industries. More supervision and audits on small and medium-sized companies made little difference to the total tax revenues, as did individual career prospects. To avoid non-enforcement, some indicators of performance assessment were added to encourage the frontline tax agents to investigate tax evasions. The greater the audit revenue, the higher the points. In practice, the mechanisms preventing tax officials from weak enforcement were largely ineffective but brought new problems. For instance, tax agencies might audit selectively based on the amount of possible arrears payment, belated payment and fines. Negotiated Enforcement When the costs and benefits of enforcement were high, tax bureaus preferred negotiation to balance the interests of the different parties. The new risk-based audit system might allocate the auditing of large businesses. The national campaign of combatting illegal VAT invoices might also lead the audit to some medium and large businesses. Auditing medium or large businesses brought high costs and benefits. Firstly, the audit process demanded more professional and experienced tax officials as the large businesses had more complicated business styles, markets and organizations. What was more, if large businesses were punished, the bad reputation might undermine the economic performance of the companies and thus bring less income. Since tax payments were significant in relation to revenue collection, less income of large businesses resulted in less revenue of the local tax authorities. Tax officials also felt hesitate to audit and punish large businesses because that would break the social 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 253 network and bring high social cost. In contrast to higher organizational and personal costs, the detection could also bring high benefits to the tax authorities. On the one hand, it could achieve more compliance in local communities if the large businesses were audited and punished. On the other hand, large businesses would mobilize their political resources to avoid audit and punishment. Tax agents were able to accept bribes from wealthy businesses. A win-win strategy—i.e. negotiated enforcement—was thus adopted. X Company, established in 1995, administered by the second tax branch of Bureau P,76 produced quick-frozen products and exported to other countries. Its tax payment reached 9.55 million RMB in 2015. The audit bureau started to investigate X Company’s 2014–2015 tax return in early 2016. It found that X Company evaded 0.01 million of VAT and 0.5 million of corporate income tax. Besides paying arrears, the company was also fined.77 14. According to Article 64, Law of PRC on the Administration of Tax Collection, “When a taxpayer does not file tax returns, or does not pay or underpays the taxes, the tax authority shall pursue the payment of the taxes unpaid or underpaid and the belated fee, and fine with the amount from 50 per cent to five times of the taxes unpaid or underpaid”.78 We suggest a fine of 50 per cent of VAT unpaid. That is 5677.43 = (2717.68*0.5 + 8637.17*0.5). 15. According to Article 32, Law of PRC on the Administration of Tax Collection, “When a taxpayer fails to pay taxes or a withholding agent fails to remit taxes within a prescribed time limit, the tax authority shall, in addition to ordering the taxpayer or withholding agent to pay or remit the taxes within the prescribed time limit, impose a 76 The second tax branch of Bureau P majorly governed export enterprises. 77 Tax Audit Report of the Case of X Company, issued in 2016. 78 The translation is from Fabao database, Peking University, http://en.pkulaw.cn.ezproxy.cityu. edu.hk/display.aspx?cgid=252598&lib=law. 254 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 belated fee at the rate of 0.05 per cent per day of the amount of taxes in arrears, on the day the tax payment is overdue.”79 X Company should pay the belated fee. Based on the above decisions, X Company should pay arrears, the belated fee and fines. The fine would be recorded in the file and the information was public. For an export company, it could damage its reputation, as well as slow the export VAT refund. Thus, X Company sought help from the local government. X Company’s application and the reply from the vice president of the county below show how the audit decision was changed. Application for Pleading with the Tax Audit Case of X Company County Government: X Company is the provincial leading enterprise of the ocean and agricultural industry. The annual amount of export sales reaches 45 million USD. The company paid a total of 11 million RMB annually.80 X Company will soon obtain a well-known trademark of China due to the stable development in recent years. In the special inspection of aquatic products from 2014 to 2015, the audit bureau found illegal records in the accounting books. X Company has regulated its finance and taxation operation under the instruction of the audit bureau. It was required to pay 0.50 million for the arrears of corporate income tax, 0.01 million for the VAT arrears, and 5677.43 for fines. The record of the penalty will lower our credit level of export tax refunds, and have a bad impact on the credit level of finance at the bank. To support local enterprises, X Company earnestly requests county government to put good words in front of the county state tax office. We wish that we can be exempted from the fines. In the margin of the report, the vice president of the county wrote to the county-level state tax office: 79 Id. 80 The amount of tax payment here was more than the amount I collected from County J. It might be because X Company included the amount of export refund. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 255 X Company is one of the top ten export enterprises in the county. If the company is fined, its export will be affected on the one hand, and the county might also fail to meet the revenue target on the other hand. Please support the company. Negotiated enforcement means that the audited taxpayers and tax authorities negotiate the penalties of illicit behaviors. After negotiation, the tax authorities reduced the sanctions. Tax officials or the chief of the tax bureau gave renqing (help) to the leaders of the local government and the owner of the company. They could get renqing back in the future when tax bureaucrats also needed help from them. It is a good way to build up a political network. Negotiated enforcement, however, has potential risks—for example, when the higher tax office or the SAT oversaw the case, tax officials and the chief of the audit bureau would be punished because the negotiated enforcement actually violated the tax law. Thus, the chief of the audit bureau and the president of Bureau P might prefer to submit these risky cases to the higher-level audit office. The chief of the audit bureau in Bureau P (P8) gave an example: Last year, a large company tried hard to persuade us not to be convicted of illegally issuing invoices. I reported it to the president. Considering the enforcement risk, we transferred the case to the city tax office. A tax official in Bureau R (R4) expressed a similar worry: There are too many discretions. Last year, I audited a state-owned enterprise and found that it evaded nearly one million. The county government asked the company to pay the arrears. Then they also negotiated on the amount of belated fee. In fact, we have faced high enforcement risks when the higher tax office checks the quality of law enforcement. 256 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 THE PATTERN AS A TOOL TO ORGANIZE PREVIOUS STUDIES Tax officials made decisions based on their mental calculations of the costs and the benefits. The cost dimension consists of the need for manpower, financial resources and facilities support, as well as the possibility of being removed from position and of losing social capital—i.e. guanxi. The benefit dimension includes the organizational benefits that increase the tax revenues and voluntary compliance of the local community, and the individual benefits of gaining promoted, accepting bribes or gaining social capital. I incorporated organizational, legal and social factors to operationalize the costs and benefits of enforcement. The two dimensions work together to shape the decisions of the frontline tax officials. Compared with Stigler’s optimum enforcement of laws, which only consider the organizational cost,81 this typology will be more applicable and complete to explain the uneven enforcement in an authoritarian state where the rule of law is underdeveloped and state capacity is limited. The typology can also be used to explain previous studies on law enforcement in China. He’s study on migrant businesses in Zhejiang village in Beijing suggested that the state undertook sporadic campaigns to prevent license violations from getting worse, but not to eliminate them; the state benefited.82 The decision was shaped by the conflict of interest between the upper level government, which concerned more on social stability, and the lower level government, which focused on immediate interest.83 Hence, these sporadic campaigns 81 Stigler, supra note 56. 82 Xin He, Why Do They Not Obey the Law? A Case Study of a Rural-Urban Migrant Enclave in China, 199 (J.S.D. Dissertation 2003). 83 Id. at 158. 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 257 fall under creative enforcement in my typology, where the costs of the action were low while the benefits were high. This happened when the lower level officials were required to regulate the business in a short period, which was not costly, and they were able to extract interest from the migrant businesses. However, they soon turned to weak enforcement because they knew that eliminating migrant businesses would reduce their benefits and the enforcement costs would become high. The failure of the sporadic campaign, 84 ad hoc enforcement85 or campaign style enforcement86 can be understood under my typology. Another strand of studies found that targets played a critical role in shaping the decisions of legal officials. Kostka categorized policy outcomes for China abiding by environmental targets based on whether the outcomes were desirable and anticipated.87 In my case, I also argue that the local tax officials concentrated more on fulfilling the tax revenue target, the outcome of which could be anticipated and was desirable for the upper level tax office, while they side-lined other targets, the outcome of which could be anticipated but was undesirable. This caused other unintended and undesirable consequences—for example, engaging in statistical inflation. Of course, the target system also generated some unintended and desirable outcomes, like enhancing local tax compliance. The two typologies thus interacted. My typology, however, provides the logic of how different consequences were generated, while Kostka only provided the outcomes. 84 See Xin He, Sporadic Campaigns as a Means of Social Control, 17 COLUM. J. ASIAN L., 121 (2003). 85 See BENJAMIN VAN ROOIJ, REGULATING LAND AND POLLUTION IN CHINA: LAWMAKING, COMPLIANCE, AND ENFORCEMENT; THEORY AND CASES 352 (Leiden Univ. Press 2006). 86 Nicole Ning Liu et al., Campaign-Style Enforcement and Regulatory Compliance, 75 PUB. ADMIN.REV.85 (2014). 87 Kostka, supra note 5. 258 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 In short, some of the major theories discussed in previous research on enforcement would gain from considering them within my typology and the interaction of the costs and the benefits. While many existing theories narrow their applications so as to focus on a specific aspect of enforcement problems, my typology can be more broadly applied. CONCLUSIONS This article examines the decision-making process of the frontline tax bureaus in China. It suggests that the pressures of tax revenue targets, political networks and a flawed legal system have shaped the frontline tax bureaucrats’ considerations on costs and benefits. Drawing on the cost-benefit analysis, four types of enforcement decisions are found: symbolic, creative, weak, and negotiated. When the costs and benefits were low, tax agencies tended to symbolically enforce the task. Symbolic enforcement, however, wasted organizational resources, led to dishonesty and neglected important work that might not favor grassroots tax authorities. Tax agencies paid more attention to meeting their tax revenue targets, since this incurred fewer costs but brought substantial benefits. When the amount of tax payment was high, tax agencies and local governments might control the amount of revenue by requiring the large businesses to implement this year’s tax liability in the next year. When tax revenue was unable to meet the target, local governments tried to mobilize the municipal tax office to lower the target in order to enhance the performance record. When tax revenue targets were somewhat difficult to implement, tax agencies or local governments might seek help from the larger businesses to pay taxes in advance, or audit taxpayers about whom they were 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 259 suspicious. Focusing on the future performance evaluation, tax authorities did not maximize tax revenue; instead, they controlled the amount of tax revenue. Tax authorities were not motivated to detect tax evasions since the costs were high and benefits low. Even though tax authorities had revised the evaluation indicators of organizational performance, introduced risk-based responsive regulation and launched annual campaign regulations on illegal tax invoices, it remained hard to enhance the motivation for detecting small violations. Tax authorities negotiated penalties with taxpayers and the local government when the larger businesses were audited. Sanctions like belated fees, fines and criminal sentences were regarded as bargaining chips for tax agencies to establish their political network. If the negotiation had an obvious enforcement cost, tax authorities were more likely to pass the cases to the higher tax office. This study suggests that the frontline tax officials preferred more measurable and quantitative work—for example, annual tax revenue targets—while they were demotivated to implement qualitative targets like auditing local businesses. Moreover, information asymmetry made the higher-level tax officials and the SAT unable to recognize the illegal behavior of tax agencies.88 To discipline frontline tax officials is never easy since most are constrained in a vicious circle. Only when the institutional environment is changed can the outcome of law enforcement be substantially enhanced. 88 Wang & Hu, supra note 22; Kostka, supra note 5. 260 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 APPENDIX Tax Officials (N=34) (P1-P21 are from Bureau P, R1- R13 are from Bureau R) P1. Male, 42 years old, bachelor’s degree, chief of fifth tax branch, from County J, worked in tax authorities for 20 years, CPC member, worked as tax administrator, chief of administrative office, and vice chief of tax branch before; P2. Male, 46 years old, bachelor’s degree, vice chief of fifth tax branch, from County J, worked in tax authorities for 25 years, CPC member, worked as vice chief of a tax branch and auditor of audit bureau before; P3. Male, 40 years old, bachelor’s degree, vice chief of fifth tax branch, from County J, worked in tax authorities for 19 years, CPC member, worked as auditor in audit bureau before; P4. Male, 55 years old, college’s degree, tax administrator of fifth tax branch, from County J, worked in tax authorities for 31 years, worked as tax administrator of a tax branch before; P5. Male, 53 years old, college’s degree, tax administrator of fifth tax branch, from County J, worked in tax authorities for 30 years, worked as tax administrator of a tax branch before; P6. Female, 29 years old, bachelor’s degree, tax administrator of fifth tax branch, from County J, worked in tax authorities for 4 years, CPC member; P7. Male, 25 years old, bachelor’s degree, tax administrator of fifth tax branch, from the downtown of the city, worked in tax authorities for less than 2 years, CPC member; P8. Male, 53 years old, bachelor’s degree, chief of audit bureau, from County J, worked in tax authorities for 31 years, CPC member, worked as chief of a tax branch before; P9. Female, 37 years old, master’s degree, vice chief of audit bureau, from County J, worked in tax authorities for 10 years, CPC member, worked as tax administrator of a tax branch before; P10. Male, 52 years old, college’s degree, auditor of audit bureau, from County J, worked in tax authorities for 30 years, CPC member; P11. Male, 54 years old, bachelor’s degree, chief of first tax branch, from County J, worked in tax authorities for 32 years, CPC member, worked as vice chief of a tax branch before; P12. Male, 50 years old, college’s degree, vice chief of first tax branch, from County J, worked in tax authorities for 30 years, CPC member; 2018] WHY LAW ENFORCEMENT IS WEAK IN CHINA 261 P13. Male, 45 years old, bachelor’s degree, chief of sixth tax branch, from County J, worked in tax authorities for 23 years, CPC member; P14. Male, 40 years old, bachelor’s degree, chief of department of administration of tax collection, from County J, worked in tax authorities for about 20 years, CPC member; P15. Male, 45 years old, bachelor’s degree, vice chief of department of administration of tax collection, from County J, worked in tax authorities for about 20 years, CPC member; P16. Male, 45 years old, bachelor’s degree, chief of tax service hall/center, from County J, worked in tax authorities for about 20 years, CPC member; P17. Male, 30 years old, master’s degree, staff of department of policies and regulations, from County J, worked in tax authorities for 3 years, CPC member; P18. Male, 52 years old, college’s degree, tax administrator of sixth tax branch, from County J, worked in tax authorities for 30 years, CPC member; P19. Female, 45 years old, bachelor’s degree, chief of fourth tax branch, from County J, worked in tax authorities for 20 years, CPC member; P20. Female, 40 years old, bachelor’s degree, staff, worked in department of administration of tax collection for 19 years, CPC member, from County J; P21. Male, 45 years old, bachelor’s degree, staff, worked in department of administration of tax collection for 20 years, CPC member, from County J; R1. Male, 56 years old, college’s degree, chief, from County J, worked in tax authorities for 36 years, CPC member, worked as chief of department of policies and regulation of city tax office before; R2. Male, 51 years old, bachelor’s degree, vice chief, from County J, worked in tax authorities for 32 years, CPC member; R3. Male, 34 years old, bachelor’s degree, taking charge of tax administration, from County J, worked in tax authorities for 11 years, CPC member; R4. Male, 35 years old, bachelor’s degree, taking charge of tax service hall/center, worked in tax authorities for 10 years, CPC member, from County J, worked as tax administrator before; R5. Male, 34 years old, bachelor’s degree, tax administrator, worked in tax authorities for 6 years, CPC member, from downtown of the city; 262 COLUMBIA JOURNAL OF ASIAN LAW [Vol. 31:218 R6. Male, 40 years old, bachelor’s degree, tax administrator, worked in tax authorities for 5 years, from County J, CPC member, worked as tax administrator in the other city before; R7. Male, 34 years old, bachelor’s degree, tax administrator, worked in tax authorities for 7 years, CPC member, from County J; R8. Female, 29 years old, bachelor’s degree, staff, worked in administrative office for 6 years, from the other county of the city, CPC member; R9. Female, 28 years old, bachelor’s degree, staff, taking charge of data analysis for 5 years, from County J, CPC member; R10. Male, 32 years old, bachelor’s degree, tax administrator, worked in tax authorities for 9 years, from County J, CPC member; R11. Female, 25 years old, bachelor’s degree, staff, taking charge of issuing tax invoices in tax service center for 3 years, from County J, CPC member; R12. Female, 30 years old, master’s degree, staff, worked in department of administration of tax collection in city local tax office for 5 years, from County J, CPC member; R13. Male, 45 years old, bachelor’s degree, chief of department of tax policies and regulations in city local tax office, worked in tax authorities for 25 years, from County J, CPC member. Copyright of Columbia Journal of Asian Law is the property of Columbia Journal of Asian Law and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.