Deficits Pro forma and Substantive: Multilevel Policy Implementation in China’s Administrative Approval Intermediaries Reform

Abstract

Policy implementation deficits tend to emerge and expand as implementation moves downward through multilevel administration. This article uses China as an example and examines its policy implementation at both its provincial and prefectural levels under the administrative intermediaries reform policy. By analyzing data from 225 prefectures in China and their component provinces, we identify two types of implementation deficits: deficits pro forma and substantive deficits. We highlight three factors mitigating both types of deficits and their exacerbation from the provincial level to the prefectural level: self-imposed requirements, top-down pressure, and demonstration effects from upper levels. We find that while self-imposed requirements have no binding effect, rigid executive directives from upper levels have the strongest effect on deficits mitigation while demonstration effects also play an important role in deficit reduction. Our research contributes to the implementation literature by distinguishing between pro forma and substantive deficits and by focusing on the middle layer of government as well as the phenomenon of deficit exacerbation across layers.

Keywords

policy implementation substantive deficits administrative approval intermediaries reform China

Introduction

Implementation has always been at the core of every public service organization, as one of necessary steps in the policy process (Lasswell, 1956; Lynn, 1999; O’Toole, 2000; Sabatier, 1986; Wright, 1978). It carries out policy design bringing policy proposals to its expected fruition and makes revisions with the prescribed lessons from program evaluations. Case studies on policy programs (Allison & Zelikow, 1971; Ferman, 1990), the theoretical top-down/bottom-up discussions (Berman, 1977; Mazmanian & Sabatier, 1981), and the scientific-oriented hypotheses testing researches all dedicated to find out solutions for successful implementation and remedies for policy failures (Ahlers & Gunter, 2015; DeLeon & DeLeon, 2002; Goggin, 1990; O’Toole, 2000; Saetren, 2014).

Admittedly, some local discretions improve policy outcomes by increasing the local productivity and adapting the policy to local needs (Kwon, 2014; Meyers et al., 2007; Tummers & Bekkers, 2012). A number of policies worked largely as they were designed, some of which were pretty much successful (DeLeon & DeLeon, 2002; Kettl, 1993). But they are not always positive (Maynard-Moody et al., 1990). Faulty implementation is commonplace (Chan et al., 1995; Lin, 1996; Sabatier, 1986; Zhan et al., 2013). As Pressman and Wildavsky (1973) contended, national policies will have 50% erosion after moving through the government hierarchy. Implementation theory has long noted that downward policy implementation increases the probability of deficits (Pressman & Wildavsky, 1973). In this sense, implementation deficits preserve an enduring importance on policy execution studies.

Some recent discussions tended to bring implementation topics in a context of collaborative governance with multiple players and networks involved. Horizontal interorganizational relationships such as attitude toward collaboration, the extent of mutual trust, and collaborative roles engagement are seen as essential to implementation system (Hill & Hupe, 2002; Meier & O’Toole, 2001). The classic school of top-down/bottom-up controversy remains a structured way to understand implementation deficits, especially in multilevel context. Top-downers assert that central authority is capable of guaranteeing effective policy implementation by ensuring adequate policy design (O’Toole et al., 2012), clear and consistent objectives and means (Matland, 1995; Van Meter & Van Horn, 1975), legal mechanisms (Pressman & Wildavsky, 1973), committed and skillful implementers, and resources (Siciliano et al., 2017; Tummers & Rocco, 2015). Some bottom-uppers claim that street-level bureaucrats often get in the way of successful implementation by utilizing their discretion, distorting polices and directing them to their own ends (DeLeon & DeLeon, 2002; Lipsky, 1971; Sabatier, 1986; Zhan et al., 2013).

Situations in the U.S. context and places outside are somewhat similar. Strategic local implementers coped with constraints from their different levels of administrative superiors. Local implementers unlikely conform to “either the spirit, expectations, rules, regulations or program components” (Odden, 1991). Their broad discretion facilitated local actors to act in their preferred manners despite performance pressures they faced (Farooqi & Forbes, 2020). Extant literature mainly focused on the impact factors of local bureaucrats’ behaviors such as bureaucratic imperatives set by organizational machinery (Lynn et al., 2001), local knowledge and attitudes (Spillane et al., 2002), workloads, other pressures (May & Winter, 2007), and so forth. Few studies addressed the strategies how local implementers respond to these factors in a multilevel context.

Middle-Level Implementation Deficits

Policy deficits are usually considered as a flaw in policy goal fulfillment (Ahlers & Schubert, 2014; Sabatier & Mazmanian, 1979). Policies can be eroded to a great extent in the process of multilevel implementation (N. Zhang & Rosenbloom, 2018). The probability of implementation deficits increases as the number of administrative layers increases (Elmore, 1979). Although some studies have examined the “layers impacts,” most have only paid attention to a single layer either at the top (Goggin et al., 1990) or at the bottom (Hill & Hupe, 2003; Hupe, 2014). The middle levels in particular have been ignored in the “level of analysis” of implementation deficits (Goggin et al., 1990). Middle levels are administrative layers between the top and the bottom levels. They can also be distinguished between the upper and lower layers. For instance, middle level in the U.S. context may refer to the state level, between the federal and the street level (Farooqi & Forbes, 2020). Detailed studies on middle levels are essential for understanding and designing effective implementation process. How would subnational implementers behave differently when confronting the pressure of upper-level executive orders in tension with the local bureaucratic interests to protect their “little money lockers” (off-budgetary funds; Gong, 2006)? Would the upper layer mandates help alleviate deficits? The roles of middle levels, the mechanisms they transform central commands, and how lower layer implementers strategically respond to upper layer’s imperatives require additional analyses.

In a multilevel context, particularly for policies conflicting local bureaucratic interests, it is likely that central policies have already been distorted before they reach the bottom as passing through the middle levels, as middle levels may bend policies in their preferred manners. Within the middle levels, speaking of a two-layer structure, the upper and lower layer, both play multiple roles. The upper layer is involved in policy formulation and design whereas the lower layer is charged with direct implementation. As policy implementers, subnational administrators at middle levels have to implement the central policy and to demonstrate their performance to the central authority. For instance, local implementers have to execute the policies so as to deal with their performance evaluation. However, they are also policy makers because successful implementation requires adaptation and adjustment of the central policy to local circumstances. Therefore, both the upper layer’s performance and its adaptations will have an impact on the policy implementation at the lower layer. Besides the extant theoretical and descriptive studies (May, 2015; O’Toole, 2000; Sabatier, 1986; Van Meter & Van Horn, 1975), few studies have quantitatively examined how a central policy is distorted in the middle levels.

The lower layer of middles levels plays multiple roles as well. They act as policy implementers, adjusters, and as strategic self-interest defenders. Unlike street-level bureaucrats at the lowest level, who are loosely bounded to bureaucratic hierarchical discipline (O’Brien & Li, 1999; Thomann et al., 2018), the majority of public officials at middle levels are strictly subject to the bureaucratic authority and discipline (Nielsen & Moynihan, 2017). For instance, Soss et al.’s (2011) study depicted that welfare managers in Florida have to deal with the pressures from policy mandates.

As policy implementers, they have to comply with upper layer policies and fulfill their responsibilities (Barrett & Fudge, 1981; McLaughlin, 1987; O’Brien & Li, 1999). However, as rational actors and even as policy targets sometimes, they may seek to protect their vested interests (Caiden, 2017; B. Y. J. Zhang, 2011) and intentionally distort policies in practice (May & Winter, 2007; Smith, 2009). In this sense, managerial influences on frontline workers are largely weak as demonstrated in Riccucci’s (2005) research on welfare reform.

Thus, the lower layer constantly confronts a dilemma of job fulfillment and self-interests protection. By constantly striking a balance between the two interests, they often strategically adjust their behaviors. Bureaucratic theory maintains that lower-level administrators, whether motivated by personnel promotion opportunities (L. A. Zhou, 2007) or subject to disciplinary systems (Christensen & Opstrup, 2018; Rothstein, 2015), have to accomplish their obligations and demonstrate their loyalties to the upper level. However, in practice, they sometimes ignore the pressure from upper levels and produce implementation deficits (Huberman, 1973). Smit (2005) showed that local teachers secretly distorted educational polices and left a wide gap between rhetoric and practice. Few studies explore in detail under what conditions and to what extent will the local implementers be loyal to central policies, comply with the upper-level mandates and exercise self-discipline: under as opposed to the conditions in which they will incline to their vested interests while sacrificing the policy objectives. It remains unknown whether self-discipline would facilitate deficits mitigation.

Within the multilevel framework, middle-level implementers’ strategic responses call for careful analysis on the deficits patterns. Existent studies largely take implementation deficits at a whole (Ahlers & Schubert, 2014; Sabatier & Mazmanian, 1979; B. Y. J. Zhang, 2011). This holistic view may well conceal the richness and thickness of implementation deficits produced by different factors at different levels. In hierarchical, particularly authoritarian systems, subnational implementers rarely resist central polices overtly (O’Brien & Li, 1999). This may illustrate why central policies often partially succeed and also indicate a necessity for further conceptualization and study of implementation deficits.

Partial fulfillment of policy goals may exist with respect to both speed and allegiance because efficiency and goal achievement have been regarded as two essential elements for successful policy implementation (DeLeon & DeLeon, 2002; Matland, 1995; O’Toole, 2000). Some implementers may respond quickly but bend policies to their own interest. Quick and symbolic responses have long been found in implementing educational policies in a multilayered system (McLaughlin, 1987): or even worse, intentional delay and sabotage policies (Chu & Huang, 2013; X. Zhou, 2010).

Neither of these situations is perfectly effective. Sacrifice in speed or in quality thus should be categorized as different types of implementation deficits. For instance, deficits in timeliness are pro forma whereas deficits in appropriate policy application, such as service delivery, are substantive. Few studies quantitatively measured implementation deficits. Moreover, extant studies usually took implementation deficits as a whole, few of which distinguish deficits into their subcategories. The benefits of categorizing deficits into pro forma and substantive ones may help explore implementation deficits in a more subtle and detailed way, and to uncover local strategies in enforcing central policies.

The remainder of this article is organized as follows: First, it discusses policy implementation structure and deficits in China. Second, it introduces our case, China’s administrative approval intermediaries reform and explains its fitness for theory building. Third, it constructs a theoretical framework and puts forward a series of research hypotheses. Fourth, data, method, and results are presented. Finally, the article discusses its findings and presents conclusions.

Policy Implementation Structure and Deficits in China

With a multilevel and homogeneous local administrative structure, China is suitable for examining implementation deficits at middle levels. A prevailing conceptualization of China’s policy process is centralized decision-making with decentralized policy implementation (Harding, 1981; Lieberthal & Oksenberg, 1988; Manion, 1991; Qian & Weingast, 1997; X. Zhou, 2010). Such a structure has local institutional factors such as legal mechanisms, bureaucratic structure, and similar contextual factors well controlled. At the same time, an extensive flexibility is left for middle-level policy makers and implementers to adapt central policies to their own circumstances. This produces variation in interpretation of upper-level mandates, performance, and implementation strategies.

China preserves a five-level policy making and implementation structure—the central, provincial, prefectural, county, and the township levels (Figure 1). Its hierarchical implementation proceeds level by level downward. Normally, after the central authority issues its national policy, the provincial-level governments start to plot and release their provincial policies. Replicating the central policy, provincial-level governments usually have the flexibility on how to proceed in assigning the tasks to the next lower level, that is, the prefectural-level governments. Similarly, the prefectural governments issue the policies at the prefectural level, some of which set deadlines for their departments. The county-level governments may have their own policies as well. In the end, theoretically though not necessarily in practice, a central policy may develop into multi-policy subsystems with a scale of more than 30 provincial policies, 333 prefectural, and more than 2,800 county-level policies.

Figure 1.

The hierarchy of administrative levels in China.

The scope of such a variety increases the difficulties of regulation and supervision from the top. As Zhan et al. (2013) argued, implementation deficits in China have been an enduring problem. This administrative structure leaves subcentral government with significant discretion on how to achieve the central-level policy goals thereby exacerbating the complexity of policy implementation (N. Zhang & Rosenbloom, 2018). Extant literature discusses the impediments to successful policy implementation in China, such as policy design, strategic agency (Ahlers & Schubert, 2014), and the lack of clear directives and information (Sabatier & Mazmanian, 1979). As for policy design, some scholars attributed implementation failure to inadequate policy details. The lack of information may also hinder successful implementation (B. Y. J. Zhang, 2011). Because subcentral governments as strategic agents have distinct preferences over various policies, they may selectively implement central policies that suit their own needs (Cai, 2004; Chan et al., 1995; O’Brien & Li, 1999; Schubert & Ahlers, 2012). However, factors such as policy design and subordinate-level preferences are difficult to capture in a quantitative analysis, thus creating problems for assessing their real impact.

In this article, we purposively chose the implementation of a special type of policy as our key example: China’s administrative approval intermediaries reform. Such a reform has a very clear goal to begin with, which is to first release (i.e., identify and list) the number of administrative approval intermediary services in local areas. Such a policy goal is easy to measure and not prone to measurement errors. Furthermore, local preferences will be completely reflected in the speed of their release and the number of approval intermediary services that they release. There is little asymmetric information, when local governments release their own intermediaries list identifying the approval requirements. All these characteristics make this policy an ideal case to study the implementation deficits at middle levels.

China’s Administrative Approval Intermediaries Reform

The administrative approval intermediaries reform is an interesting case to study in the sense that the central policies are well defined and local implementers share the similar legal directives and incentive schemes. This largely controls factors such as “policy objectives” and “legal mechanisms.” The major objective of the policy was part of a comprehensive reform launched in 2001 to improve bureaucratic efficiency and efficacy by streamlining administrative approval procedures. The State Council has eliminated 2,012 administrative approval items through six rounds of reforms since 2002. Nevertheless, the central government soon realized that reforming inside the government system failed to enhance the administrative approval efficiency substantially (N. Zhang, 2016).

The administrative approval intermediaries were soon identified as a core hindrance. Administrative approval intermediaries are agencies that provide professional certified services such as technical reviews, appraisals, tests, consultations, required by approval administrative units for a variety of processes. For instance, an applicant who plans to apply for a building permit from a local jurisdiction may be required by the building department to submit some supporting materials such as an environmental impact appraisal report together with his or her application. The applicant may need to hire intermediary agencies to provide such materials due to their requirement for professional expertise. Such agencies are an example of intermediaries we focus on in this article. Because these intermediaries are central actors in administrative approval procedures, we call them “administrative approval intermediaries.”

Administrative approval intermediaries function as third and middle parties between the government and the applicants. Each level of government preserves certain administrative approval jurisdictions. The difference between administrative approval and administrative approval intermediaries is that the former is the jurisdictions belonging to bureaucratic departments, whereas the latter are the third party agencies providing intermediary services required by the administrative approval procedures. Again, taking the building permit, for example, the central-level building authority is in charge of approving and issuing building permits for construction projects above a certain scale. The provincial-level building authorities retain jurisdiction for construction projects of a middle scale. The prefectural-level building authorities have even lower jurisdictions. Those intermediaries providing environmental impact appraisal services for building permits needed by applicants in cities are usually local intermediaries of a small scale.

Administrative approval intermediaries were initially introduced by the Chinese government for the purpose of encouraging social forces to participate in regulation. The introduction of intermediary agencies to provide supporting services during the administrative approval process is to help transit the Chinese omnipotent state into a limited regulatory state. Currently, administrative intermediaries in China are a mix of private enterprises and quasi-official agencies. Benefiting from their close relationship with local governmental departments, quasi-official intermediaries either monopolize local markets or often break the rules and regulations by raising the market prices with prolonged service provision periods. Subnational governments may benefit from this malpractice as some intermediaries are funded by subnational governments so that the profits of the intermediaries can be transferred back to the governments. The collusion between those quasi-official intermediaries and subnational governments would make any reforms on the intermediaries against the interests of subnational governments. Figure 2 explains the relationship between applicants, quasi-official intermediaries, and subnational governments taking the prefectural governments as example. To cope with the collusion between intermediaries and local-level governments and cultivate healthy market competition, the central government launched a comprehensive reform on regulating administrative approval intermediaries on April 27, 2015.

Figure 2.

The relationship between applicants, quasi-official intermediaries, and prefectural-level governments.

In this research, we focus on the initial efforts of policy implementation of this reform, the release of intermediary services lists by local-level governments in their local markets. Because only by knowing what and how many intermediary services existed in local markets, the number and services provided by those quasi-official agencies could be pinpointed, and be transparent to the high level of authorities to take further regulation.

When launching this reform, the central government set a deadline for the national departments to release their intermediary services lists and announced that successive lower levels of governments should launch their own reforms. To implement the central policy, the provincial governments issued reform notices, specifying the specific requirements on the lists issuing, including the content of the central policy, provincial departments in charge and involved, some of which set deadlines for prefectural governments to release their intermediary lists. The prefectural governments subsequently released the reform notices at the prefectural level, some of which set deadlines for themselves to release the intermediary lists. A wide variation often exists in “how the same national policy is implanted at local levels” (Matland, 1995). Without clear requirements on deadline in the central policy, vast local disparities were exhibited on the speed the provincial governments released their reform notices whether the provincial governments set deadlines for their prefectural subordinates, whether the prefectural governments set deadlines for themselves, and how well the provincial governments perform in implementing the national policy. All these governmental levels and variations may influence the varieties of implementation deficits at prefectural level. We will specifically explore the policy implementation deficits from provincial to prefectural level in this research as the number of prefectural governments provide a suitable sample size. Moreover, although there are thousands of county- and township-level governments, we do not have county- and township-level data because this reform has seldom been launched below the prefectural level.

As mentioned above, this administrative approval intermediaries reform is a case of inner bureaucratic battles where the policy implementers are at the same time the policy targets because authorizing and protecting quasi-official administrative approval intermediaries may well be in the economic and administrative interests of local governments. Thus, the reform launched by the central government on intermediaries is largely against local implementers’ interests.

Such an unfavorable policy would inherently increase the possibility of resistance from subnational implementers via implementation deficits. Local implementers face a dilemma between complying with the upper levels’ requirements and protecting their vested interests by reducing the quality of implementation. The implementation problems are therefore more evident in addressing this seemingly intractable and contested problem (May, 2015). China’s multilevel context also provides us with a prime opportunity to observe what has happened in the middle layers and how the battle plays out with top-down pressures and local distortions.

Typology and Hypotheses: Deficits pro forma and Substantive Deficits

Objectives fulfillment has usually been deemed as a common way to judge whether a policy implementation is successful (Sabatier & Mazmanian, 1979; Talen, 1996; Winter, 2012). The degree to which the policy objectives are not successfully achieved indicate the extent of deficits, though the standard of judging how many deficits accumulated could be considered as implementation failure is mainly subjective.

In examining the implementation deficits in the administrative approval intermediaries reform, we find it necessary to differentiate between two different types of deficits (Figure 4). As mentioned above, speed and quality both matter to successful implementation. Theoretically, policy makers would always prefer quick and successful implementation, and ideal implementation should be quick and faithful. But this is not necessarily—or even typically—the actual situation. In reality, some implementers may respond fast but perfunctorily; others may act slow but more dutifully. With limited financial, staff, and informational resources, tradeoffs between implementation speed and quality may occur and vary among jurisdictions. Erosion in either way will result in implementation deficits. Yet, deficits from speed erosion and from quality erosion are different. We assume that implementation quality is preferred to speed for policy makers, who would prefer slower but loyal implementation rather than fast but perfunctory execution, if the quick as well as high quality option is not available and timeliness is not essential. Thus, as shown in Figure 2, we develop a deficits typology and distinguish two kinds of implementation deficits: deficits pro forma and substantive deficits; the former places emphasis on speed erosion and the latter on quality erosion.

Deficits Pro forma

We define deficits resulting from perfunctory enforcement as deficits pro forma. The responsibilities of policy implementers are to enforce policies even if their own interests might be undermined by the policies (N. Zhang & Rosenbloom, 2018). In our case of intermediaries reform, we consider time delays as implementation deficits pro forma. Subnational governments have been granted discretion and flexibilities to decide how fast they respond to the national policies (Lipsky, 1971). To satisfy their superiors and act like faithful implementers, subnational officials sometimes yearn to show enthusiasm (L. A. Zhou, 2007). We indeed observed that some provincial and prefectural governments soon released their lists of intermediaries while others were lagging behind. The time delay in the intermediaries reform probably reflects a wait-and-see strategy by subnational governments during policy implementation to test on the seriousness of the top-down reform. However, from a central government perspective, such time delay erodes the momentum passion and opportunity windows (Kingdon & Thurber, 1984) for the reform, thus a policy implementation deficit due to speed.

Substantive Deficits

Substantive deficits are the real undermining of the ultimate objectives of a policy. In our case, the substantive deficits are deeds by subnational governments that intentionally defeat the policy goal and prevent the central government gaining awareness of the actual local situations. One key objective of this reform is to investigate the exact and complete administrative approval intermediaries services in local markets. In particular, the focus of the reform is to determine what and how many intermediary services are monopolized by intermediaries that are attached to governmental departments. Such information provides an important basis for the central government to clear out intermediary services that are quasi-official and to push the reform to the next step.

However, this intention contradicts local departments’ self-interests such that they are unwilling to disclose the real situation accurately and completely. Subnational governments could potentially release a partial list in which intermediary services are incompletely identified. Services that are not listed will not be discontinued or reformed, which in the aggregate could seriously undermine the policy objective. In this sense, subnational governments produce substantive deficits that erode policy objectives in quality.

We analyze three generic factors as the impact factors of policy implementation deficits below, namely, self-imposed pressure, pressure from immediate upper-level governments, and demonstration effects from immediate upper-level governments too. These factors work for both types of implementation deficits. Figure 3 displays the deficits typology and the timeline of the reform.

Figure 3.

Core stages and the timeline of the policy implementation.

Figure 4.

Theoretical framework and typology.

Self-Imposed Pressure Hypotheses

We first assume that the most direct pulse on implementation deficits is from subnational governments themselves (Sabatier & Mazmanian, 1979). Implementation deficits could be alleviated if subnational governments take the central policy seriously and put pressure on their component departments. They would issue their reform notices quicker and better if they take the central policy more seriously, or at least they act as if they truly value the policy. Because the reform needs governmental departments to launch a thorough investigation in their specific fields, some subnational governments even set up deadlines for their departments to complete their tasks and provide feedback; however, not all governments have done so for their departments. Therefore, we consider such a deadline-setting behavior as self-imposed pressure on subnational governments, which should be able to reduce implementation deficits.

Hypothesis 1a (H1a): Self-imposed pressure will reduce subnational governments’ deficits pro forma.

Hypothesis 1b (H1b): Self-imposed pressure will reduce subnational governments’ substantive deficits.

Upper-Level Pressure Hypotheses

It is expected that pressures from the immediate upper-level government would lower policy implementation deficits as well because local implementers would value their superiors’ reaction due to China’s cadre responsibilities and evaluation systems (Ahlers & Schubert, 2014; O’Brien & Li, 1999). Executive directives from immediate upper-level governments will impose pressure on lower levels. In our case, pressure from immediate upper levels could arrive in two forms: one soft and the other hard. The soft pressure depends on how early a province releases its reform notice. The earlier the provincial reform notice is released, the stronger the pressure a province is imposing on their prefectural governments. Such pressure is soft in the sense that the provincial reform notice does not necessarily pose rigid requirements for when the prefectural governments should complete their tasks.

For the hard pressure from upper levels, we found that some provincial notices set a deadline for their prefectural governments, specifying the timelines for when prefectures should issue their lists. We assume that prefectural governments will have to work harder if the provincial governments set a deadline for them to finish their jobs, compared with a case when there was no such deadline. Hard pressure is presumed to have a greater impact on alleviating deficits than soft pressure. Furthermore, we assume that the hard pressure from intermediate upper level of government will have a greater impact than that from prefectures than self-imposed pressures. Interviews with prefectural governments disclosed that deadlines-setting by the upper level are rigid executive directives. If not compliance with the deadlines, the lower levels may suffer from the deduction on performance evaluation punishment, and even from the potential chances for promotion. However, pressure from a level of government on themselves is relatively flexible because they would not embarrass themselves either for their limited capabilities or unwillingness to emphasize effective implementation.

Hypothesis 2a (H2a): Provincial soft pressure is negatively related to prefectural deficits pro forma.

Hypothesis 2b (H2b): Provincial soft pressure is negatively related to prefectural substantive deficits.

Hypothesis 3a (H3a): Provincial hard pressure is negatively related to prefectural deficits pro forma, and will have a greater impact than that from a prefectural level itself on lowering the probability of deficits pro forma at the prefectural level.

Hypothesis 3b (H3b): Provincial hard pressure is negatively related to prefectural substantive deficits, and will have a greater impact than that from prefectural level itself on lowering the probability of substantive deficits at the prefectural level.

Upper-Level Demonstration Effect Hypotheses

Although most of provinces soon issued their notices and announced that they would follow the central government’s policy strictly, distinct disparity existed on their lists issuing dates. According to Mazmanian and Sabatier’s (1981, pp. 20–21) seminal definition on policy implementation and the fact that local implementers care about their superiors’ reactions (Ahlers & Schubert, 2014; O’Brien & Li, 1999), we assume that demonstration effects by provincial governments do matter. The lower levels would decide how to proceed with implementation by looking at how their immediate upper-level superiors treat the reforming tasks. The central policy of intermediary reform required all lower levels of governments to investigate intermediary services under their jurisdictions. Prefectural governments would play close attention to the provincial governments’ reform progress. They are more likely to be committed implementers if they observe that their immediate upper levels take their own jobs seriously and set a good example for them. If lower levels detect any perfunctory attitude, they are most likely to follow this style.

By demonstration effects, we refer to the impacts of both the implementation speed and the quality of the immediate upper level to the lower level. The faster the immediate upper-level governments issue their lists, the stronger the positive demonstration effects, and the lower the probability that lower levels will have implementation deficits. Thus, we first test the impact of the speed of immediate upper level’s implementation on their lower levels’ implementation deficits. We assume that the slower the immediate upper level, the larger the implementation deficits at the lower levels. Furthermore, perfunctory superiors are less likely to have committed subordinates. In our case, we hypothesize that large provincial substantive deficits will lead to large prefectural implementation deficits.

Hypothesis 4a (H4a): Provincial deficits pro forma are positively correlated with prefectural deficits pro forma.

Hypothesis 4b (H4b): Provincial deficits pro forma are positively correlated with prefectural substantive deficits.

Hypothesis 5a (H5a): Provincial substantive deficits are positively correlated with prefectural deficits pro forma.

Hypothesis 5b (H5b): Provincial substantive deficits are positively correlated with prefectural substantive deficits.

Finally, we assume that the deficits pro forma at one level of government have a positive relation with its substantive deficits because enforcement delay implies a disinterested or resistant inclination which would further lead to the quality erosion and substantive deficits.

Hypothesis 6 (H6): Deficits pro forma of a level of governments are positively correlated with their substantive deficits.

Data and Method

This study was conducted from December 2017 to March 2019. We collected and extracted data from various official documents on the intermediaries reform at the national, provincial, and prefectural government levels from April 2015 to May 2017. The time frame spans from when the central policy was issued to when prefectural responses were stabilized because few prefectures issued their lists after March 2017. Thus, the reform gradually stopped at local levels though the central government did not announce to terminate the reform. Thus, we believed that data collected during this time frame could potentially reflect the actual local implementation situation. Our variables of interest include the time when these governments issued their intermediaries reform notices, the time they released their intermediaries lists, the number of intermediary services in the lists, and whether governments set a deadline for their constitutive departments to follow. Overall, these official documents cover one document from the State Council, 31 provincial documents, and 225 prefectural documents,¹ from which we derived our key dependent and independents variables.²

Dependent Variables

Our dependent variables are the two types of implementation deficits at the prefectural level. On prefectural deficits pro forma, prefectural governments were supposed to launch reforms after the State Council launched the national reform on April 27, 2015. We thus use the time difference between the releasing of intermediary lists by the prefectural government and the provincial government releasing to measure deficits pro forma. We assume that greater the time difference, the later that the prefectural government issued its intermediary lists following its provincial government’s lead, thus bigger the deficits pro forma.

We use the difference between the number of intermediary services listed by the prefectural government and that by the provincial government to measure the substantive deficits. For instance, the provision of environmental impact appraisal as an intermediary service exists at all local levels: the central, the provincial, the prefectural level, and so forth. Thus, the total number of all kinds of intermediary services at each level should not be significantly different from one to another, while allowing for small disparities. We thus posit that bigger the difference between the number of intermediary services a prefectural and its provincial government released, respectively, greater the substantive deficits this prefectural government has.

To showcase one of our dependent variables, we plot the substantive deficits at both the provincial and prefectural levels (Figure 5). It is evident that the central reform policy has been eroded as it goes downstream from the provincial level to the prefectural level. An essential deficit gap exists between these two levels. This prompts us to further explore the layer impacts: What are the factors at the provincial and prefectural levels that might explain the uptick of implementation deficits?

Figure 5.

Substantive implementation deficits.

Independent Variables

As we have explained above, there are three sets of independent variables. First is self-imposed pressure, and we measure it using a dummy variable that indicates whether the prefectural government has set up a deadline for their subordinates to release the intermediary services. Second is soft and hard pressure from provincial governments. We measure provincial soft pressure as the time lag between the provincial notice and the national notice. We assume that earlier the provincial reform notice is released, the stronger the pressure this province is tending to impose on their prefectural governments. Provincial hard pressure is measured by its deadline-setting as well, a similar measurement to that at the prefectural level.

Finally, on demonstration effects, we use provincial-level deficits pro forma and substantive deficits as the corresponding measurements. Their exact definitions are parallel to those at the prefectural level. Specifically, provincial deficit pro forma is measured by the time lag between provincial lists and the national notice, and provincial substantive deficits as the difference between the number of intermediary services on provincial lists and that on the national list.

Control Variables

We further augment the intermediaries reform data with prefectural control variables from China’s Cities Statistical Yearbook. To avoid the problem of reverse causality, all data on control variables are from 2014, 1 year before China’s approval intermediaries reform started. These variables cover the economic (Wang & Ap, 2013), employment (May & Winter, 2007), and financial conditions of each prefecture (Brynard, 2009; Huang et al., 2017) that may impact their policy implementation.

Specifically, on economic conditions, we include variables such as GDP per capita and the industrial structure (secondary industrial and tertiary industrial shares); on employment conditions, we have employment data for the private and public sectors to measure the market force on intermediaries and the sufficiency of official staff input in city-level reforms; finally, on financial conditions and the size of the government, we have bank loans and government revenues at the prefectural level. For control variables, we postulate that a prefecture with relative good conditions would show smaller implementation deficits. The descriptive statistics are summarized in Table 1.

Table 1.

Descriptive Statistics.

Variables		M	SD	Minimum	Maximum	Observations
Dependent variables	Prefectural substantive deficits (no.)	95.85	105.43	−163	344	225
Dependent variables	Prefectural deficits pro forma (no. days)	398.6	233.47	−252	715	225
Independent variables	Self-imposed pressure: Prefectural hard pressure	0.54	0.5	0	1	225
	Upper-level Pressure A: Provincial hard pressure	0.26	0.44	0	1	225
	Upper-level Pressure B: Provincial soft pressure (no. days)	357.8	88.26	0	468	225
	Demonstration effects A: Provincial deficits pro forma (no. days)	316.6	136.88	16	581	225
	Demonstration effects B: Provincial substantive deficits (no.)	243.5	102.97	56	374	225
Control variables	GDP per capita (in 1,000 Yuan)	52.61	49.67	9.27	482	225
	Industrial share (%)	48.88	9.74	16.1	75.5	225
	Service share (%)	38.36	8.11	19.8	74.9	225
	Private sector employment (per working people)	1.02	0.57	0.05	3.77	225
	Public sector employment (per working people)	0.31	0.11	0.06	0.67	225
	Bank loans (in 1,000 Yuan)	51.17	69.49	6.62	682.45	225
	Government revenue (per capita, in 1,000 Yuan)	4.62	5.72	0.67	62.69	225

Note. GDP = gross domestic product.

Model Estimation

We use regression methods to help understand reasons behind the two types of implementation deficits, while controlling for various differences among prefectures. The model specification for implementation deficits pro forma, Tlag_i, is as follows:

\begin{array}{l} T l a g_{i} = α + β_{1} d d l_{i} + β_{2} N o t i c e_{j} + β_{3} d d l_{j} + \\ β_{4} T l a g_{j} + β_{5} d n u m_{j} + Z^{'} δ + ε_{i} \end{array}

(1)

where i is a prefecture, j is a province, and ddl_i is whether the prefectural government sets up a deadline for its consisted departments to complete the reform, that is, self-imposed pressure. Notice_j is the provincial soft pressure, indicated by how soon a provincial government publishes its reform notice following the national notice. And, ddl_j is similar to ddl_i, capturing the hard pressure from the provincial government. As for demonstration effects, Tlag_j is the provincial deficits pro forma, whereas dnum_j is the substantive deficits at the provincial level. Finally, Z stands for prefectural-level controls, ε is the error term, and others are parameters to be estimated.

For the substantive implementation deficits (dnum_i), we use the same model specification as Equation 1, except that we further add prefectural deficits pro forma as one additional explanatory variable. We postulate that the deficits pro forma affect the substantive deficits, not vice versa. Deficits pro forma reflect the reluctance of policy implementers to reform that is logically prior to the occurrence of substantive deficits. Here, it is assumed that the lower levels can identify intermediary services as required because they have the best knowledge on local markets than their superiors. After including deficits pro forma into the substantive deficits regression, we break the impact of all other variables on substantive deficits into two parts: one direct and the other indirect through deficits pro forma (i.e., the mediator). In the robustness checks, we further test this assumption.

One feature to note about Equation 1 is that since it uses prefectural deficits pro forma as the dependent variable, including the demonstration effect (provincial deficits pro forma) as an independent variable allows us to study the problem of implementation deficits exacerbation at the same time. This is because deficits exacerbation is essentially the difference between prefectural deficits pro forma and provincial deficits pro forma. Using such a difference term as the dependent variable will produce the same coefficients as Equation 1 does (except for β₄):

\begin{array}{l} T l a g_{i} - T l a g_{j} = α + β_{1} d d l_{i} + β_{2} N o t i c e_{j} + β_{3} d d l_{j} + \\ (β_{4} - 1) T l a g_{j} + β_{5} d n u m_{j} + Z^{'} δ + ε_{i} \end{array}

(2)

The same logic applies to the regression of substantive deficits.

A second feature to note about Equation 1 is that we choose not to include provincial fixed effects to begin with. Including those fixed effects would automatically drop interesting provincial-level variables, such as Notice_j, ddl_j and Tlag_j. Because these variables offer important insights into what provincial-level factors explain the implementation deficits at the prefectural level, we choose to include them in our regression instead of provincial fixed effects. Later, we show that including these meaningful variables gives us similar model goodness-of-fit to the specification using provincial fixed effects, an interesting finding that supports our choice of explanatory factors. Furthermore, we do not employ multilevel models here because we fail to find any random slopes when estimating a hierarchy linear model, while including only random intercepts is inferior to the fixed effects model used.

Results

Table 2 summarizes the regression results for both deficits pro forma and substantive deficits. There are two versions of results using ordinary least squares (OLS): with or without provincial-level explanatory variables. Based on the R² statistics, adding provincial variables improves the goodness-of-fit of the model significantly. Thus, we mainly focus the results including those provincial variables.

Table 2.

Regression Results on Two Types of Implementation Deficits.

	Deficits pro forma			Substantive deficits
Variables	OLS_1	OLS_2	FE	OLS_1	OLS_2	FE	OLS_3
Prefectural factors
Deficits pro forma				0.111*** (0.028)	0.079*** (0.013)	0.091*** (0.014)
Hard pressure (ddl_i)	12.953(30.783)	−12.869(27.927)	−6.681(26.662)	8.216(12.624)	−5.441(6.933)	−8.211(7.245)	−6.459(7.261)
Provincial factors
Soft pressure (Notice_j)		−0.630*** (0.147)			0.119*** (0.042)		0.069(0.044)
Hard pressure (ddl_j)		−138.682*** (34.770)			−28.986*** (10.294)		−39.958*** (10.485)
Deficits pro forma (Tlag_j)		0.266* (0.150)			0.038(0.030)		0.059** (0.030)
Substantive deficits (dnumj)		−0.264(0.174)			−0.869*** (0.045)		−0.890*** (0.045)
Controls
GDP per capita	−0.026(1.052)	−0.595(0.770)	−1.230* (0.669)	−0.585(0.595)	0.242(0.296)	0.090(0.231)	0.195(0.303)
Industrial share	−7.825*** (1.837)	−6.816*** (1.810)	−5.263** (2.532)	−3.583*** (0.985)	−1.463*** (0.534)	−2.122*** (0.685)	−2.002*** (0.535)
Service share	−12.301*** (3.042)	−10.737*** (2.731)	−6.124* (3.387)	−5.670*** (1.470)	−1.226* (0.683)	−2.924*** (0.775)	−2.075*** (0.651)
Private sector employment	−10.954(29.580)	−40.842(31.022)	−17.014(27.149)	18.381(11.816)	−1.503(5.864)	2.334(6.403)	−4.734(6.831)
Public sector employment	2.280(1.855)	2.818(1.743)	−1.845(1.847)	−2.809*** (0.708)	−0.855** (0.354)	−0.749* (0.422)	−0.632* (0.376)
Bank loans	−0.275(0.585)	−0.099(0.510)	−0.625(0.403)	0.372(0.280)	0.075(0.134)	0.252(0.154)	0.067(0.133)
Government revenue	7.408(9.703)	7.126(6.900)	14.574** (5.854)	0.864(5.248)	−4.053(2.737)	−4.314* (2.373)	−3.489(2.703)
Provincial FE	No	No	Yes	No	No	Yes	No
Observations	225	225	225	225	225	225	225
R ²	.128	.267	.460	.190	.785	.810	.762

Note. Robust standard errors in parentheses. OLS = ordinary least squares; FE = fixed effects; GDP = gross domestic product.

p < .1. **p < .05. ***p < .01.

In terms of deficits pro forma, our results (OLS_2) show an insignificant correlation between deadline-setting from the prefectural governments and the deficits pro forma at prefectural level. This means that self-imposed pressure does not expedite local implementers’ responses to central polices, and indicates that hard pressure from one level of governments per se may not be binding on themselves. Our data reveal that a large number of prefectural governments have failed to meet the deadlines they set for themselves.

As for the pressures from the immediate upper level, the results show that provincial soft pressure is significantly negatively correlated with prefectural deficits pro forma. This means that the stronger the soft pressure from the immediate upper level, the faster the lower levels would finish their tasks, smaller deficits pro forma. Similarly, a significantly negative correlation existed between deadline-setting by provincial government and prefectural deficits pro forma. This means that stronger the provincial hard pressure, the faster the prefectural government would accomplish their jobs. A much larger coefficient of the hard pressure than the soft pressure on prefectural deficits pro forma indicates that hard pressure from above would remarkably advance lower level’s enforcement efficiency.

Considering the demonstration effects, the results confirm that the immediate upper level have certain demonstration effects on the lower levels. First, the provincial deficits pro forma have a marginally significant positive correlation with prefectural deficits pro forma. This indicates that laggard provincial implementers will largely be followed by laggard prefectural executors. Second, however, provincial substantive deficits are not significantly correlated with deficits pro forma at the prefectural level. This means that the implementation quality of the immediate upper level does not affect lower levels’ speed. In short, the demonstration effects on lower levels’ deficits pro forma tend to result from the same type of deficits, while not from something else.

As for the substantive deficits, similar to the deficits pro forma, the correlation between deadline-setting and substantive deficits at prefectural level is not significant. This indicates that pressure from a level of government on itself does not help improve implementation quality. In contrast, significant negative correlation existed between provincial deadline-setting and substantive deficits, which means that hard pressures from the immediate upper level may help greatly ease the substantive deficits and contribute to higher-quality implementation.

Different from what is seen in prefectural deficits pro forma, provincial soft pressure is significantly positively correlated with prefectural substantive deficits. This means that the stronger the soft pressure from the provincial government, the bigger the prefectural substantive deficits are. Such a counterintuitive result is due to the fact that deficits pro forma as a mediator have absorbed the negative correlation between provincial soft pressure and prefectural substantive deficits; overall, there is no significant correlation between these two variables.

The provincial demonstration effects on prefectural substantive deficits display an interesting picture. The provincial deficits pro forma are not significantly correlated with prefectural substantive deficits in any direct way, similar to the story that provincial substantive deficits are not significantly correlated with prefectural deficits pro forma. However, the provincial deficits pro forma can still impact prefectural substantive deficits indirectly through prefectural deficits pro forma. On the other hand, provincial substantive deficits are significantly negatively correlated with the prefectural substantive deficits. This means that provincial substantive deficits seem to counteract prefectural substantive deficits, opposite to what we have found regarding the provincial demonstration effects on prefectural deficits pro forma. A better explanation here is that prefectural implementers perform poorly no matter how provincial superiors have behaved. When provincial substantive deficits are small, prefectural substantive deficits are big. When provincial substantive deficits are big, prefectural substantive deficits are big as well. The latter is simply because if provinces fail to implement substantially, it leaves little room for prefectures to have serious substantive deficits as there is nothing or little for them to implement.

Finally, we find a significant positive correlation between prefectural deficits pro forma and substantive deficits. This supports our hypothesis that a slower implementer is most likely a perfunctory actor.

With respect to our control variables, prefectures with higher secondary industrial and tertiary industrial shares see smaller deficits pro forma and substantive deficits, which are consistent with our expectations. Furthermore, prefectures with higher public sector employment also see smaller substantive deficits, likely due to a higher level of official staff input. Interestingly, GDP per capita has no correlation with implementation. This is contrary to prevailing perspectives that the lack of resources will hinder full enforcement (Xu, 2011). It suggests that the level of economic development has little to do with successful implementation if the policy is unwelcomed by implementers.

To test the robustness of our regression results, we further run provincial fixed effects regression while dropping those variables that only vary at the provincial level. Judging from the R² statistics, it is interesting to find out that the pressure from and the demonstration effects at the provincial level capture the majority of provincial influences on the prefectural implementation deficits. Admittedly, it is more difficult to predict the prefectural deficits pro forma than the substantive deficits because the exact time when a prefectural government releases its intermediary list is simply more random than the number of intermediaries it will report.

We conduct a second robustness check by dropping deficits pro forma in the substantive deficits regression. As a result, no mediator is included in the regression now, and the coefficients show the total marginal effect of variables of interest. Except for the provincial soft pressure and deficits pro forma as noted above, all other variables tend to have very similar coefficients and results. However, dropping deficits pro forma in the regression ignores the positive correlation between deficits pro forma and substantive deficits, even after controlling for all the other variables in the regression. There are likely other omitted variables that impact both deficits pro forma and substantive deficits.

To control for the positive correlation between deficits pro forma and substantive deficits, we employ a method called seemingly unrelated regression (SUR) to estimate the deficits pro forma regression and the substantive deficits regression together. The SUR results clearly indicate a significant and positive correlation between the two deficits. Nevertheless, such a positive correlation disappears once including deficits pro forma in the substantive deficits regression, and then estimating the two regressions together. The SUR results are very similar to the OLS counterparts, further suggesting that our OLS estimation is robust.

Discussion

By examining how provincial and prefectural governments influence the policy implementation at the prefectural level, we discuss why a central policy has been eroded at a certain administrative layer as it moves downstream from provinces to prefectures. Our results show that hard pressure like deadline-setting by a level of governments does not have binding force on themselves. Nevertheless, hard pressure from the immediate upper level would significantly contribute to deficits alleviation, both on deficits pro forma and substantive deficits. The immediate upper level is closest to the implementer. Its executive capabilities specified by the mandates of the policy have direct pressures on the implementer. Threats on potential promotion, reduction on future budget, deduction on performance evaluation are all among the resources that the upper level of government has to force obedience from the lower level. A more interesting finding is that hard pressure from the immediate upper level has a much greater effect on deficits pro forma than on substantive deficits. This discloses a tactic that subnational implementers use in response to unfavorable policies. They would at least behave like faster rather than committed implementers to display disingenuous seriousness and decisiveness when they do not welcome the policies. To put it differently, it is difficult to guarantee high quality implementation even when the upper-level authorities impose hard pressure on the lower levels. This finding may enrich our understanding on principle agency theory in terms of the strategies how agents respond to their principles. Given the information asymmetry, confronting hard pressure, to alleviate deficits pro forma by implementing the policy faster which is relatively easier for the principle to monitor may cover the agent from not challenge the deficits substantive, which is difficult to monitor from top.

The demonstration effects reveal that deficits pro forma from the immediate upper level have a direct impact on the deficits pro forma at the lower level. If the immediate upper level responds to the central policies slowly, the lower level would be influenced by this and act slowly as well. This effect also translates to the lower level’s substantive deficits, though in a weakened way. Slower lower levels’ implementers typically distort the reform policy and contribute to poor quality implementation.

Substantive deficits from the immediate upper level have a much greater impact on the lower levels’ substantive deficits. The negative correlation between upper and lower levels’ substantive deficits reflects again the images of poor prefectural implementers. It suggests that the central policy has encountered essential obstacles at prefectural levels. The policy implementation failed largely when it went downstream to prefectural levels. No matter how well the provincial governments finished the jobs, the prefectural implementation enforced the policies poorly. This suggests that demonstration effects only work in a symbolic way. The speed of the immediate upper level’s enforcement has positive demonstration effects on the lower level; whereas the implementation quality from the immediate upper level has no effect on the lower level. This portrays the image of local implementers as choosing to quickly respond to the central policies to show their attitude and as decisive implementers so as to cover up their poor implementing qualities.

Moreover, subnational implementers in industrialized and developed areas are in general effective and committed. This is probably because these areas preserve a higher level of marketization and subnational governments intervene the market to a lesser degree. These areas have mostly gone through several rounds of reforms since China’s opening-up and have relatively deeper understanding and a better starting point on the central reforms. Yet, the degree of development has a greater impact on easing local deficits pro forma than on substantive deficits. After all, this reform runs counter to local implementers’ interests so that even those reform pioneers in developed areas chose to resist the central policy.

Conclusion

In the context of implementing an unfavorable central policy, subnational implementers serve as both policy executors and targets in this reform, especially when the central policy harms their interests. In the course of policy implementation, subnational implementers have to constantly strike a balance between fulfilling responsibilities and self-interested protection. By distinguishing implementation deficits as deficits pro forma and substantive deficits, the behavior of strategic local implementers is uncovered. Local implementers are able to lower deficits pro forma by behaving as decisive executors and expressing their commitment to the upper-level authorities. In the meantime, they can do little to alleviate substantive deficits so as to minimize the loss of their vested interests. In the process, the pressure from one level of governments has little binding force on themselves, whereas the pressure from the immediate upper-level authorities exhibit a greater effect on the lower levels in different ways. Hard pressure from the immediate upper level may greatly decrease lower levels’ deficits, whereas soft pressure functions on easing substantive deficits through deficits pro forma alleviation. Demonstration effects from the immediate upper level merely work on deficits pro forma in a weakened way. In our studied case, the policy implementation of intermediary’s reform failed when it went two layers from the central government down to the prefectural level. This captures the implementation process in a more accurate way by portraying central authority’s penetration when confronting strong local resistance. Beside factors we discussed, the lack of any serious inspections and evaluations to encourage local compliance may be another reason to explain these deficits. We discussed the effects of self-imposed, that of upper level. Besides, lower level’s effects may also play a role, as one level of government would probably consider the actual situation of lower level such as capabilities, financial resources in policy making process to avoid unrealistic policy design. This deserves future studies with more micro-level data.

Footnotes

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: National Natural Science Foundation of China (Grant No. 72004223). This research is also supported by the Fundamental Research Funds for the Central Universities, and the Research Funds of Renmin University of China.

ORCID iD

Nandiyang Zhang

Notes

Author Biographies

Nandiyang Zhang is an associate professor in the School of Public Administration and Policy, Renmin University of China. Her current research interests include administrative reforms, public administration, transitions in contemporary China, and China’s local governance.

David H. Rosenbloom is distinguished professor of Public Administration & Editor-in-Chief, Routledge Public Administration and Public Policy Series, Department of Public Administration and Policy, School of Public Affairs at American University (Washington, DC).

Changgui Dong is an associate professor in the School of Public Administration and Policy, Renmin University of China. His research focuses on public policy, China’s governance, and quantitative methods.

References

Ahlers

A. L.

Gunter

(2015). Effective policy implementation in China’s local state. Modern China, 41, 372–405.

Ahlers

A. L.

Schubert

(2014). Effective policy implementation in China’s local state. Modern China, 41(4), 372–405. https://doi.org/10.1177/0097700413519563

Allison

G. T.

Zelikow

(1971). Essence of decision: Explaining the Cuban missile crisis (Vol. 327). Little, Brown Boston.

Barrett

Fudge

(1981). Policy and action: Essays on the implementation of public policy (Vol. 756). Routledge.

Berman

(1977). The study of macro-and micro-implementation. Public Policy, 26(2), 157–184.

Brynard

(2009). Mapping the factors that influence policy implementation. https://repository.up.ac.za/handle/2263/12231#:~:text=The%20key%20to%20implementation%20performance,and%20administrative%20role%2Dplayers%20concerned

Cai

(2004). Irresponsible state: Local cadres and image-building in China. Journal of Communist Studies and Transition Politics, 20(4), 20–41.

Caiden

G. E.

(2017). The role of the public bureaucracy in policy implementation in five ASEAN countries. Cambridge University Press.

Chan

H. S.

Wong

K-k.

Cheung

Man-keung Lo

(1995). The implementation gap in environmental management in China: The case of Guangzhou, Zhengzhou, and Nanjing. Public Administration Review, 55, 333–340.

10.

Christensen

J. G.

Opstrup

(2018). Bureaucratic dilemmas: Civil servants between political responsiveness and normative constraints. Governance, 31(3), 481–498.

11.

Chu

Huang

(2013). Policy implementation: Implement and policy choice. Journal of Jishou University, 34(4), 63–67.

12.

DeLeon

(2002). What ever happened to policy implementation? An alternative approach. Journal of Public Administration Research and Theory, 12(4), 467–492.

13.

Elmore

R. F.

(1979). Backward mapping: Implementation research and policy decisions. Political Science Quarterly, 94(4), 601–616.

14.

Farooqi

Forbes

(2020). Enacted discretion: Policy implementation, local government reform and education services in Pakistan. Public Management Review, 22(8), 1217–1239.

15.

Ferman

(1990). When failure is success: Implementation and Madisonian government. In Palumbo

D. J.

Calista

D. J.

, & Policy Studies Organization (Eds.), Implementation and the policy process: Opening up the black box (pp. 39–50). Greenwood Press.

16.

Goggin

M. L.

(1990). Implementation theory and practice: Toward a third generation. Scott Foresman.

17.

Goggin

M. L.

Bowman

A. O. M.

Lester

J. P.

O’Toole

L. J.

Jr. (1990). Implementation theory and practice: Toward a third generation. Scott Foresman.

18.

Gong

(2006). Corruption and local governance: The double identity of Chinese local governments in market reform. The Pacific Review, 19(1), 85–102.

19.

Harding

(1981). Organizing China: The problem of bureaucracy, 1949–1976. Stanford University Press.

20.

Hill

M. J.

Hupe

P. L.

(2002). Implementing public policy: Governance in theory and practice. Wiley Online Library.

21.

Hill

M. J.

Hupe

P. L.

(2003). The multi-layer problem in implementation research. Public Management Review, 5(4), 471–490. https://doi.org/10.1080/1471903032000178545

22.

Huang

L. C.

(2017). Hayek, local information, and commanding heights: Decentralizing state-owned enterprises in china. American Economic Review, 107(8), 2455–2478.

23.

Huberman

A. M.

(1973). Understanding change in education: An introduction. Experiments and Innovations in Education No. 4. https://eric.ed.gov/?id=ED082330

24.

Hupe

(2014). What happens on the ground: Persistent issues in implementation research. Public Policy and Administration, 29(2), 164–182.

25.

Kettl

D. F.

(1993). Searching for clues about public management: Slicing the onion different ways. Jossey-Bass.

26.

Kingdon

J. W.

Thurber

J. A.

(1984). Agendas, alternatives, and public policies (Vol. 45). Little, Brown.

27.

Kwon

(2014). Motivation, discretion, and corruption. Journal of Public Administration Research and Theory, 24(3), 765–794.

28.

Lasswell

H. D.

(1956). The decision process: Seven categories of functional analysis. Bureau of Governmental Research, College of Business and Public.

29.

Lieberthal

Oksenberg

(1988). Policy making in China: Leaders, structures, and processes. Princeton University Press.

30.

Lin

A. C.

(1996). When failure is better than success: Subverted, aborted, and non-implementation [Paper presentation]. The Annual Meeting of the American Political Science Association, August 1996; San Francisco.

31.

Lipsky

(1971). Street-level bureaucracy and the analysis of urban reform. Urban Affairs Quarterly, 6(4), 391–409.

32.

Lynn

L. E.

Jr. (1999). A place at the table: Policy analysis, its postpositive critics, and future of practice. Journal of Policy Analysis and Management: The Journal of the Association for Public Policy Analysis and Management, 18(3), 411–425.

33.

Lynn

L. E.

Jr. Heinrich

C. J.

Hill

C. J.

(2001). Improving governance: A new logic for empirical research. Georgetown University Press.

34.

Manion

(1991). Policy implementation in the People’s Republic of China: Authoritative decisions versus individual interests. The Journal of Asian Studies, 50(2), 253–279.

35.

Matland

R. E.

(1995). Synthesizing the implementation literature: The ambiguity-conflict model of policy implementation. Journal of Public Administration Research and Theory, 5(2), 145–174. https://doi.org/10.1093/oxfordjournals.jpart.a037242

36.

May

P. J.

(2015). Implementation failures revisited: Policy regime perspectives. Public Policy and Administration, 30(3–4), 277–299.

37.

May

P. J.

Winter

S. C.

(2007). Politicians, managers, and street-level bureaucrats: Influences on policy implementation. Journal of Public Administration Research and Theory, 19(3), 453–476.

38.

Maynard-Moody

Musheno

Palumbo

(1990). Street-wise social policy: Resolving the dilemma of street-level influence and successful implementation. Western Political Quarterly, 43(4), 833–848.

39.

Mazmanian

D. A.

Sabatier

P. A.

(1981). Effective policy implementation. Free Press.

40.

McLaughlin

M. W.

(1987). Learning from experience: Lessons from policy implementation. Educational Evaluation and Policy Analysis, 9(2), 171–178.

41.

Meier

K. J.

O’Toole

L. J.

Jr. (2001). Managerial strategies and behavior in networks: A model with evidence from US public education. Journal of Public Administration Research and Theory, 11(3), 271–294.

42.

Meyers

M. K.

Vorsanger

Peters

B. G.

Pierre

(2007). Street-level bureaucrats and the implementation of public policy. The Handbook of Public Administration, 153–163.

43.

Nielsen

P. A.

Moynihan

D. P.

(2017). Romanticizing bureaucratic leadership? The politics of how elected officials attribute responsibility for performance. Governance, 30(4), 541–559.

44.

O’Brien

K. J.

(1999). Selective policy implementation in rural China. Comparative Politics, 31(2), 167. https://doi.org/10.2307/422143

45.

O’Toole

L. J.

Jr. (2000). Research on policy implementation: Assessment and prospects. Journal of Public Administration Research and Theory, 10(2), 263–288.

46.

O’Toole

L. J.

Jr. Peters

Pierre

(2012). Interorganizational relations and policy implementation. In Peters

B. G.

Pierre

(Eds.), The SAGE handbook of public administration (pp. 292–304). SAGE.

47.

Odden

A. R.

(1991). The evolution of education policy implementation. In A. Odden (Ed.), Education policy implementation (pp. 1–12). SUNY Press.

48.

Pressman

J. L.

Wildavsky

A. B.

(1973). Implementation: How great expectations in Washington are dashed in Oakland; or, why it’s amazing that federal programs work at all. University of California Press.

49.

Qian

Weingast

B. R.

(1997). Federalism as a commitment to preserving market incentives. The Journal of Economic Perspectives, 11(4), 83–92.

50.

Riccucci

(2005). How management matters: Street-level bureaucrats and welfare reform. Georgetown University Press.

51.

Rothstein

(2015). The Chinese Paradox of high growth and low quality of government: The cadre organization meets Max Weber. Governance, 28(4), 533–548.

52.

Sabatier

P. A.

(1986). Top-down and bottom-up approaches to implementation research: A critical analysis and suggested synthesis. Journal of Public Policy, 6(1), 21–48.

53.

Sabatier

P. A.

Mazmanian

(1979). The conditions of effective implementation: A guide to accomplishing policy objectives. Policy Analysis, 5, 481–504.

54.

Saetren

(2014). Implementing the third generation research paradigm in policy implementation research: An empirical assessment. Public Policy and Administration, 29(2), 84–105. https://doi.org/10.1177/0952076713513487

55.

Schubert

Ahlers

A. L.

(2012). County and township cadres as a strategic group: “Building a New Socialist Countryside” in three provinces. The China Journal, 67, 67–86.

56.

Siciliano

M. D.

Moolenaar

N. M.

Daly

A. J.

Liou

Y. H.

(2017). A cognitive perspective on policy implementation: Reform beliefs, sensemaking, and social networks. Public Administration Review, 77(6), 889–901.

57.

Smit

(2005). Teachers, local knowledge, and policy implementation: A qualitative policy-practice inquiry. Education and Urban Society, 37(3), 292–306.

58.

Smith

(2009). Political machinations in a rural county. The China Journal, 62, 29–59.

59.

Soss

Fording

Schram

S. F.

(2011). The organization of discipline: From performance management to perversity and punishment. Journal of Public Administration Research and Theory, 21(Suppl. 2), i203–i232.

60.

Spillane

J. P.

Reiser

B. J.

Reimer

(2002). Policy implementation and cognition: Reframing and refocusing implementation research. Review of Educational Research, 72(3), 387–431.

61.

Talen

(1996). Do plans get implemented? A review of evaluation in planning. Journal of Planning Literature, 10(3), 248–259.

62.

Thomann

Hupe

Sager

(2018). Serving many masters: Public accountability in private policy implementation. Governance, 31(2), 299–319.

63.

Tummers

Bekkers

(2012). Discretion and its effects: Analyzing the experiences of street-level bureaucrats during policy implementation. file:///C:/Users/Evans/Downloads/Metis_180041.pdf.

64.

Tummers

Rocco

(2015). Serving clients when the server crashes: How frontline workers cope with E-government challenges. Public Administration Review, 75(6), 817–827.

65.

Van Meter

D. S.

Van Horn

C. E.

(1975). The policy implementation process: A conceptual framework. Administration & Society, 6(4), 445–488.

66.

Wang

(2013). Factors affecting tourism policy implementation: A conceptual framework and a case study in China. Tourism Management, 36(June), 221–233.

67.

Winter

S. C.

(2012). Implementation perspectives: Status and reconsideration. In Peters

B. G.

Pierre

(Eds.), The SAGE handbook of public administration (pp. 265–278). SAGE.

68.

Wright

D. S.

(1978). Understanding intergovernmental relations. Duxbury Press.

69.

(2011). The fundamental institutions of China’s reforms and development. Journal of Economic Literature, 49(4), 1076–1151.

70.

Zhan

W.-H.

Tang

S.-Y.

(2013). Contextual changes and environmental policy implementation: A longitudinal study of street-level bureaucrats in Guangzhou, China. Journal of Public Administration Research and Theory, 24(4), 1005–1035.

71.

Zhang

B. Y. J.

(2011). Tracking the implementation of green credit policy in China: Top-down perspective and bottom-up reform. Journal of Environmental Management, 92(4), 1321–1328.

72.

Zhang

(2016). Political Elite Coalition and Local Administrative Reform in China—A case study of Shunde under Wang Yang. Journal of Contemporary China, 25(98), 277–291.

73.

Zhang

Rosenbloom

D. H.

(2018). Multi-level policy implementation: A case study on China’s administrative approval intermediaries’ reforms. Australian Journal of Public Administration, 77(4), 779–796.

74.

Zhou

L. A.

(2007). Research on the promotion championship model of Chinese local government officials. Economic Research Journal, 7, 36–50.

75.

Zhou

(2010). The institutional logic of collusion among local governments in China. Modern China, 36(1), 47–78.