Abstract
Effective policy implementation is a core component of the Chinese political system’s adaptability and stability. A thorough investigation of local implementation mechanisms, however, is often hindered by an almost exclusive concentration on implementation efficiency. This article introduces a new analytical framework and suggests focusing on the interactions between the different administrative tiers—counties, townships, and villages—to understand local policy implementation in terms of procedural and outcome effectiveness. It argues that the triangle of central policy design, institutional constraints, and strategic agency of local implementers explains cases of effective policy implementation that can be observed throughout China. By way of studying the “Building of a New Socialist Countryside” in four cases, this article shows how effective policy implementation can be the result of what students of local governance have so far rather treated as obstructive factors, namely performance and cadre evaluation, financial scarcity, limited public participation, and the focus on models.
Keywords
The local state has been subject to increased scrutiny by China scholars since the early 2000s, reflecting a new concern on the part of the Chinese central state to resolve the “three peasant problems” 三农问题 by pushing ahead rural development and urbanization. The tax-for-fee reforms, initiated as early as 2001, the abrogation of agricultural taxes in 2006, the launch of the Building a New Socialist Countryside 社会主义新农村建设 program around the same time, and the institutionalization of a system of fiscal transfers to replenish local coffers must all be seen against this background. Today, a complex process of urban transformation is taking place in the Chinese countryside and extending down to village level, which includes the rapid commodification of land (Trappel, 2011), the systematic relocation of peasants to “new villages” 新农村 and “new neighborhoods” 新社区, either in the countryside (in situ urbanization) or in the urban peripheries (Chen, Liu, and Zhang, 2004; Hsing, 2010; Wang, Feng, and Xie, 2004; Yew, 2012), the dilution of the hukou system and the standardization of rural and urban architecture at astonishing levels. 1 So far, China scholars have pointed out how the well-known flaws in the Chinese bureaucratic system hinder or even sabotage local policy implementation and have shed light on the many negative consequences that it entails with respect to land and labor rights, distributive fairness, and the allocation of public money. Their findings, in most cases, paint a bleak picture: given a political system that does not hold its officials sufficiently accountable, the local state is mostly depicted as the epitome of China’s perennial governance crisis: it is dysfunctional, non-transparent, corrupt and, most certainly, highly undemocratic.
However, things are getting done in the local state: new roads, irrigation and fresh water drainage systems, schools, and houses are being built and the provision of public goods, such as education, vocational training, and social welfare, is improving. At the same time, local bureaucracies have become more professional and service-oriented, and administrative procedures more institutionalized and law-based. Government and cadre performance assessment, despite being designed as an internal mode of top-down control for securing the compliance of local agents, gradually integrates feedback mechanisms to ensure some level of inclusion of popular demands and public supervision, and even though most upper-level mandates are still underfunded in the local state, 2 they have been met with rising levels of government money and much creativity (e.g., private sector alliance-building) by local officials. In many localities this has substantially benefited the needs of the local populace, but this certainly does not mean that all needs have been satisfied and all policies have been implemented efficiently.
Nevertheless, scholars working in the field of China’s local state have expressed pessimism about its performance with regard to output efficiency and responsiveness to the people’s needs. They may agree that things are getting done, but they also point out that central goals are compromised or even displaced by local agency during the implementation process (see below). They also insist that the way things are being done is not sustainable and reflects the widely remarked dysfunctionalities in China’s political system. In fact, many scholars follow the landmark study by Kevin O’Brien and Li Lianjiang on “Selective Policy Implementation in Rural China” (1999), which showed how unpopular policies are implemented as a result, ironically enough, of institutional reforms originally meant to improve implementation efficiency and accountability—most notably the cadre responsibility and evaluation systems. These, as the authors argued, induce local cadres to consider the reactions of their superiors rather than the needs of the population, with the result that they tend to lose sight of public needs and demands and carry out projects which can be measured. Interestingly enough, at the end of the article, O’Brien and Li noted that “more research is needed to clarify under what conditions street-level discretion serves policy implementation and under what conditions it allows local officials to displace central goals with other interests” (O’Brien and Li, 1999: 182), which suggests that the implementation of popular policies is possible, even if somewhat unlikely.
Scholars have also tried to shed more light on the nexus between agency and institutional constraints in the shaping of local politics. For instance, Cai Yongshun found that local cadres create an “irresponsible state,” “in which politicians make decisions and allocate resources so as to serve personal political interests which take precedence over public concerns, resulting in the waste of public resources” (Cai, 2004: 22). He drew attention, in particular, to the cadre management system, which rewards “image-building projects” to the detriment of economic development (and the public interest) without putting any mechanisms in place to control or rein in local officials. Graeme Smith, for his part, discovered “an internal logic of corruption and selective policy implementation” (Smith, 2009: 29) by informal cadre networks which penetrate the institutional fabric of county and township governments. These networks build a “shadow state” that connects personalized rule, bureaucratic behavior, and corruption, and has little interest in serving the populace, because it does not have to contend with meaningful public participation or supervision. In addition, Smith stated that “the persistence of these practices, organizational forms and social ties can be a source of corruption and de-legitimization for the local state, but they are also a source of assets and resources, and a basis for coordinated actions when the local state chooses to implement the policies of higher levels of government” (Smith, 2009: 30). In the latter case, as he argued, however, a local government would only implement policies that were important for the annual performance and cadre evaluations, and that would raise revenue and benefit individual cadres financially. Like Cai, Smith expressed strong doubt that policy implementation in the local state could be efficient or responsive.
Ben Hillman (2010) adopted the notion of “factionalism” to conceptualize local cadres’ behavior under conditions of fiscal decentralization and “increased opportunities for profiteering” resulting from the “institutional dynamics” in the local state. Like Smith, Hillman founded his assessments on observations in one county. He identified factions that crisscrossed different government and party bureaus, controlled local policy making, and successfully hijacked the cadre management system. For Hillman, their cultural bonds “are an important source of factional strength and resilience in the face of central government efforts to instill greater fiscal discipline, strengthen accountability mechanisms, and punish the most egregious cases of official misconduct” (Hillman, 2010: 7). Although prone to corruption, “factions can help to coordinate policy action” and “make things happen” (15) and they have sustained the Chinese political system “by organizing and stabilizing political competition, supplementing formal institutional deficiencies and rewarding underpaid local élites with spoils” (18). Hillman is more circumspect when it comes to assessing local policy implementation and appears to be skeptical about its quality.
Zhou Xueguang (2010) introduced “collusion” as an alternative tool for analyzing local state politics and policy making. Similar to Smith and Hillman, Zhou finds that collusion encapsulates best a local state characterized by personal networks of officials stretching across different functional units within a local government. Collusion, according to Zhou, is the result of the “institutional logic of the Chinese bureaucracy,” which gives rise to three paradoxes that have a critical impact on local policy implementation: the paradox of uniformity in policy making and flexibility in implementation; the paradox of incentive intensity and goal displacement; and the paradox of bureaucratic impersonality and the personalization of administrative ties (Zhou, 2010: 55). In fact, collusive practices, such as the diverting of earmarked funds for other purposes, are as much “an indispensable coping strategy for local governments to implement state policies” (69) as they are often responsible for inefficient policy implementation (or “goal displacement”). They are “adaptive mechanisms” (74) that remedy the monitoring problem that accompanies the centralization of decision-making processes and the incentive structure to make local cadres respond positively to upper-level demands. Zhou Xueguang’s analysis may perhaps be the most sophisticated in the line of recent studies that have dealt with the local state’s predicaments. It shows that many tradeoffs are generated by the effort to reconcile the central state’s political agenda-setting and its claim to overall control on the one hand, and local state agency that must be left with considerable autonomy to implement policies on the other. “Collusion,” like Hillman’s factionalism, may or may not lead to good policy implementation in the sense that public needs and demands are satisfied, but in any case, it bespeaks “organizational failure” in the Chinese political system (74). 3
The studies of the local state discussed above are representative of the rather negative conclusions of China scholars that have emerged during the last ten years or so. 4 Interestingly enough, this picture conflicts with a different stream of research that has found high degrees of learning capacity and adaptability in the Chinese political system at all administrative levels with respect to the formulation and implementation of policies that resolve relevant problems. The recent literature on policy innovation and experimentation, as represented by Sebastian Heilmann (Heilmann, 2008a, 2008b, 2009; Heilmann and Perry, 2011) and Wang Shaoguang (Wang, 2008, 2009), indeed suggests that local political elites play a much more positive role in policy implementation, and that the institutional environment they face is much more helpful in bringing about tangible results for the populace than is usually contended in the China studies field. Heilmann, for his part, has emphasized that local policy experimentation has allowed the central state to shirk responsibility for policy failures by blaming local governments, while at the same time elevating successful local best practices and successful policy experimentation to national models for emulation. For Heilmann, the two formulae, “maximum tinkering under the shadow of uncertainty” and “experimentation under hierarchy,” encapsulate the logic of the Chinese policy process, which results in effective policy implementation and overall system stability through high levels of adaptive learning at all levels of the administration. Wang Shaoguang has offered support for Heilmann’s argument, contending that adaptive governance is also exhibited in the field of public goods provision down to the villages, about which Heilmann himself actually expressed doubt (Heilmann, 2008b: 23).
China scholars clearly have differing opinions on China’s policy process and policy implementation in the local state, and on the way this affects overall system stability. One explanation for this puzzle seems to derive from the fact that micro-political studies of the local state tend to focus on cadre agency and portray local officials as rational (selfish) opponents of central state policies. Macro-political or “top-down” studies, like the “experimentation” literature cited above, focus on processes and institutions, with the latter figuring as effective incentives for policy innovation and a powerful means of enforcing cadre compliance. Both “schools” find that things are getting done in the local state, but scholars who are looking into the “political machinations” of the local state seem to see this as happening “by default” while those who are concerned with innovation and experimentation as core features of the Chinese policy process tend to view effective (and responsive) policy implementation as a major indicator of adaptive governance in contemporary China and as offering an explanation for much of the current system’s stability.
We argue that a new approach is necessary to bridge this gap in perception—an approach that combines systematic policy research with analyses of institutional change and cadre agency in the local state, conceptualizes implementation in terms of procedural quality and cadre responsiveness to public demands, and helps us to understand better how and under what conditions local policy implementation can be effective. Hence, effective policy implementation is different from efficient policy implementation, which implies the existence of an external (“objective”) standard of measurement for the best way of allocating public funds and responding to public needs. Generally speaking, and in accordance with systems theory à la David Easton (1967), effective policy implementation (EPI) is measured by the capacity of a political system to generate system stability by adapting to demands from its environment and, consequently, produce outputs that entail (specific and diffuse) support for the system. With specific reference to the Chinese case, EPI means that
a policy has been implemented in accordance with central and upper-level guidelines that respond to a relevant problem;
this policy has been successively adjusted to local requirements and conditions;
policy implementation has been checked by performance assessment and cadre evaluation procedures;
public support has been mobilized through goal-defined inclusion of public demands;
and that, ultimately, tangible results “on the ground” have been achieved.
EPI does not mean ideal implementation in the sense that all upper-level requirements have been fulfilled, that performance assessment has been uncompromisingly rigorous, that all public demands have been satisfied, or that all resources have been allocated cost-efficiently to respond to these demands (see below). Hence, effective policy implementation is not measured against external benchmarks stemming from internationally accredited indicators of good governance or subjective expectations of what should be achieved or what is best in terms of local development and public goods provision—or what is meant by implementation efficiency. Nor does EPI require full democratic supervision by the public in the process of making, implementing, and evaluating policies. It is ensured by the nexus of top-down performance assessment and cadre promotion, limited public mobilization and deliberation, and the pressure of public demands that feed into the political system by telling it what kind of policies must be initiated. Hence, EPI is more concerned with procedural effectiveness than outcome efficiency, although outcomes must still meet public demands to the extent that a critical degree of system support is ensured.
This approach to assessing local policy implementation and system stability requires a detailed analysis of how local governments handle policy adaptation, project launching and coordination, budget allocation, cadre and performance evaluation, and the mobilization of popular support as well as the deliberation and inclusion of public demands (Ahlers and Schubert, 2009; Schubert and Ahlers, 2012a). As we discovered when conducting our research in different Chinese counties during the last few years, there is much more to local policy implementation than un(der)-funded mandates, bureaucratic collusion, and factionalism leading to illegal deviations and embezzlement of public funds, the violent resettlement of villagers to spur on enforced urbanization, and “empty shells” showcasing local development that is of no use to the populace. Instead of pointing at systemic failures (or dysfunction) and bad policy implementation, 5 we have taken an interest in explaining why county, township, and village cadres have driven forward local development and public goods provision in many parts of China and how, in this context, strategic cadre behavior locks in the operational logic of the Chinese political system, ensuring its stability and, arguably, legitimacy. We believe that an analytical focus on effective policy implementation, as it has been defined, helps to make sense of the Chinese political system’s resilience by (local) adaptive governance, and that China scholars would be well advised to think more in terms of effectiveness in policy implementation and the assessment of system stability by policy outcomes that may be suboptimal in terms of procedural and output efficiency, but still good enough to make the system persist.
In the following sections, we take a fresh look at the implementation of the “Building a New Socialist Countryside” (XNCJS) policy to highlight the procedural and output effectiveness in four counties. 6 The analysis is roughly structured by conventional policy analysis, focusing on the implementation and evaluation stages at the local level, that is, the county, township, and village tiers. We examine the strategic agency on the part of local cadres in setting up XNCJS projects and trace the processes and outcomes of policy implementation against the background of local development blueprints, fiscal constraints, cadre evaluation, and types of goal-oriented inclusion that are supposed to ensure that public demands are met and a critical degree of popular feedback is put in place. We conclude by generalizing from our findings on the distinctive features of the Chinese policy process in the local state and make a final argument for conceptualizing effectiveness as a concept to enhance our understanding of the relationship between local policy implementation and system stability in contemporary China.
“Building a New Socialist Countryside”: Content, Design, and Outcomes
The XNCJS policy program was officially promulgated by the central government in 2006 as an initiative to spur rural development in the aftermath of the tax-for-fee reforms of the early 2000s (Central Government of the People’s Republic of China, 2006; State Council, 2006). 7 It primarily entailed the merging of previous policy reforms and newer initiatives, for example, agricultural modernization, infrastructural development, rural governance innovation, the expansion of social welfare, the strengthening of rural education, fiscal reform, and so forth, into a more comprehensive conceptual framework to “support the countryside” 支农. Simultaneously, transfer payments and funds earmarked for these purposes were increased dramatically: according to official figures, during the Eleventh Five-Year Plan period, XNCJS-related central government expenditure (termed “government spending on agriculture, rural areas, and farmers”) rose from RMB 339.7 billion in 2006 to RMB 857.97 billion in 2010. 8 The complementary decision to abolish the agricultural tax by 2006 was not only epochal in that it relieved the Chinese peasantry of a burden it had shouldered for more than 2,500 years; it also marked a far-reaching paradigm shift, in that, from now on, theoretically, “central and local finance began to cover rural public goods and services, with the state replacing the farmers as the main provider of rural public goods and services” (Ye, 2009: 140). Hence, the tax income of local governments decreased, while their responsibilities grew. The central government pledged to compensate local governments for their losses in revenue by means of extended transfer payments through the intergovernmental transfer system and by providing substantial earmarked XNCJS funding. Parallel to these measures, budget allocation and monitoring mechanisms were partly altered (see below). Funding mechanisms were also adjusted to local circumstances: for poorer areas, the largest part of the XNCJS budget was composed of increasing central (and sometimes provincial) subsidies, but the more developed and well-off regions or counties had to generate and allocate the bulk of XNCJS funds by themselves.
Notwithstanding this increased central impetus and funding, XNCJS was especially formulated and designed to encourage new conceptual initiatives for rural development on the ground. Given that the central state’s blueprint was not very specific—one could even argue that central advice was “intentionally vague” (Ahlers and Schubert, 2009: 57)—XNCJS should rather be regarded as a macro policy or a policy framework: implementation was required by all local governments nationwide, while specific targets, prioritization, and methods were subject to local concretization. This entailed each locality having to set up its own specific development blueprint 规划, projects 项目, coordinating mechanisms, and evaluation guidelines, in order to reconcile local needs and capabilities with short- and long-term development goals, the latter having to correspond with nationwide objectives of rural transformation. In fact, XNCJS as designed by the central state left ample leeway for local policy adaptation and innovation.
Provincial guidelines communicated the regional priorities of XNCJS-related rural development further down through the governmental hierarchy to the lower administrative tiers. Target quotas and performance indicators as well as the definition of “one item veto” 一票否决 criteria, 9 created a corridor for local, that is, county and township, policy implementation. It was precisely at the county level that the most crucial adaptation took place and that ultimate project implementation was put on track. Acting as a pivotal unit in the local administrative hierarchy, counties coordinated overall XNCJS implementation within their jurisdictions, deciding on project selection and the allocation of funds. They also managed township “collaboration” in the process of identifying project localities and steering project management on the ground.
In all four counties that we investigated—Dingnan 定南 (Jiangxi province), Mizhi 米脂 (Shaanxi province), Nan’an 南安 (Fujian province), and Qingyuan 庆元 (Zhejiang province) (for further details, see the Appendix)—comprehensive local XNCJS blueprints existed and were strongly propagated within the government apparatus. The Dingnan and Qingyuan county governments presented the most comprehensive local XNCJS strategies. In both these counties, XNCJS projects were spread spatially and the local governments boasted of their local achievements and recognized “model practices” 模式. For instance, Qingyuan, the least developed county in Zhejiang province, made successful use of its backwardness in terms of industrialization and transportation connections to promote its “intact” eco-environment, “organic agriculture,” and “green enterprises,” and to sharpen its profile as “China’s No. 1 eco-environment county” 中国生态环境第一县. Mizhi county, for its part, had the most modest overall XNCJS program and hosted the only case in which one of the planned measures, a minor urbanization (or resettlement) scheme included in Mizhi’s five-year XNCJS plan, had not been implemented, at least on the occasion of our last visit in 2009. The coordination and implementation of all the other measures, mostly pertaining to infrastructural development, had nonetheless been carried out with tangible results, while special funding was granted to “pilot” and “model” villages. The county government in Nan’an could already point to successful cases of rural industrialization before the introduction of XNCJS, which then served as a frame of reference to rearrange and streamline its development blueprint and evaluation system.
In terms of outcomes, the overall impression we gained from internal data and reports, from our respondents (officials at county, township, and village levels, as well as village residents), and also from our numerous visits to project sites was that overall socioeconomic development had accelerated in all four counties, although to quite different degrees, during the past decade. 10 Locally associated with sannong policies, or XNCJS in particular, were the enforcement of basic infrastructure projects (for example, road construction, electrification and wiring, biogas and solar power generating facilities, extension of irrigation systems, improvement of sanitary facilities, information and communication technology, etc.), the intensification and commercialization of agriculture, the extension of social services in the villages (including the New Rural Cooperative Medical System, the reformed minimum living allowance scheme, improved elderly care facilities, cuts in school fees, etc.), the endorsement of local folk culture, and, very often, decentralized urbanization and “town-ization” in combination with reformed local administration schemes. Even if there was much more discontent than we were able to detect and XNCJS implementation suffered from legitimacy problems (particularly with respect to democratic participation and supervision), there was no reason to doubt that local policy implementation had so far been relatively effective in all four counties when measured against the county governments’ development plans and projections and the visible benefits, limited as they may have been from an external perspective, for the local populace.
Tracing Procedural and Outcome Effectiveness at the Local Level
Coordinating and Steering Project Implementation
County governments are the core units responsible for local policy implementation. They enjoy considerable autonomy throughout a large part of this process. Although the content of rural policy remained largely unchanged, the new XNCJS initiative changed the scope and variation of project funding for rural development which counties could apply for at higher levels. This led to an increasing need for bureaucratic coordination, not only to acquire funding from above, but also to carry out decision making, task assignment, and project allocation between leaders, bureaus, and departments. Hence, some fairly radical adjustments were made to the organizational set-up of the county government apparatus. Generally speaking, the change to project-based administration 项目管理的模式 represented a major shift in the approach to rural development that has taken shape over the past ten years alongside the introduction of XNCJS. At the same time, the counties’ relationships with superior levels where they apply for project funding, has—in most cases—changed with the implementation of the “province administers the county” 省管县 administrative reform (Liu, 2008; Wang and Li, 2010). Informal negotiations between the county government and upper levels (prefecture and province) on funding, policy targets, and evaluation indicators, for example, became more common and, to some extent, institutionalized (Ahlers, 2014). As a result, in all four cases, the XNCJS-related decision-making and coordination process at the county level was redesigned and modified. Two new coordinating bodies were established: the XNCJS Leading Small Group 新农村建设领导小组 and the XNCJS Office 新农村办公室 (or 新农办). The XNCJS Leading Small Group, chaired by the county leaders and comprising all departments relevant for XNCJS as well as the County Organization Department 组织部 and the local Development and Reform Commission (DRC) 发展和改革委员会 (or 发改委), discussed and made decisions on major issues concerning XNCJS implementation, such as the selection of development plans and projects submitted by township governments for funding, the final allocation of these funds, and the coordination of government bureaus in cases where there were overlapping competencies or competing interests. 11 The XNCJS Office, for its part, was in all four cases affiliated with the Bureau of Rural Works 农村工作部 (or 农工部), a Communist Party unit, 12 which highlighted the importance attributed to XNCJS. Its main responsibilities were, again, the coordination 协调 of tasks and projects for which different government departments had overlapping responsibility, the collection of data to document project implementation, the identification and management of pilot and model villages, etc. In this way, the XNCJS Office to a great extent operated as the executive body of the XNCJS Leading Small Group (Ahlers and Schubert, 2009; Schubert and Ahlers, 2012a).
This administrative reorganization was often accompanied by strategic local modifications of formal and informal institutions and communication flows. For example, because of the overlap of space and personnel, the XNCJS Office and its hosting parent unit, the Bureau of Rural Works, were usually both called and referred to as one and the same body. In fact, the latter unit had been meaningfully upgraded in all four counties by the establishment and affiliation of XNCJS Offices. The most noteworthy elevation happened in Dingnan county, where the Bureau of Rural Works’ director was holding a threefold post: he also served as a member of the county’s CCP Standing Committee and as deputy county commissioner in charge of rural issues. Consequently, his bureau not only acted as the pivotal authority 枢纽 for all the decision-making and coordination concerning XNCJS projects in Dingnan, including the allocation of project funding, but became almost as powerful as the Organization Department; the Bureau of Rural Works was being given additional responsibility for “party building” activities at the grassroots level and substantial authority for designing and carrying out XNCJS project and performance evaluation—both fields that traditionally belong to the domain of the Organization Department. In comparison, Qingyuan county’s Bureau of Rural Works and XNCJS Office were not empowered to such an extent, but were still crucial for the basic XNCJS coordination. As one of the Bureau of Rural Works’ leading cadres stated, “although there are no formal institutional provisions for interdepartmental coordination meetings steered by our bureau, it has become common practice for us to hold these meetings” (Interview, Sept. 12, 2009).
The other county-level bodies mentioned above also played an active part in XNCJS management. For example, in the four cases that we investigated, the County Organization Departments integrated XNCJS into the local Party School curricula, pushed forward sannong-related grassroots party activities, and, to differing degrees, incorporated XNCJS into their cadre and department performance evaluations. This was, of course, primarily carried out in accordance with higher-level evaluation guidelines, but these were sometimes substantially modified by the county governments (see below). The stronger XNCJS was sold as a core party task and as appealing to basic cadre virtues, so that something almost like missionary zeal among cadres at all levels could be observed. Administrative streamlining, enforced communication and clearing between government departments, and the generation of ideological coherence by cadre training and “thought work” 统一思想 ensured that policy measures attached to XNCJS were spread throughout the counties, XNCJS projects were put on track relatively smoothly, resources and efforts were applied reasonably well, and supervision was installed, generating tangible outcomes in terms of local infrastructural development and public goods provision.
Concerning outcomes, money, of course, was critical. In the three poor to moderately developed counties—Dingnan, Mizhi, and Qingyuan—the bulk of XNCJS funding was provided by higher-level subsidies. Earmarked funds 转移支付 covered between 50 percent (Dingnan) and 70 percent (Mizhi) of the expenditure on single projects, and the necessary local share (see below) was largely taken from general transfer payments 专项补助. 13 In contrast, in Nan’an county, the most advanced of the four research sites, XNCJS-related projects were only supported at the average rate of 35 percent at most and although transfer flows were hidden in local finance statistics, it could be assumed that they were insignificant, compared with those in the other three cases, because of the logic of the tax system. Almost all the interviewees in Nan’an took the opportunity to lament the lack of higher-level funding, although subsidies for rural development were said to have increased continuously since the launch of XNCJS.
But whether the payments were perceived as generous or meager, financial monitoring in all four counties was said to have been critically expanded, especially since subsidies increased under the heading of XNCJS. In general, respondents agreed that the newly enhanced mode of disbursing funding on the basis of project applications made the diversion of (earmarked) transfer funds much harder. In some provinces, such as Jiangxi, the management of XNCJS projects has become a “one item veto” evaluation indicator. In Zhejiang province, the effective use of public funds is rewarded. In all the four localities that we visited, special bodies were established to assist the XNCJS Leading Small Group in the process of project finance. In line with central legislation, a Leading Small Group for the Integration and Coordination of Rural Support Funds 支农资金整合协调领导小组 was responsible for overseeing the proper distribution of project funds (Ministry of Finance, 2006). Furthermore, in Qingyuan, the Bureau of Finance also established a so-called “information exchange mechanism” 信息互通机制 to supervise all finance transactions related to rural development, including those involving nongovernment actors, such as farmers cooperatives. This was explained to us as an additional means of avoiding illegitimate multiple applications for governmental subsidies. Improving procurement and the management of limited budgets were hot topics of discussion among county cadres, especially in the poorer provinces (Ma, 2009: 36), where local leaders could well make their mark by promoting local best practices, as was reported in Shaanxi (Interview, leading official, Department of Rural Works, Yulin city, Sept. 2, 2009).
County governments were generally obliged to mobilize part of the funding earmarked for XNCJS-related measures by themselves. Interestingly enough, however, what might at first be seen as an additional constraint on their maneuvering space was turned into a resource for county governments’ steering capacity. The practice of raising complementary funds to match project money coming down from above 配套资金 encompasses all administrative levels in China. It is based on the expectation that, although the necessary funding could possibly be provided entirely by upper levels, the mobilization of the local peitao 配套 helps to prevent the abuse of the transfer system and to spur on goal-oriented, efficient project management. Following a successful project application, the provincial or city government allocates only a portion of the estimated project budget to the county. After completion of the project, a county is eligible to apply for partial reimbursement 报销, but often enough the local contribution is not compensated. In poor or less-developed counties, such as Dingnan, Mizhi, and Qingyuan in our research, the matching requirement is generally moderate compared with that in more affluent localities; securing compensation later, however, is difficult everywhere (Ahlers and Schubert, 2009: 48–49; Schubert and Ahlers, 2012a: 71–74).
In fact, county governments made use of the same tool and determined the amount of matching funds to be contributed or mobilized by the townships which, for their part, then went to the villages. However, the townships in Dingnan, Mizhi, and Qingyuan, in general, only had to provide limited peitao as they did not have much money of their own to contribute.
14
Generally speaking, township budgets were subjected to strict controls by the county governments and township governments did not have any leeway to steer their own agendas.
15
Matching was a different matter for the villages in all four counties; some were relieved of any such requirements, some contributed physical labor or a fixed sum per household, while others took the necessary project budgets completely out of their collective coffers or raised private donations,
16
and ultimately made higher contributions to local policy implementation than the township level (especially in Nan’an). To further control townships and villages, a special “county-level reimbursement system” 县级报账制 was adopted in the mid-2000s. A leading cadre from the Bureau of Rural Works in Mizhi described its operation: At the current stage, central funding for XNCJS is not paid all at once, but divided into three payment stages. First, 30 percent is given to start up XNCJS projects 项目的启动; at the second stage, 40 percent of the sum is disbursed to support the project process under way; and the last 30 percent is paid after the completion of the project. This supports the real time monitoring of project development and progress. (Interview, Sept. 1, 2008)
The matching could sometimes be negotiated, but only in the region of a few percentage points. Basically, for most projects at all levels the prevailing principle was: “Show that you are really willing and able to implement a project by mobilizing some kind of matching.”
Townships and villages, hollowed out by fiscal centralization since the early 2000s, were generally heavily dependent on financial support provided by the county, much in the same way that the counties, in turn, were dependent on higher-level governments, and this gave the county enough leverage to secure compliance from its subordinates. Township governments were ascribed by our respondents a “missing link” or nexus function 承上启下 between the county and village levels, basically in terms of preparing, overseeing 监督, and checking 验收 project implementation. In most respects, power and functions were clear-cut: the townships were reduced to a kind of cooperating agency that helped villages during the application process but were not allowed to have a say in the ultimate decision-making process. 17 The townships therefore usually tended to ensure that as many of their concerns as possible were integrated in the county government’s development blueprint beforehand, because negotiating changes was virtually impossible once plans had been made and targets set. Although township representatives did not admit that any bargaining took place, there was evidence to suggest that substantial informal negotiations and lobbying had occurred between the two levels before the allocation of upper-level funding was finally settled. Nevertheless, county and township governments cooperated rather well in all four counties under observation.
Township administrations communicated to the villages the XNCJS projects they could apply for, according to the county’s annual plans. They helped the county government to collect grassroots data and to gauge the feasibility of proposed village-level projects (Ahlers and Schubert, 2009; Schubert and Ahlers, 2012a). Particularly for this purpose, the traditional “village tutoring system” 包村制度 was revived during the early 2000s and, in most cases, specifically linked to XNCJS: at the township level, leading and functional cadres were made responsible for one or more of the villages within their administrative jurisdictions. 18 After the townships had recommended the villages they deemed eligible to apply for XNCJS funding, the final choice was then made by the county government. Most importantly, county- and township-level respondents unanimously admitted that, beyond all formal selection criteria, several additional aspects strongly influenced their decisions: the quality and performance record of the township and village “leading team” 领导班子; their ability to guarantee “solidarity” 团结 (that is, cohesion) and the smooth implementation of projects; the possible previous model status of a village—seen as evidence of good policy implementation; the fit of village plans with development goals set by the county government, including the villages’ drive to pursue modeling strategies and, generally, the scope of the funding that could be spent. Interestingly enough, respondents at village and township levels reported that applying for projects had become much easier since the launch of XNCJS, and that the majority of applications submitted were successful.
Popular Mobilization and Goal-Oriented Inclusion of Public Demands
XNCJS as a macro policy, which had to be adjusted to local conditions, was propagated nationwide to provide for popular mobilization and input. This bottom-up feature was, however, often diluted by the requirement of local governments to find a distribution formula for the limited funding of projects in the villages. Although villages designed their own development plans and could decide to apply for projects on their own, overall project definition and ultimate selection was based on further strategic considerations of township and county governments, as described above, and public participation in the policy process was therefore limited from the very beginning. Nevertheless, a certain degree of inclusion of the people was always present for the purpose of ensuring smooth policy implementation. In times of heightened sensitivity to “social stability” 社会稳定, discontent arising from faulty policy implementation, leading to petitioning 上访 (or 信访) or street protests by village residents, had to be avoided at all costs. Cadres agreed unequivocally that any occurrence of shangfang within their jurisdictions would seriously blacken their performance records and, hence, their prospects of promotion. The possibility of protest, be it covert or overt, individual or collective, thus constituted a serious constraint for local policy implementation. For this reason, as local officials often argued, they preferred to negotiate a deal rather than to force a policy measure upon villagers, even if this measure was legitimate and supported by the majority of the local population. In some cases, township or county governments invested more money than had originally been allocated for informally negotiated compensation payments or for “buying compliance.”
19
As a township party secretary put it: If village residents oppose a project, you have to consider whether it serves the majority of the villagers. Sometimes necessary measures will touch upon individual rights and some people will, of course, not agree that you may use their land, for example, for road building in the course of a village renovation project, so you have to make a detour. If you can’t reach a compromise, the government now cannot but take one step back, for example, by paying more compensation to these demanding people 难缠的人. The government sometimes has to buy over the opinion leader 收买意见领袖. (Interview, J. township, Sept. 23, 2008)
Although the truth of all these stories in which local governments accommodate villagers’ demands is often hard to establish, it can be hypothesized that noncompliance on the part of villagers is often met with acquiescence rather than suppression by local governments, simply in order to prevent it from evolving into an uncontrollable conflict. As far as could be ascertained, it was only when shangfang had actually started and could no longer be contained by negotiations and compromise that township and county cadres would shift from the mode of conflict prevention to pro-active conflict management, implying the use of force. 20
Hence, local governments attempted to mobilize public support to anticipate or accommodate resistance to XNCJS measures and to counter villager agency at an early stage. Strategies of inclusion by establishing more communication with villagers and their (limited) participation in the policy process were crucial for legitimizing XNCJS. They ensured input and feedback channels that, arguably, made policy implementation more responsive and effective. On the one hand, county and township governments had to formally adhere to the established institutions of grassroots autonomy: for each XNCJS project, for example, “democratic consent” by the villagers’ self-government authorities had to be solicited. This required at least a simple majority (or up to 85 percent in Dingnan and Qingyuan) of villagers in a formal decision made by the Village Assembly 村民大会 or the Village Representative Assembly 村民代表会议 to apply for a certain XNCJS project. On the other hand, new institutions were established to ensure villager consent and smooth implementation once a project was launched in the village. In Dingnan, for example, local authorities strategically utilized the social and political fabric of the village communities for this purpose. The county government (in line with earlier prefectural regulations) institutionalized so-called “Village XNCJS Administration Councils” (VAC) 村民新农村建设理事会. Comprising a handful of members elected by the villagers from their ranks, VACs were responsible for XNCJS project communication, acclamation, and implementation within a village (Dingnan County Government, 2008; Ganzhou City XNCJS Office, 2007; Schubert and Ahlers, 2012a). In Jiangxi, where XNCJS project funds were usually allocated for special “project sites” (dian, 点), roughly covering thirty households, a village could sometimes host up to about 20 of these dian. VACs sought support for the village development blueprint to meet the threshold mentioned above before the village assembly made a formal decision on the plan, and they cooperated closely with village officials who had to face their superiors and come up with strong villager support. The councils then became responsible for ensuring sound implementation until the plan was fulfilled, including the persuasion of recalcitrants. This was mainly done by “pulling social strings” 拉关系, that is, activating relatives and friends to take part in what is neatly called “doing work” 做工作 in China’s cadre terminology. Most importantly, VACs took care of project funds, composed of XNCJS dian subsidies and contributions made by the villagers themselves. If the VACs carried out their tasks successfully, they were soon placed in charge of all potential follow-up projects and general maintenance management, since term limits had not yet been established for elected members. From the perspective of local governments, these subunits at the village group level were a valuable instrument to assist village leaders in bringing XNCJS projects on track. They served as transmission belts to communicate to villagers what was to be done and, at the same time, solved problems that came up in the course of the implementation. In other words, they helped to ensure effective policy implementation.
As another strategy, Nan’an county promoted the cooperation of village elites in XNCJS projects in a way that merged dimensions of financial and political management. For this purpose, the county government sought to recruit private entrepreneurs as village party secretaries. In general, the argument that village cadres should act as role models for an “entrepreneurial village spirit” has long formed part of the national rhetoric (State Council, 2006). 21 In Nan’an county, it was acknowledged that “according to a survey, most of the successful cases of XNCJS in our county are in those villages which are headed by village party secretaries who had once gone out to do business or who are engaged in other business activities. This shows that the selection of good village cadres, especially party secretaries, should be the focus of XNCJS” (Nan’an County Bureau of Rural Works, 2011: 17). Private entrepreneurs—once they decided to join in the administration of a village—were immediately admitted to serve as officials with little formal procedure. These individuals often held dual posts 一肩挑 as village directors and village party secretaries. They played a prominent role in sketching the development blueprints for their villages, acting often enough as their sponsors. One village party secretary, for instance, explained to us that he had recently invested more than RMB 300,000 in one year in all kinds of projects, including XNCJS measures, in his village. In 2011 alone, he had contributed RMB 200,000 to a school building project (Interview, XD village, Sept. 5, 2011). Most of these political entrepreneurs had a vision which they were trying to realize in their villages, using all the material and nonmaterial resources available to them. For example, another village party secretary used the infrastructure and personnel of his information technology company to create an online village information and trading system, selling agricultural products and providing public services for rural residents. Once firmly established in the village that he administered, the system was successively expanded to neighboring villages and townships, and even beyond. At first sight, these “entrepreneurial Party Secretaries” (Yan, 2012) seemed to be fairly independent of their county and township superiors, who, in their own best interests, preferred to cooperate with them rather than attempt to rein them in. However, village party secretaries were kept in check because the upper levels controlled administrative procedures that were impossible to circumvent (regulations concerning land acquisition or redesignation, labor relations, taxation, etc.). In addition to seeking the contributions of entrepreneurs who were strategically recruited to serve as village party secretaries, the Nan’an county government also looked for private investment 民间投资 to support its local XNCJS policy and targeted local business people to contribute their fair share to the development of the environment in which they thrived (Ahlers, 2014).
Managing Cadre and Performance Evaluation
The evaluation regime is another crucial institution in the policy process which county governments were able to partly transform from a constraint on their agency into a central tool for effective local policy implementation. 22 Across China, this generally encompasses individual cadre evaluation 干部考评 and unit-based (departmental, bureau) performance assessment 目标 / 项目考核. All government bureaus and offices are screened regularly by their superiors, that is, townships by counties and counties by city/prefectural authorities (and sometimes also by provincial governments), typically holding the units’ leading cadres 领导干部 responsible for the results. Township governments, for their part, assess the work of village cadres. Although these processes are mainly structured in a top-down fashion, monitoring is also conducted horizontally, that is, county bureaus are evaluated by a county-level assessment procedure, and township offices are screened by an analogous procedure at that level. The targets to be tested are prescribed by the annual responsibility contracts 责任书, signed by the different departments. Due to its status as a core national policy, XNCJS has found its way into all categories of the evaluation system, although its prominence in individual cadre evaluation varies. While all localities evaluate subordinate leaders’ achievements in rural development in general, such as, for example, the growth rate of rural incomes and the state of infrastructural development in the villages, some, such as Dingnan county, referred to more specific local “XNCJS goals” as evaluation indicators. For project evaluation many targets and indicators were already determined by upper levels, but others could be adjusted to local conditions and a county’s particular XNCJS strategy. In fact, beyond simply filling out given templates, county governments in most cases created their own assessment schemes and methods based on locally defined criteria (Ahlers and Schubert, 2009; Schubert and Ahlers, 2012a).
XNCJS assessment was directed by three key departments, with minor differences between the four sites under investigation. XNCJS-related project, performance, and cadre evaluation was conducted either by
a county-level Evaluation Office directly under the umbrella of the Organization Department according to Shaanxi province’s “three in one” 三位一体 scheme, assisted by the Bureau of Rural Works in the case of Mizhi;
a separate county-level Evaluation Office, assisted by the Bureau of Rural Works and reporting to the Organization Department as in Qingyuan;
the DRC together with the specific departments, especially the Bureau of Rural Works and the Organization Department as in Nan’an;
or was
concentrated in the hands of the Bureau of Rural Works (as in Dingnan), which carried out quarterly inspections of XNCJS projects and even designed a XNCJS-focused cadre evaluation scheme in cooperation with the Organization Department.
The Bureau of Rural Works acted as the authoritative center, not only overseeing and coordinating all aspects of XNCJS project implementation, but sometimes also the individual or departmental performance evaluations in a county. It directed the setting of the indicators and benchmarks and transferred this catalogue to the XNCJS Leading Small Group and the local Evaluation Committee 考核委员会 for a final decision (Ahlers and Schubert, 2009; Schubert and Ahlers, 2011).
With regard to the customization of evaluation contents, the Dingnan county government’s modification of performance criteria pertaining to the superior Ganzhou city’s XNCJS scheme serves as a telling example: the Ganzhou city government had connected the assessment of “XNCJS village renovation” 新农村建设村庄整治建设考评 carried out by the counties within its jurisdiction to the benchmarks for the construction of new residential quarters and new villages, the construction of eco-civilized model villages, and the construction of “civilized roads.” 23 In the evaluation guidelines for townships and dian, the Dingnan county government then decided to omit the “civilized road” category and transferred the related 15 points 分 to a new section called “industrial development.” This section included three other important XNCJS-related goals, which were particularly emphasized in the county: large-scale agricultural specialization, farmers’ professional cooperatives, and the systematic provision of new income opportunities for rural residents. In this way, the county government was able to exploit its latitude to import or even duplicate indicators that had already appeared in the county’s other XNCJS-related guidelines in order to communicate, control, and reward the implementation of important local policy targets. Another example is the practice of “categorizing” the evaluation 分类考核 of lower units according to their development status, as happened in Nan’an and Qingyuan counties. 24 Taking into account the economic disparities within its jurisdiction, in late 2008 the Qingyuan county leadership, for instance, removed the “iron” appraisal criteria, such as GDP growth, industrialization, and attraction of investment, from the evaluation guidelines for the less-developed townships, and replaced them with criteria based on ecology-friendly development. Furthermore, in 2011, in order to safeguard its “No. 1 eco-environment county” status, Qingyuan even upgraded environmental protection and ecological conservation to countywide “one item veto” criteria 生态环保一票否决.
As XNCJS assessment became part of both program and cadre performance evaluation, it created a heavy additional workload for the local bureaucracies. Their work typically included computing data, producing detailed statements and reports, answering inquiries from above and, finally, receiving and accompanying higher-level inspections. A deputy township commissioner in Nan’an explained that his locality would be visited by more than thirty delegations from upper levels every year (Interview, JD township, Sept. 5, 2011). Although the evaluation 考核 was therefore primarily employed as a cascading disciplinary resource, local cadres reported relatively smooth cooperation in adapting and modifying performance criteria, not least between the county and township governments. These two levels actually depended on each other for the delivery of good results, since township performance was assessed by the counties and the counties needed the townships to produce good results to show to their superiors at the prefectural and provincial level. 25 Hence, negotiations on target indicators between these two levels seemed to occur quite frequently. Township cadres indicated that they could discuss target fulfillment indicators with the county government if they had had to face certain unforeseeable hardships (such as natural disasters or an economic crisis) during the review period, which cut into their project funds or led to sudden slowdowns that threatened local development objectives. In Dingnan, for instance, because of the financial crisis in 2008–2009, economic targets for townships were adjusted even during the running evaluation procedure. It was also reported that, with the exception of one-item-veto criteria, suboptimal results in one project field could be offset by excellent performance in other areas.
Despite the heavy burden that personnel and performance evaluation imposed on them, most of the local cadres that we spoke to considered it to be an inseparable part of the policy process. From the county level to the village level, respondents stated that policy implementation would not work without some sort of “pressure” and “control.” During recent years, the interviewees estimated, rural development yardsticks and their assessment had become more reasonable, and “relatively scientific.” The leeway for the flexible adaptation of performance criteria was regarded as ensuring the “adequacy” and “feasibility” of set targets, and making it easier for local implementers to fulfill their tasks; benchmarks, it was unanimously stated, could usually all be accomplished through “hard work.” Collusion between and across the different administrative tiers, as in the case of villages, townships, and counties that faced city- or provincial-level inspections, was admitted and considered legitimate. Upper levels, for their part, showed only limited interest in discovering problems and in intervening in subordinate levels’ affairs. As long as the one-item-veto criteria were not violated and the major policy goals were achieved, some ineffectiveness in carrying out performance evaluation was tolerated by all levels of government.
Although this confirms the findings of the China scholars who have criticized institutionalized performance evaluation as a distorting element of the policy process in the local state (see above), our impression was that its positive effects on policy implementation should not be underestimated. Evaluation had a disciplining influence on local cadres because, by definition, it always produced a ranking of performance quality and (non-) performers, entailing sanctions and rewards. While leading cadres merely had to be careful not to fall out of favor with their superiors, and, in a positive sense, had to take care to advertise their aptitude for potential career promotion, performance assessment created pressure by encouraging competition between the different governmental departments. Every year, a list of assessment results was made public, particularly highlighting the first and the last three positions on the list. Beyond this name and shame dimension, manifest consequences comprised awards for the first rank as well as “calls to order,” demotion in case of subsequent failure to complete projects, and financial punishment for poorly performing bureaus, meaning the reduction of their annual budgets. 26 This placed a bureau under great stress and seriously curtailed its prospects for excelling in the future (Ahlers, 2014). Evaluation rankings produced pressure to perform, since no entity wanted to finish at the bottom of the final list. The only way to avoid being stigmatized as nonperforming and to escape the ensuing consequences was to perform well, that is, effectively, which also included the anticipation and accommodation of local public demands, in order to avoid a negative “social stability” record.
Concluding Remarks: EPI and System Stability in Contemporary China
Our study of XNCJS implementation in four counties in contemporary rural China has shown that local governments, operating under financial stress, must and do respond to policy guidelines from upper levels, come up with their own development blueprints, and deliver results. The service orientation of county and township governments has become a major reference point for cadre legitimation since tax and fee extraction has been abrogated. The positive outcomes of policy implementation, in terms of tangible improvements of the rural infrastructure, public goods provision, and housing conditions, ensure the maintenance of local government autonomy vis-à-vis upper levels (shielding them from criticism and interference) and, moreover, bring about rewards, notably cadre promotion and additional funding. Negative outcomes, however, result in sanctions, most notably the loss of funding and blocked career paths.
Policy implementation in the local state is subject to strict fiscal regulations and cadre evaluation to ensure implementation effectiveness. There is leeway to bend the regulations, just as there is flexibility to adjust the indicators, but there is no way to completely circumvent these strictures. At the same time, the populace is increasingly demanding and occasionally presses hard for positive responses on the part of local governments. Not delivering public goods, job opportunities, and financial assistance in the case of broad-scale relocation measures means provoking protest and undermining social instability. Local cadres must perform in order to protect their autonomy and safeguard their career prospects, and they must do so by being responsive to both upper-level requirements and the demands of the populace. Although local governments may still divert or even waste public money, deceive upper levels, and oppress villagers, such behavior becomes increasingly costly for local cadres. Concluding from our experiences in four counties, we argue that in the face of central state policy guidelines, fiscal centralization, cadre and performance evaluation, and the increasing demands of the populace, local governments in contemporary China cannot but engage in effective policy implementation.
In the process of shaping and implementing local policy, county governments act strategically by establishing their own development plans and priorities, allocating (scarce) funds, negotiating with upper levels on policy targets, closing ranks with townships and villages (or forcing them into obedience), and by managing cadre evaluation, public mobilization, and goal-oriented inclusion. They creatively steer the implementation process to ensure that policies are adjusted to local conditions, bureaucratic frictions (both horizontally and vertically) are minimized, and interdepartmental cooperation is enhanced, that control mechanisms are put firmly in place, and cadre coherence and unity are ensured. Very often, local cadres also take pride in delivering results that serve the people, adding a dimension of idealism to their mainly strategic behavior that responds to policy demands and hard institutional constraints.
Effective policy implementation, defined as procedural and output effectiveness, results from a three-level game played out between (1) centrally designed policies and institutional control mechanisms affecting all administrative tiers; (2) strategic agency on the part of local governments, most notably at the county level, which is able to transform institutional constraints in political resources, although unable to avoid or defy control by upper levels; and (3) public demands that cannot be ignored and that must be anticipated and accommodated. In this way, effective policy implementation creates win-win situations for all parties concerned, that is, the central state, local governments, and the populace, in the sense that
the central state, as the principal, sees its political goals sufficiently taken care of and implemented by its local agents;
local governments (local cadres) secure their positions and prospects for promotion vis-à-vis the central state and their immediate administrative superiors by accepting their roles as state agents on the one hand and enjoying considerable autonomy in adapting, incorporating, and implementing upper-level guidelines on the other;
the local populace enjoys tangible policy results, which means better public goods provision, and generally speaking, better living conditions (including, e.g., access to urban housing and better sanitation) than before, both in terms of subjective perception and objective conditions.
At the systemic level, EPI can be seen as a mode of “structural coupling” (à la Luhmann) between the political system and its environment: the system reacts to external stimuli with internal adjustments but without changing the “operative logic” of the system. In David Easton’s terminology, EPI results from a feedback loop that connects the input and output functions of the system so that demands are processed and critical degrees of specific and diffuse support maintained that allow the system to persist. Or to put it more mundanely, the Chinese political system has learned to adapt in such a way that effective policy implementation in the local state is made possible—and, arguably, enforced, most of the time, in most places.
We are well aware that this perspective on the local policy process will not go uncontested by China scholars, many of whom have reported on ineffective (but mostly, inefficient) policy implementation in rural and semi-urban China (see above) in recent years. However, many of the conflicting viewpoints in the above quoted literature can be explained and, arguably, reconciled by making a precise distinction between effective and efficient policy implementation. To what extent public demands feed into the process of local policy implementation must certainly be explored more systematically in future research as effective policy implementation cannot be isolated from a political system’s overall responsiveness to these demands. But we argue that it is fair to say that effective policy implementation, though not necessarily the best implementation, is good enough to sustain the Chinese political system, at least for the time being.
Footnotes
Appendix
The fieldwork was conducted as part of a joint research project on County and Township Cadres as Strategic Groups in Contemporary Rural China, funded by the German Research Foundation (Deutsche Forschungsgemeinschaft) and the German Ministry of Education and Research. Data were gathered on four consecutive field trips, which comprised three research stays in Qingyuan county (2008–2010), two visits in Mizhi county (2008 and 2009), one single but longer stay in Dingnan county (2010), and one single and shorter stay in Nan’an county (2011). Altogether, our field research yielded 20 interviews in Fujian, 36 in Jiangxi, 47 in Shaanxi, 58 in Zhejiang (161 in total), mainly with government and party officials at the provincial, prefectural, county, township, and village levels, but also with village residents and Chinese scholars working in similar fields. Internal documents comprising policy guidelines, work reports, and evaluation regulations, and fiscal statistics, among others, were also collected and completed the pool of primary data.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) and the Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung, BMBF).