Introduction
In current public administration theories, street-level bureaucrats (SLBs) have two contradictory but equally impressive faces. The first face is revealed by studies of SLBs’ policy entrepreneurship. Scholars have intensively discussed the phenomena of SLBs defining the content of public policies (Riccucci Reference Riccucci2005), creating significant rules-in-use (Lipsky Reference Lipsky1980), and bending inappropriate rules to serve citizens (Maynard-Moody and Musheno Reference Maynard-Moody and Musheno2003). SLBs are sometimes “citizen agents” more than “state agents” (Maynard-Moody and Musheno Reference Maynard-Moody and Musheno2000). They may even comprise a political group that leverages policy innovations (Arnold Reference Arnold2015; Lavee and Cohen Reference Lavee and Cohen2019). In this vein, SLBs can be as active as social activists. However, the opposite face is emphasized by bureaucracy research, indicating that SLBs are literally bureaucrats. They play the blame game, treat people as cases rather than human beings, and care mostly about risk avoidance (Denhardt Reference Denhardt1981; Hummel Reference Hummel1977). To a great extent, SLBs have to perform as bureaucrats to protect themselves because they are “unlike the top bananas” who have formal authority (Hood Reference Hood2011, p. 32).
The contradictory faces in SLB stories lead to a significant question: how can SLBs be active as policy entrepreneurs when they are required to rigidly follow bureaucratic principles as frontline bureaucrats? This question is especially important regarding SLBs with mandatory power. To better respond to increasing social concerns over law enforcement, the devices and requirements of law enforcement records have been developing globally in recent decades. Street-level work is no longer a supervisory blind spot; however, law enforcement officers are still typical SLBs with the two opposing faces. The connection between the faces is a necessary missing piece in understanding SLBs’ policy entrepreneurship.
More importantly, if the face of a drab and timid bureaucrat is rational for the sake of job security, systematic rationales should prompt SLBs to choose risk while somehow protecting them from potential punishment. Therefore, both opposing faces can be observed. Exploring these rationales helps us bridge the findings of two major fields in the literature and suggests some cues to better motivate SLBs.
To reveal these rationales, we should change traditional dichotomized thinking, such as that related to entrepreneurs/bureaucrats, policymakers/implementors, or citizen/state agents. Tiny adjustments in everyday administration could cause dramatic policy changes. Over a long time span, initiative may emerge from a superficial lack of change. The integration of innovation and risk avoidance may present a bigger picture of frontline bureaucrats’ policymaking, since they should promote change while also keeping their jobs.
In this study, we employ a historical case of frontline taxi officers in Guangzhou promoting a policy innovation throughout years of tiny administrative adjustments from 1992 to 2011. A momentarily unsolvable legislative problem was eventually changed based on these undramatic routines. Because taxi administration is associated with permit issuance and supervision, the relevant law enforcement has worldwide sensitivity. The tension between public and private interests easily leads to social discontent, protests, and lawsuits (Esbenshade et al. Reference Esbenshade, Shifrin and Rider2019; Xiong Reference Xiong2018). Taxi officers must strategically manage complicated claims, demands, and pressures from diverse groups. The activities related to managing tension are responses to the core concern of SLB studies. More importantly, in contrast to the diversity of political contexts, the norms and logic of bureaucracy create common constraints and opportunities faced by frontline bureaucrats in different countries. Thus, the details of SLBs’ strategic actions inside the bureaucracy can provide more threads for understanding SLBs’ policy entrepreneurship than merely focusing on legislative processes. This is especially important in countries that have a bureaucratic tradition or a long history of a civil official system. In this research, the specific manner in which a new policy is promoted by bottom-up bureaucratic skills rather than through legislative advocacy is interpreted as bureaucratic entrepreneurship.
Policy entrepreneurship: individual discretion or collective mindfulness
SLBs play two roles: state agent and citizen agent. In the role of state agent, SLBs are needed to ensure that rules and orders are adhered to. Bureaucrats have impersonal characteristics and focus on state goals and administrative processes (Hummel Reference Hummel1977). Almost all bureaucratic practices, such as the blame game, self-preservation and red tape, are experienced by the frontline administration (Hood Reference Hood2011). In addition, SLBs are citizen agents. They are not completely impersonal gears in a cold-blooded machine that merely follows mandatory principles (Maynard-Moody and Musheno Reference Maynard-Moody and Musheno2000). In many cases, SLBs help citizens by initiating strategies, motivations, and norms. This discretionary space is sometimes intentionally created by formal political institutions to improve flexibility (Stazyk and Goerdel Reference Stazyk and Goerdel2011), enabling the bureaucracy to be more efficient without being choked by red tape.
Regarding SLBs’ role switching, individualism is a prevalent but questionable approach. An individualist interpretation focuses on individual SLBs’ discretion and describes a role switch as a personally free choice. However, many frontline bureaucrats work in command-and-control organizations with predetermined legal requirements and procedure-oriented supervision. When confronting the conflict between lawfulness and reasonability, SLBs usually mirror “high-reliability organizations” based on “collective mindfulness” to protect themselves with organizational support (Homberg et al. Reference Homberg, Vogel and Weiherl2019, p. 29), especially when they play the de facto legislative role under the risk of being charged with malfeasance. The idea of collective mindfulness complements the prevalent individualist studies on the basis of three different assumptions.
First, the use of discretion is a type of collective action. Superficially, administrative discretion is individually implemented. However, officials in the same position are an organizational group rather than independent individuals (Blau Reference Blau1954). This is important when SLBs have to step into gray areas without the protection of written laws because they then need their colleagues to “have their backs.” To practice collective mindfulness, SLBs need internal connections to be able to discuss problems with colleagues or even lobby senior bureaucrats who have formal authority. Insider activism exists in a bureaucracy, although it differs from legislative activism (cf. Olsson and Hysing Reference Olsson and Hysing2012).
Second, being a citizen agent is a result, not merely a motivation. SLBs sometimes wish to prioritize serving people—even by bending the rules (Maynard-Moody and Musheno Reference Maynard-Moody and Musheno2000). However, the norms of bureaucracy are procedure-oriented. From the view of bureaucracy, rule-bending and the abuse of power are two sides of the same coin. Administrative accountability cannot rely merely on practitioners’ consciences because such accountability is usually associated with punishment (Finer Reference Finer1941). In almost all bureaucratic systems, a department is assigned to prevent rule-bending. In addition, related officials at higher levels do not always share the same feelings as those of SLBs but have the formal authority to punish SLBs’ rule-bending (Hummel Reference Hummel2006). Being a citizen agent is not an individual, free choice. The existence of a citizen agent always means that the SLB has successfully applied certain strategies to survive administrative supervision.
Third, collective mindfulness depends on the acquiescence of the bureaucratic structure. The individualist interpretation of policy entrepreneurship sometimes leads to a one-sided focus on SLBs’ “people work.” However, SLBs are still inside, rather than outside, the bureaucratic network. If the relevant departments are not willing to support the action plan of an organization, such work can hardly be practiced (Wilson Reference Wilson1989). Regarding policy entrepreneurship, the development of collective mindfulness is both an organizational and an interorganizational affair. The risk taken by SLBs is probably in vain if other departments stop or reverse discretionary decisions through legal authority.
SLBs take individual initiatives, but their actions are always associated with the perspectives of their peer group, supervisors, and related departments in the bureaucratic network. The purposes of policy innovation are sometimes carried out by administrative routines without dramatic political events. In this sense, the examination of SLBs’ entrepreneurship should include an analysis of SLBs as a group rather than as individuals.
Methodology and methods
This research employs an interpretive methodology. The methodology of historical interpretation (Pierson Reference Pierson2005) and the storytelling methods of Maynard-Moody (Reference Maynard-Moody, Musheno, Yanow and Schwartz-Shea2006) in SLB studies are the two intellectual foundations of this project. In contrast to other individualist frameworks, Pierson’s methodology emphasizes policy changes that occur because of administrative routines, not merely as dramatic political events. The absence of conflict does not signify the absence of change (Derthick Reference Derthick1979). A “snapshot” individualist study without historical context may miss significant elements of a real story. In addition, the “present” is influenced by the “past” rather than merely by the current choices. “When things happen within a sequence of a process affects how they happen and constrain possible outcomes at a later point in time” (Tilly Reference Tilly1984, p. 14). Therefore, the historical sequence of nonconflicting changes in routines is the basis of this project. The interpretation explains how frontline bureaucrats understand the relationship between their slow-moving frontline routines and the promotion of the policy process from “invisible” to “visible” changes.
In addition, a historical interpretation is not a grand narrative. Particularly in the field of SLB studies, how SLBs understand the problems of routines and what changes they expect are important. Maynard-Moody’s methods of storytelling are thus applied. By asking officers to recall stories and comparing their recollections to archives, researchers can determine what impressed SLBs about policy changes in different historical periods. This process creates triangulation with the records of formal archives and mass media. More importantly, SLBs’ stories help researchers clarify the types of tiny changes in administration that are related to policy formulation and on the basis of what logic. Since the historical sequence of formal policy changes is verifiable in the archives, the understanding of SLBs within oral history is significant for developing the meaning of administration.
These interpretivist methodologies are applicable to this research for two main reasons. First, how SLBs strategically play two opposite roles depends on their understanding of the meaning of their activities. The strategic switch in roles is because people believe that doing so helps achieve their purposes, say self-preservation or policy innovation, rather than because of functional causes such as administrative orders. Regardless of objective truthfulness, even a so-called “false” belief still leads to real actions. In this regard, the epistemology of this research should be intentionalist rather than functionalist. An interpretivist summary of the co-constructed meaning in the specific situation of policy innovation is more suitable than other positivist methods that attempt to prove the meaning objectively true or false. Second, policy entrepreneurship is signified by intentions and associated efforts, not by results. To policy entrepreneurs, no strategy is destined to succeed in the sense of a causal mechanism; however, even an unsuccessful policy innovation attempt would not override the intentions, efforts, or roles of policy entrepreneurs. When SLBs develop a strategy for promoting a policy change, they construct a meaning system to understand themselves encompassing the roles of citizen and state agents, which cannot be revealed by merely positivist results. This is why scholars choose the method of storytelling and interpretivist epistemology to summarize SLBs’ perceptions in complex situations (Maynard-Moody and Musheno Reference Maynard-Moody, Musheno, Yanow and Schwartz-Shea2006). The interpretation of SLBs’ understanding is the core of understanding their role-playing.
In intentionalist research, the evidentialist truthfulness of officers’ beliefs is less relevant than honesty. An existing “mis-understanding” is still a real understanding, while bureaucratic idle talk may be safe, correct, and “evidential” in law but not in heart. Especially in terms of the nature of law enforcement officers, such as “siege mentality culture,” “hard to track population,” and risk-averse tendency, the usual manners for improving reliability through performance records may conversely distort the trustworthiness of intentionalist research (cf. Butterfield et al. Reference Butterfield, Edwards and Woodall2004; Cohen and Cohen Reference Cohen and Cohen2023; Weick et al. Reference Weick, Sutcliffe, Obstfeld, Staw and Cummings1999). A storytelling study needs to be carefully designed to encourage honesty.
This project began in 2006 through friend-of-friend referrals to taxi officers. From 2006 to 2012, the researchers collected multiple historical archives (Table 1). The main issues and historical sequences are summarized from these formal documents. Then, the researchers organized two rounds of interviews to collect officers’ stories and perceptions associated with the topic. However, because the topic of this study is sensitive and taxi officers are not a large team, the face-to-face interview method may have a strong bias: in terms of trustworthiness, interviews involving as many officers as possible may be better. However, once the number of interviewees is close to the total number of the team, the participating officers can be easily inferred regardless of whether they are anonymous. The usual methods of improving data trustworthiness may conversely distort the collected data.
Table 1. Types of collected materials
To reduce officers’ worries and collect truthful stories, an officer was invited back to the team to collect related stories from officers and retell these stories in 2011. Then, after reviewing the collected stories and summarizing further questions, we asked another officer to follow up and complement them in the same way. When so doing, we did not know any of the officers’ personal information, the number of interviewees, or which story happened to whom. All officers could safely share their stories (see Table 1 for the number of stories). The trustworthiness of the stories was not evaluated by the report of the rate or frequency but depended on mutual authentication among different types of historical records. The Supplementary Appendix provides further details about the design of the data collection to improve trustworthiness and research ethics.
Because perceptions and stories were integrated into transcripts, process coding was applied to clarify the meaning units of stories, while topic coding was employed to summarize the officers’ perceptions and logic in their stories (Saldaña Reference Saldaña2013, pp. 87; 96). Then, the logics that emerged were integrated into the historical context.
The taxi administration sector: law enforcement of the taxi industry in china
Taxi officers are frontline law enforcement bureaucrats. In China, different cities name this type of team with little nuance. For example, the city of Changsha calls it the “taxi law enforcement team” (chuzu zhifa dadui). Some cities in Liaoning Province name the team the “taxi administrative enforcement team” (chuzuche guanli xingzheng zhifa dadui). In Guangzhou, the official name of the team has been changed multiple times. During the time span of this research, the original name back to the 1990s was mostly used – “the section of taxi administration” (chuzuche guanli ke) in the passenger traffic administration department (keyun jiaotong guanli chu), affiliated with the Traffic Committee of Guangzhou. After 2009, this department was expanded as the Traffic Law Enforcement Bureau (jiaotong xingzheng zhifa ju) and was affiliated with the newly formed Municipal Transportation Bureau; however, the duties of taxi officers remain largely unchanged (cf. Guangzhou Municipal Traffic Law Enforcement Bureau 2020).
The team of taxi officers is usually made up of 18 officers from the municipal team plus 4 officers assigned to each district (the number of districts changed several times, but there were 12 districts during most of the research period). According to the Interim Measures for the Taxi Administration of Guangzhou (Guangzhoushi Chuzuche Keyun Guanli Zanxing Banfa) 1998 and other local laws, the “section of taxi administration” is the lowest-ranking taxi-relevant administration and law enforcement section – there are no lower-level affiliated agencies. Thus, taxi officers are frontline policy implementers and supervisors, performing tasks such as qualification examination/license issuances, taxi safety checks, and other supervisory affairs related to the taxi business. Taxi officers usually organize their issues into three categories: taxi cars, safety checks, and customer complaints. In this regard, officers have both frontline administrative interactions with citizens in offices (e.g. qualification examinations and license issuance) and law enforcement on the streets (e.g. action related to stop-inspect-fine illegal taxi business or modified taxi cars).
The assignments of taxi officers mostly require direct interaction with people. According to the laws, taxi officers’ work is associated with four groups of people: taxi drivers, license owners (individuals and companies), taxi company managers, and taxi customers. All of these examinations and supervisory tasks are labor intensive. Officers’ salaries are the main operating costs.
Importantly, taxi officers have three considerable types of discretion. First, the schedule of law enforcement actions is left to their discretion due to its superficial triviality, but it may create significant influence. A typical example is when visiting a taxi company to conduct a safety check. According to the laws, all main managers and managers of workplace safety in taxi companies should pass a national examination of workplace safety (four times a year) to remain qualified. If taxi officers arrange the check schedule after an exam, managers in fact have one more chance to pass the exam before the safety check.
Second, legal documents sometimes leave space for taxi officers’ discretion. For example, State Council Communique 2005 No. 1, Chap. 10, Item 6 requires taxi officers to “teach taxi drivers how to protect their lawful rights” but also states that they should “criticize and educate drivers who gather a crowd to make trouble” and then fine those who “lead in making trouble, block traffic, disturb the social order, endanger public security, or infringe on public or private property rights or other people’s safety.” The lines between these categories to some extent depend on street-level judgment.
Third, street-level law enforcement sometimes relies on taxi officers’ professional judgment. For example, if a safety hazard is identified, the amount of time allowed to a company manager or taxi driver to eliminate the hazard is based mostly on taxi officers’ professional experience (the time could be between a week and half a year, depending on the difficulty). Additionally, if a taxi customer files a complaint that would result in a fine for the driver, taxi officers need to professionally estimate the trustworthiness of the driver’s and customer’s words. Taxi officers need to determine whom to trust based on the evidence. Their decisions will influence the procedure of the case.
In summary, taxi officers are a group of law enforcement SLBs supervising a particular industry. Theoretically, they have no legislative authority to make policies but rather merely enforce laws. However, in some conditions, they may promote a new policy in their everyday work.
The hollow state after privatization: the taxi administration in Guangzhou, China
The “hollow state” means a state that is not able to integrally enforce its administrative authority (Terry Reference Terry2005). Regarding the history of taxi administration reform in Guangzhou, China, the hollow state was once a tangible problem. As in many other metropolises, this problem in the taxi administration of Guangzhou was caused by the use of privatization as a solution to the public financial deficit in the 1990s.
In 1992, the city of Guangzhou decided to build a mega-underground metro system. This ambitious development project required enormous city government funding in a short time. To meet this instant demand for public funding, the city government pursued a specific financial policy called “Ten Approaches to Financing” (Rong Zi Shi Tiao), in which the privatization of taxi services was considered an important approach. Based on this policy, Guangzhou organized three auctions in 1992, 1995, and 1998 and issued a total of 1,872 taxi operation licenses to bidders. Each taxi car was to be registered with a license. Guangzhou announced that this reform would promote taxi services and business competition. Through auctions and fee collection during this privatization reform, the city government quickly raised 885 million RMB (He and Wang Reference He and Wang2012, p. 63).
Overall, the marketization of the taxi business was common in China in the 1990s. However, most Chinese cities managed taxi licensing as a type of administrative permit system. License holders could run taxi businesses but did not have property rights in the licenses. Licensing was managed by the local government.
The case of Guangzhou was different. The urgent demand for public funding pushed privatization in Guangzhou much further: license ownership was privatized. After bidders obtained licenses during auctions, the licenses worked like a title to private property. The private owners of the licenses were able to sell or lease their licenses to others. In addition to the taxi driver's license, a qualified driver without an operating license had to buy or rent one to legally operate the taxi business. More importantly, because these licenses were treated as private property, the license trades or rent prices were considered marketized affairs outside the control of the taxi administration.
In addition, the city government informally promised that after the three auctions, it would not issue new licenses for a long time to protect the interests of the winning bidders. The city government strictly kept its promise to the initial investors and issued no new licenses from 1998 to 2007. Even worse, since this privatization was carried out in haste, the original marketized license price was calculated based on the instant budget gap for metro development rather than the market value. Thus, the value of the licenses was seriously underestimated in the auction.
In this regard, the privatization of the taxi business not only removed this business from governmental management but also fixed the number of valid licenses. The prices and rental costs of licenses consequently skyrocketed. The rental price of a taxi operation license in 1993 was approximately 2,000 RMB but rose to over 10,000 RMB in 2000. License rentals cost drivers five times more after privatization than before. In 2008, the price for purchasing the ownership of a license reached as high as 600,000 RMB. After privatization, private owners determined the market with limited governmental control. Owners’ interests overpowered the interests of drivers.
Disputes between taxi drivers and license owners escalated during this situation. On November 12, 23, and 25 and December 1, 2008, taxi drivers staged several strikes. In particular, the strike on November 23 at the International Convention Center significantly raised tensions. During this year, the police department repeatedly warned of the serious risk to public security due to the ineffective taxi administration through the official internal reports “Public Security Information Express” (Gongan Xinxi Kuaibao) No. 2392, 2399, and 2534 (He and Wang Reference He and Wang2012, p. 71).
After that, the public discontent derived from taxi privatization was widely reported by both the Chinese and international media. For example, BBC News (2008) reported that 10,000 taxi cars joined the strike. Taxi drivers were preparing to present a petition in front of the building of the Traffic Committee, which was the location of taxi officers’ offices. According to BBC, 70% of the city’s taxi drivers had joined the strike. The main reason was the high rental price. South China Morning Post reported that “about 70 percent of Guangzhou’s 17,821 taxis went on strike yesterday” and called it “the largest industrial action since 1989.” The newspaper also summarized that the high “car rent” and “illegal taxis” were “two key grievances of the licensed drivers” (Zhai Reference Zhai2008).
The grievances of taxi drivers not only created conflicts but also reduced the quality of taxi services and hurt customers’ interests. For example, BBC (2008) reported that “customers had to wait over 30 minutes for a taxi.” The most influential official Chinese media, Renmin Wang, called the implementation of taxi policies in Guangzhou “embarrassing” and summarized that when the situation was only good for rent-seeking license owners, taxi drivers would be naturally not sufficient. Then, the interests of both customers and drivers were seriously damaged (Luo Reference Luo2008). VOA reported the interview of two taxi companies. Even the companies that did not like to talk about the strike still agreed that it was difficult for customers to wait for a taxi (Lin Reference Lin2008).
The city government was also frustrated with the situation. On the one hand, most governmental officials sympathized with taxi drivers and were unsatisfied with taxi services. For example, on local TV news, the mayor publicly appealed to the public and the related departments of the city government to pay attention to drivers’ claims rather than criticizing the drivers for their actions (Sohu 2008). On the other hand, because the high rental price was created and legitimized by the previous privatization policies to satisfy urgent fundraising needs, the city government had limited lawful policy instruments to use to directly affect the situation. According to a brief record of a mayor’s meeting (General Office of Guangzhou Municipal People’s Government 2007, p. 4), at least approximately two years before the strike, the city government realized that the high rental price created by the privatization led to “urgent problems,” but the conclusion of the meeting was “there are still no clear solutions.” Then, the forecasting conflicts unfortunately happened.
Even the license owners were to some extent discontent because they hoped the industry could be kept in order and healthy. According to the report of Renmin Wang, at least three taxi companies and their taxis were attacked in 2008 (Luo Reference Luo2008). The high rental price also pushed more drivers to rent illegal taxis without licenses, intensifying the market chaos. Reuters also reported that “dozens of police grappled to restore order as more than 100 cab drivers carrying bricks and rocks intercepted and smashed unlicensed taxis in Guangzhou” because “the rising rental fees … threatened drivers’ livelihoods” (Reuters Staff 2008). To relieve the tension, the Guangzhou taxi companies once collectively reduced the monthly rent by 800 RMB. However, it did not help stop the turbulence. From the archives of this period, we can see the confusion and the feeling of helplessness at that time. Not only did the aforementioned mayor’s meeting record “no solution” as the meeting conclusion, but also the then information report of think tanks “Lingdao Juece Xinxi” (“Information for Leaders’ Decisions”) from Beijing summarized that the 800 RMB rent reduction seemed useless in relieving the tension. The information report used a question as the title, “Why did Guangzhou taxi drivers still strike after the new policy” but did not provide an answer (Information for Decisions Magazine 2008).
During this period, taxi drivers, license owners, customers, and government departments were to some extent dissatisfied with the situation in the taxi industry. Taxi officers were confronted with a dilemma. On the one hand, they were the administrators of a particular industry and were responsible for improving taxi service quality. This general discontent placed great pressure on taxi officers. On the other hand, the taxi administration was merely a frontline section. Taxi officers had neither the authority to change the privatization policy for urgent fundraising nor the power to override license owners’ property rights to reduce rental prices.
From bureaucrats to policy entrepreneurs: fundamental conditions
The key argument of this study is that we should change the dichotomized thinking and take an integrated understanding of both SLB’s policy innovation (e.g. policy entrepreneurship) and bureaucratic self-preservation (e.g. blame games). Typically, in studies on policy entrepreneurship, the responsibility of policy entrepreneurs is supposed to be given. However, in the study of bureaucracy, the attribution of responsibility is a dynamic gaming process. Historically, we now know that the problems created by privatization were finally settled by the new regulation issued by the local congress. However, during past drivers’ strikes, taxi drivers did not gather and announce their petitions at the congress but, rather, in front of the taxi officers’ office building.
The phenomenon that the public blames SLBs for general policy problems is global. In the West, it is sometimes called the “absence of role differentiation,” noting that the public does not make distinctions between civil and political officials when blaming the failure of the whole public sector (Frederickson and Frederickson Reference Frederickson and Frederickson1995). Especially in the context of China, SLBs more frequently become default blame takers because the challenges created by policy design may be usually interpreted as poor policy implementation (Gao Reference Gao2015).
However, scholars have also realized that SLBs have strategies to avoid blame—in both Britain (Hood Reference Hood2011) and China (Li et al. Reference Li, Ni and Wang2021). When confronted with social discontent, rational SLBs probably continue to pass their responsibilities until they lose the blame game. Therefore, to bridge the two academic threads, the conditions under which SLBs lose the blame game while taking responsibility as policy entrepreneurs are fundamental.
Similar to other SLBs, taxi officers seemed both powerful and helpless. They were directly in charge whenever disputes related to the taxi business arose, but also helpless if the causes of the disputes were outside their authority. Usually, top-down law-changing or bottom-up rule-bending are alternative solutions. However, when these coping strategies are not available, SLBs are likely to become policy entrepreneurs.
Results-oriented political pressure
In most cases, SLBs have limited opportunities to directly interact with politics. They are supposed to implement formal rules following the legal process. In so doing, process-oriented concerns could be prioritized over substantial results. However, in other cases, politics in frontline administration can be very substantial. For example, if a political campaign involves SLBs’ activities, the politics-administration relationship at the street level may suddenly change, and a street-level routine may suddenly have a political focus. More importantly, politics are, to a large extent, results-oriented.
Although systematic problems have existed since privatization in the 1990s, taxi officers did not engage in activism until 2004, when Guangzhou earned the right to host the 2010 Asian Games. Thus, passenger transportation became a crucial issue. In 2006, the city government formed a specific board called “The Leading Group of the Taxi Administrative Work of Guangzhou” (Chuzu Qiche Hangye Guanli Xiaozu), which was directly led by the deputy mayor. This group developed “a work plan for establishing a civilized taxi industry in Guangzhou” (Chuangjian Guangzhoushi Chuzu Qiche Wenming Hangye Gongzuo Fangan). This plan clearly required the taxi administration to create a “reliable, harmonious, and advanced” taxi industry “for the incoming Asian Games.” After the drivers’ strikes against taxi companies and license owners, the city government announced “eight measures for systematically implementing and promoting the healthy development of the taxi business” (Xitong Luoshi Cujin Guangzhoushi Chuzuche Hangye Jiankang Fazhan Baxiang Cuoshi); through these measures, the city government promised that the taxi administration would protect drivers’ interests (He and Wang Reference He and Wang2012, p. 71).
In this regard, the taxi administration section was no longer merely a low-level section. Rather, it was assigned to support the most important event in the city: the Asian Games. Theoretically, the conflicts of interest between private owners and taxi drivers needed a legal adjustment for a systematic solution to be created. Officers could have lawfully waited for changes in the laws. However, when the city leaders directly assigned taxi officers the task of meeting the goal without a change in the formal laws, the results-oriented political pressure reached the street level in 2008. This assignment suddenly became the top priority of taxi officers. In this situation, they could not say no to the deputy mayor and had to find some way to achieve the goal.
The failure of both top-down and bottom-up implementation
The failure of implementation is another significant condition. If a problem can be solved by a new policy or subtle rule-bending, frontline officials may be able to manage independently. However, in some situations, implementation adjustments cannot relieve the tension.
First, when a conflict involves contradictory interest groups, the rule-bending strategy would not work. Intergroup conflicts were different from state-citizen disputes because the rule-bending that involved taking a side would not be acceptable to the opposing group. For example, a taxi officer clearly expressed the idea that taxi administrators should support taxi drivers because the ultimate quality of taxi services depended on drivers. In the street world, drivers sometimes considered taxi officers their agents but treated police officers as the opposing group. Taxi drivers even preferred to call taxi officers before calling police officers when reporting a collision or street fighting.Footnote 1 However, during the conflicts between license owners and drivers, taxi officers needed to balance the interests of both groups. This balance could not be achieved merely by gaining drivers’ respect.
Second, regarding the top-down approach, it was not guaranteed that all urgent issues facing great political pressure could be addressed in the policy agenda. Rather, if an imperative policy was confronted with dilemmas in their formal approaches, both political and administrative leaders may even have relied more on SLBs. In the case of taxi reform in Guangzhou, privatization was undertaken mainly to raise public funds. However, after radical privatization, public interference with taxi services then needed additional financial investments from the public. The conflict between governmental responsibility and the lack of a budget created a prominent challenge. One taxi officer stated the following:
“The deputy mayor once said that the taxi service was a type of public transportation. It should give citizens more convenience. When he raised requirements to us, he would say something like that. However, if you truly take it seriously and apply public subsidies to the taxi industry, as other public transportation businesses have, he would say that the current budget is not enough! If the taxi business is a public service, it should be inexpensive, and the city government should cover the loss. If you consider it a private business, then the balance between price and service quality should be adjusted by the market, not by the government.”
At the frontline level, not all administrative assignments were as rational as formal policy formulation, with logical arrangements and lawful adjustments. When the goal was urgent (i.e. improving taxi services before the Asian Games) but the policy dilemma could not be logically solved (i.e. no extra public budget was available to cover private losses), frontline officers would receive elusive orders to meet a general goal without detailed information on how to achieve these administrative measures because no measures could be logically rationalized.Footnote 2 This situation created a lawful space of policy entrepreneurship for SLBs.
No room for the blame game
Although SLBs may sometimes be flexible, they are still bureaucrats. The logic of their behavior is consistent with that of bureaucracy. If the blame game works, then allowing an irrational assignment to become someone else’s problem is a rational, alternative solution. However, SLBs are not a monolithic group. Those who serve all citizens are different from those who focus on a particular group. An example of the former type is police officers, who have street-level law enforcement power in all cases related to public security. In contrast, taxi officers represent the latter. They serve a specific group and, thus, should be accountable for most affairs related to that group. Whenever conflicts occur in the taxi business, it is almost impossible for taxi officers to win the blame game against other bureaucrats.
Specifically, three administrative agencies were directly related to conflicts between taxi drivers and license owners: the police department was tasked with maintaining social order and preventing the use of violence, the price bureau assessed the fairness of price standards, and the taxi administration supervised taxi services. However, the reality of officialism is that no SLBs want to be involved in conflict, especially when officers of other agencies have already been involved and can be blamed. As a taxi officer noted, “Our officers once faced violence when they investigated illegal taxis. We called the police. However, the police officers just looked around and then left. The police officers did not support us. That brings us great troubles.” Similarly, the price bureau did not want to join in and become the one to be blamed since a systematic problem was derived from past policies associated with the urgent privatization process.
In this regard, playing the blame game is not an option for SLBs who are accountable to a particular group. If a case is thorny, these officials may even become victims of the blame game and lose the support of other SLBs. This situation forces a particular team of SLBs to develop entrepreneurship beyond bureaucratic routines.
The policy entrepreneurship of citizen agents: the art of cooperation
The basic idea of the concept of citizen agents is that SLBs not only use their discretion to implement existing policies but also initiatively act based on their judgment of the worth of citizen clients. Usually, the stories of citizen agents are related to normative judgments in frontline interactions but are much less associated with a response to political power (Maynard-Moody and Musheno Reference Maynard-Moody and Musheno2000, p. 329). If citizen agents are integrated with policy entrepreneurship and bureaucratic self-preservation, their normative but nonpolitical natures bridge the two purposes. On the one hand, SLBs attempt to lay a better foundation for future policy innovation that is ethically fair to each group of clients. Cooperation is created by social skills and norms. On the other hand, they must masquerade the underlying policy-changing concerns as nonpolitical routines to protect themselves. In this regard, bureaucratic coping in name could in fact be policy entrepreneurship.
In taxi officers’ stories, the purposes of policy innovation are usually hidden under the interaction of administrative trivialities until officers explain the real reasons. Taxi officers had no legislative authority to issue a needed new law. However, if new, unofficial requirements could be broadly accepted, they would become the de facto rules in use. Therefore, a key to policy innovation here was how to persuade multiple groups to accept a new arrangement and potential financial losses without legislative authority.
Frontline requirements in everyday administration
Taxi officers are not able to change a city’s privatization policy but could easily make some administrative procedures more “annoying.” Taxi officers collectively apply requirements to their administrative procedures. These requirements are not issued by the legislative sectors but are requested at frontline level. Theoretically, people can question these requirements; however, if they do, they encounter obstructive bureaucrats.
First, taxi officers pushed the transfer of license ownership from individual owners to taxi companies. After privatization, over 18,000 licenses were sold in the market. Approximately 2,000 of these licenses were held by individual holders. From 1998 to 2004, some licenses (approximately 1000) were frequently bought and sold. These owners did not care about the taxi business but merely considered the licenses to be a type of financial investment. “Some owners even knocked on my doors at midnight and tried to sell a license to me for 4,000 RMB, because they had just lost at the underground gambling dens,” an officer recalled.
The property rights of these owners were lawful, but they had no motivation to improve the services as long as they could obtain the desired rental price. To change this situation, taxi officers did not deny ownership but asked for an extra requirement: all licenses should be registered with taxi companies. Because individual owners would be annoyed by the process of searching for a taxi company to help them register, they would choose to sell their licenses. Especially when original owners passed away, their heirs – who were less familiar with the taxi business – would be more likely to sell their licenses.
Second, taxi officers “persuaded” taxi companies to sign employment contracts with taxi drivers, which made these drivers not only license renters but also employees.Footnote 3 This requirement greatly changed the laws applied in law enforcement practices. In the free rental market, taxi officers faced difficulties intervening in the relationship between owners and renters. However, once employment contracts were signed, labor and social security laws could be applied. Taxi officers could work with the labor union to intervene in disputes. Because drivers were employees, the labor union started to help them with illnesses and other difficulties. The relationship between license owners and drivers became warmer than it had been before.
Third, taxi officers attempted to exclude speculators from the market by treating cases of ownership transfer registration differently. They supported three types of ownership transfer: inheritance, exchange of a license for shares in a taxi company, and merger/reorganization of taxi companies. However, regarding ownership transfer for profiteering purposes, they switched to the face of the “annoying” bureaucrat to disturb and block such transfers. When the transfer approach was narrowed, the realization of money became difficult, depressing the speculative value of the licenses. After these strategies were employed, the number of taxi companies in Guangzhou dropped from 101 to 67 in 2011. Pure speculators were mostly squeezed out of the market.Footnote 4
Negotiation to relieve tension
Theoretically, taxi officers had no authority to develop formal regulations regarding administrative permits. Until the local congress legally passed the regulations, taxi officers took the responsibility to develop the so-called requirements that they employed in gray areas. Taxi officers knew the serious legal risks very well. Thus, seeking consensus from both sides was their foremost concern. Because the requirements were not originally developed through formal channels, taxi officers had to negotiate with stakeholders and hope that the informal agreement would close the legal loophole.
Drivers naturally welcomed taxi officers’ actions that decreased rental prices. However, negotiation with license owners was crucial since the new requirements directly damaged the financial value of their properties. For example, before the new requirements were implemented, if an owner used a license as a pledge to a business bank, he or she could probably obtain 400,000 to 500,000 RMB in a mortgage loan. However, after the taxi officers asked for the extra requirements, a bank probably approved only a 200,000 loan against a pledged license. In this regard, taxi officers used several strategies to balance owners’ losses in exchange for their support.
First, taxi officers personally met with the major taxi companies that had more than 200 licenses. They persuaded these companies that they were promoting the merger and reorganization of the taxi business. This situation provided a great opportunity for large companies to expand their business. As the price of a license decreased, the cost for these companies to purchase licenses also decreased.
Second, taxi officers explained that the new requirements emphasized fairness rather than merely price control. To seek the support of taxi companies, they suggested that the companies collect more service fees from drivers but reduce their rental prices. For example, taxi companies could charge drivers fees for parking, invoice books, printing, contract deposits, and so on. However, taxi officers needed to evaluate the quality of the services offered by these companies, such as the quality of repair shops, the outcomes of vehicle safety checks, and the number of customer complaints. This evaluation process led to taxi companies being categorized into four levels.Footnote 5 Taxi officers asked companies with lower-level services to charge less rent than those offering good services. All of these activities contributed to attempts to persuade taxi companies to accept their role change from a license leaser to a service manager. Good companies had a new profit collection method, while the high service fees for good services to drivers were fairer than merely high rental prices.
Third, some taxi officers personally promised taxi company owners who suffered the greatest losses from declining license prices that they would help these owners obtain new licenses in the future. Although new licenses were sold by the city government through auctions, the technical standard for bidders’ quality evaluation was drafted by taxi officers because they knew the most about taxi management. Therefore, the advice from experienced officers regarding the technical standard to some extent determined the companies that could bid on the new licenses.
Ally formulation in personal relationships
The above strategies seemed to act as an exchange based on economic rationality. However, in practice, this interaction was more personal and emotional. Taxi officers helped both taxi companies and drivers when they were trapped in difficult situations, albeit they sometimes also implied a threat of paralyzing companies’ or drivers’ business by delaying administrative processes. Taxi officers used personal influence when asking multiple groups to cooperate. This was an important reason for the success of these strategies.
For example, taxi officers had good personal relationships with the top six largest companies. Whenever the taxi administration sector received new assignments, taxi officers discussed them with the managers of these companies. After reaching an agreement, the officers arranged a large meeting at which the representatives of each company openly expressed their support for the decisions. Taxi officers were also closely connected to the Taxi Association of Guangzhou. Some former taxi officers worked for the Taxi Association after they retired. The board members of the association were mostly from large taxi companies. Thus, officers easily connected with any key person in the taxi industry through professional association channels. When taxi officers were intensively involved in a business, they achieved consensus more easily than other agencies (cf. Taxi Association of Guangzhou 2007).
Policy entrepreneurship of the state agent: bureaucratic games
Because of their situations and connections, SLBs can, to some extent, achieve goals that cannot be achieved by a formal approach. However, once SLBs start to design policies, they step into a dangerous area. Multiple governmental agencies can stop innovation both easily and legally. In the worst case, the hindrance of innovation may be counted as an achievement, while the fulfillment of the assignment for the Asian Games may be considered a crime based on the process-oriented logic of bureaucrats. SLBs must also play a strategic bureaucratic game to prevent hostile interference from bureaucratic politics. In other words, this section is about the second face of SLBs – the bureaucratic self-preservation associated with policy entrepreneurship.
Request for written support from other administrative departments
Although SLBs could have sought consensus among social stakeholders, they still needed formal documents to protect themselves from potential supervisory charges and lawsuits. If related governmental departments agreed to publicly announce their support for a street-level innovation, then these multiple endorsements would make frontline policy entrepreneurs safer. These endorsements were extremely important, especially when the specific affair was separately controlled by several agencies. Taxi officers were not able to set a maximum rental price, as this administrative power belonged to the Price Bureau. Similarly, taxi officers could persuade taxi companies to sign employment contracts with drivers but had no power to develop the detailed standards of the contract, which was regulated by the Labor and Social Security Bureau. Therefore, taxi officers had to continually communicate with the officials of these bureaus.
Success in seeking support depended on bureaucratic competencies. Frontline officers’ paperwork skills needed to be outstanding because they were confronted with another bureau. While the incoming Asian Games placed great pressure on taxi officers, it also provided them with a good reason to ask for help from other bureaus. More importantly, taxi officers promised that if their strategic actions led to any dispute, they would take responsibility.
For example, the aforementioned document on “eight measures for systematically implementing and promoting the healthy development of the taxi business” was one of the most important formal documents, which was publicly issued together by the Traffic Committee, Police Bureau, and Price Bureau. However, another governmental archive of the mayor’s meeting ([2008] No.17) was attached to this document, called the “work plan for the eight measures” (Passenger Traffic Administration Department 2009b, pp. 275–289). In this document, taxi officers and the sections in the Traffic Committee accepted the responsibilities and accountability for most of the listed items on this problem-solving schedule.
Superficially, the work plan seemed to be not rational for taxi officers because it did not reduce taxi officers’ responsibilities. However, such a written archive represented significant support. Since the related departments had already agreed on the plan and filed it in the archives of the mayor’s meeting, these departments were unlikely to criticize taxi officers for abusing power in a potential blame game. Meanwhile, by that, taxi officers also demonstrated that requests for support were not a buck-passing trick, because the responsibilities were clearly listed. The construction of trust was fatal to policy innovation attempts in the bureaucratic blame world.
Political lobbies and selective implementation within the bureaucracy
Completely separating public administration from politics is impossible, even at the frontline level. SLBs also need to nurture the relationship between politics and administration during the policy innovation process. SLBs’ understanding and discretion in practice may shape the meaning of a case and influence public opinion on related topics. This is an important reason SLBs have specific power in policy entrepreneurship cases. However, regarding self-preservation, especially in law enforcement, the expression of personal understanding is usually considered a lack of tactics. To leverage a policy innovation, SLBs should promote the change but not in their names. Bureaucratic politics is thus covered by superficially nonpolitical administration consisting of two main types: inside political lobbies and selective implementation.
Regarding lobbying activities, reporting law enforcement cases for political leaders is significant. Theoretically, this type of report is a professional affair since it is about enforcing rather than developing laws. SLBs can still use the “following orders” defense in the case of a potential blame game. More importantly, the interpretation of some influential cases makes it easier to reach political leaders. In policy implementation, SLBs must always use discretion to identify and distinguish different types of cases. If they can persuade political leaders to accept the underlying rationales behind the interpretation, they can leverage the policy innovation safely inside a bureaucracy because they are introducing professional thinking about law enforcement without any unauthorized activity. Therefore, in a policymaking process, different groups of SLBs may compete for their interpretations in the tone of professional reports to the political leader.
In a story, a taxi driver was once beaten by a night-club security guard at midnight, but the police supported the rights of the night-club owner. The driver immediately called on a large group of taxi drivers to protest in front of the police station. The judgments on which law should be applied and whether the drivers were criminals or victims were crucial but complex. Taxi and police officers had different ideas about these fundamental judgments. Taxi officers thus viewed this case as an important event and actively reported to the Traffic Committee their ideas about the problems in the taxi industry and their suggestions with an analysis of the case. The directors of the Traffic Committee and police department finally carried out a debate on this case in front of the deputy mayor, enabling street-level officers’ ideas about policymaking issues to safely reach political leaders. In this regard, SLBs could engage in political lobbying functions through a nonpolitical bureaucratic channel, which helped them advocate for their policy thinking but reduce the associated risk of doing so.
Equally importantly, frontline officers also need to know how to divert political leaders’ attention. The concerns over the local politics for a campaign and the professional administration for an industry are not always consistent. A typical story was that the mayor’s foremost concern was to ensure that most visitors during the Asian Games enjoyed convenient taxi services. He ordered the issuance of 10,000 to 20,000 new licenses in the near future so that the price and wait time of these services would decrease during the Games. However, taxi officers knew that although they needed the mayor’s support, they had to protect the long-term interests of the taxi business rather than serve only the Asian Games. Thus, they divided the ordered number of new licenses over 5 years and issued fewer than 500 new licenses in a year (Traffic Committee of Guangzhou 2007, July 30). Whenever the mayor’s office inquired about progress in this area, taxi officers played the classic bureaucratic card and reported, “we are doing so now.”
Informal carrots and sticks in bureaucratic games
The formal supervision system of accountability is a great concern when SLBs attempt to go beyond lawful routines and promote policy changes. However, not all bureaucratic self-preservation activities stay in formal channels. Since changing policy is not a legal authority of SLBs, any related litigation is a great risk to their job security. Thus, avoiding litigation is significant to the legitimacy of gray area activities. If general stakeholders accept these activities without questioning their legality, no blame means there is no blame game to lose. To win the necessary acceptance, taxi officers’ SLB features created informal capacities for promoting unity or deterrence.
On the one hand, SLBs’ social capital was a prominent advantage. Taxi officers might bring outside resources and power into the bureaucracy and then unite officials of other departments. Because taxi officers’ jobs were closely connected to company owners and managers, most had good personal relationships. While these companies lobbied other higher-level policymaking departments for their business interests, they sometimes encompassed taxi officers’ ideas in their entire advocacy package. Some taxi officers might have played an information broker role in this process. Take one of the archived open letters from the association to taxi drivers as an example. This open letter shows that the top six largest companies and the association openly supported taxi officers and a potential new policy, albeit their financial interests might also have been influenced by the policy change (Taxi Association of Guangzhou 2007). Other clerical bureaucrats might not have had this type of external assistance when confronted with the same bureaucratic games. Thus, unity with taxi officers was mostly beneficial to both clerical and street-level bureaucrats.
On the other hand, taxi officers had to use “sticks” when “carrots” did not work in a game. Almost all law enforcement agencies have reporting systems that enable SLBs to connect with higher-ranking officials at each level. If a bureaucratic game is unfortunately initiated, it is not limited to the street level. Through the reporting channel, the section chief of taxi officers may ask the director of the passenger traffic administration department for help. Then, the director may turn to the chair of the Traffic Committee of Guangzhou and finally to the deputy mayor in charge of public transportation. When frontline taxi officers attempt to report to higher authorities, they can turn the blame game into a political relay race with a great deal of meetings and paperwork. That takes a great deal of personal time and energy for any game opponent. The aforementioned story of the tension between taxi and police officers is an example. The situation was originally about different ideas regarding a taxi driver’s protest, but the street-level ideas of these two law enforcement systems finally reached the level of the deputy mayor. Thus, the high cost of expanded games became an informal deterrence against potential challenges inside the bureaucracy. Although a similar strategy is common in bureaucratic games, law enforcement SLBs can use this type of deterrence for self-preservation in controversial policymaking more conveniently than other types of bureaucrats. The existing system for reporting loopholes in the law and the obstruction of law enforcement is naturally related to the design of the law and may create a larger cost to potential opponents.
The bureaucratic entrepreneurship of policy innovation
SLBs are usually the embodiment of both hospitality and officialism. A comprehensive SLB framework should respond to both stances and their dialectic relationship. Otherwise, we face the risk of being “blind people” who emphasize each real characteristic but lose sight of the whole “elephant” of SLBs’ bureaucratic entrepreneurship (Fig. 1).
Figure 1. Framework of frontline bureaucratic entrepreneurship.
The case of taxi officers reveals significant trends in SLBs’ action logic. The dialectic relationship among the dimensions of responsibility, performance, and accountability contributes to SLBs’ specific strategies in different situations. For either unchanging routine lives or entrepreneurship, the ternary dialectic of the three dimensions is the key concern.
During everyday routines, SLBs follow the same logic as that of other bureaucrats. However, compared to other clerical employees, SLBs are confronted with real people. Discretion in the application of the laws shapes the powerfulness of SLBs’ performance. At the same time, the space for discretion is limited because of the existence of a bureaucratic accountability supervisory system. If the use of discretion is not enough to solve problems, then SLBs are helpless because of their lack of legal authority. In this regard, the typical officialism with powerful discretion but limited legal authority creates the basic ternary dialectic of SLBs’ everyday lives.
If SLBs maintain their routines, they engage mainly in their typical behaviors, as existing studies have found. However, changing conditions possibly directly affected the ternary dialectic. For example, political pressure changed the procedure-oriented responsibility of taxi officers to a results-oriented responsibility. The mayor did not allow taxi services to become the shortcoming of the Asian Games plan regardless of whether the shortcoming was based on legal processes. The identity of SLBs’ responsibility thus changed. In addition, if legal processes or bureaucratic tricks could have enabled them to meet these goals, SLBs would prefer to perform the job in a bureaucratic way to maintain their safety. This explains why taxi officers clearly realized the problem and recorded it in archives years earlier than they did their active actions. However, when no new laws were issued in time to help but the problem must be solved, bureaucratic entrepreneurship emerged.
Compared to other policy entrepreneurs, SLBs were mediating agents between society and the state. When they were activated to promote a new policy, they not only revealed their potential to form a consensus at the frontline but also played bureaucratic games for self-preservation in the state sphere. Based on the changing content of the ternary dialectic, SLBs adjusted their goals, tactics, and self-protection approaches. Thus, superficially opposing behaviors were real – taxi officers sometimes strongly fought for taxi drivers’ interests – however, at other times, they were annoying bureaucrats paralyzing the taxi business. They possibly supported drivers’ rights against police officers but also kept a low profile to “beg” for support from other departments. This circumstance did not mean that some officers were citizen agents and others were state agents. Rather, they were all both citizen and state agents depending on the content of ternary dialectics.
As policy innovation is based on the logic of bureaucracy rather than legislation, the policy development promoted by SLBs has different stages. Bureaucratic entrepreneurship diffuses rules from bureaucracy to congress through the formation of consensus among stakeholders in the industry, then in the bureaucracy, and finally in the local congress. In the story of taxi officers, policy entrepreneurship was always associated with bureaucratic self-preservation. The attempt to control the blame game and the creation of acquiescence contributed to the legitimacy foundation for future formal policy changes. The bureaucratic structure introduced special characteristics to this unique type of policy entrepreneur (Table 2).
Table 2. Comparison of bureaucratic Entrepreneurship and classic policy Entrepreneurship
In contrast to other social activists, SLBs perform in an established system: their organization is formal in the bureaucracy, and they manage assigned problems. In most cases, they are forbidden from expressing political concerns without the approval of political leaders. For example, taxi officers’ report system and the channel to reach the mayor were not always attainable to social activists who were not bureaucrats. Therefore, SLBs can play the role of policy entrepreneurs with the cover of professional administration because they are state agents.
SLBs are also different from regular bureaucrats. While SLBs have to promote policy formation, the necessary interactions become bureaucratic politics from below instead of public politics at the top (Brower and Abolafia Reference Brower and Abolafia1997). During this process, SLBs closely exchange with stakeholders with whom they have strong ties. The bargaining process may be either formal or personal depending on the context. Therefore, both rational calculation (e.g. financial compensation) and affectional actions (e.g. due to loyalty, respect, or fear) coexist in the bargaining process. For example, taxi officers avoided litigation depending on their social relations and discretion during law enforcement practices. Other clerical bureaucrats may not be able to effectively carry out the same strategies because they lack the necessary conditions to play the role of citizen agents. Therefore, SLBs play the simultaneous roles of impersonal bureaucrats and human beings with feelings. They differ from either clerical bureaucrats or social activists in terms of policy innovation (Fig. 2).
Figure 2. Features of SLB in bureaucratic entrepreneurship.Footnote 6
Significantly, the tension between legal-procedural ethics and entrepreneurship is always prominent. Without the protection of formal laws, if any group of stakeholders sue or merely threaten to sue, the involved SLBs are endangered in the gray area. Therefore, the avoidance of litigation depending on paperwork skills, and professional reputations to a large extent determines whether the promoted informal rules had sufficient legitimacy. In this regard, this type of policy is not created by overriding obstacles when the policy window is finally open. Rather, it is derived from the careful maintenance of a general consensus at each policy stage.
Conclusions: bureaucratic entrepreneurship between the state and society
SLBs are not born as policy entrepreneurs because their jobs are designed for implementation and are nonpolitical. For their career safety, SLBs prefer to stay in safe routines until retirement. However, once the conditions are in place, their bureaucratic entrepreneurship for policy innovation is activated. Their informal connection and bureaucratic skills help them achieve an informal consensus within bureaucratic politics until the legislative sector formally approves the innovation. This research happens to be conducted in China but reveals important theoretical trends for understanding the roles of SLBs in policy formation despite its special historical context.
First, Strategic Incrementalism from Below. Negotiating by doing is a prominent specialty of the policymaking of SLBs. Because the new policy is gradually formalized in the implementation process before it receives congressional approval, policy cycle procedures are to some extent reversed. Some superficially tiny changes may in fact be significant policymaking steps, such as accepting drivers as employees or persuading heirs to sell licenses. If we consider these types of stories in other contexts, they may sound indifferent to other frontline routines without any policymaking advocacy. However, officers know that the nuances are prepared for some eventual problem-solving purpose, which cannot be finished in haste. What a policymaking action is depends on the strategy rather than merely formal procedures.
Second, Diverse Types of Politicization. The characteristics of political orders from the top and bureaucratic politics between administrative departments differ during the policymaking process. For example, political pressure from the mayor mostly affected SLBs’ responsibility, while the blame game between the police and taxi officers was related to accountability. In this regard, the ternary dialectic framework of this article can help examine the structure of related political exchanges, albeit there are overlapping areas, such as the departmental debate in front of the deputy mayor.
Third, Comprehensive Role Playing in Different Contexts. The relationship between the roles of citizen and state agents is not definite. These two roles are derived from the mediating status between the state and society. SLBs should be state agents to safely keep their jobs and should be citizen agents to achieve social support for their administrative performance. The role switching depends on the context of the ternary dialectics. More importantly, SLBs’ competencies are associated with their role-playing skills. Therefore, role-playing by SLBs has special effects beyond formal procedures; however, the specific strategy may not be repeatable by others in a different context.
Based on these findings, if political leaders encourage frontline officials to be policy entrepreneurs, they can examine the ternary dialectic among responsibility, performance, and accountability: a) the assignment should be sent directly to a particular department to forestall the blame game of officialism, b) the assigned group of SLBs should already have adequate strong ties to the industry and a good reputation in it, and C) to motivate entrepreneurship, process-oriented supervision probably needs to be temporarily flexible.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S0143814X24000023
Data availability statement
This study does not employ statistical methods and no replication materials are available.
Acknowledgments
We would like to thank Dr. Guanglong Wang and the two anonymous officers for their assistance with data collection. We also want to thank Dr. Thomas J. Catlaw and Dr. Joanna Lucio for their inspiring reading lists.
Funding statement
This project is supported by the Grant from the Major Program of Key Research Institute of Humanities and Social Science, Ministry of Education of China (No. 22JJD630023) and the National Social Science Foundation of China (Grant No. 19ZDA125).
Competing interests
We declare that we have no financial or personal relationships with other people or organizations that can inappropriately influence our research.
