key: cord-0073420-frfszoga
authors: Mei, Ciqi
title: Policy style, consistency and the effectiveness of the policy mix in China’s fight against COVID-19
date: 2020-06-30
journal: Policy Soc
DOI: 10.1080/14494035.2020.1787627
sha: edcd5b911929d626e8f23ec381b1f20731ec3019
doc_id: 73420
cord_uid: frfszoga

As the first country stricken by the COVID-19 pandemic, China deployed a policy response that was chaotic at the start but effective in the end. A complete account to explain China’s COVID-19 experience should explain both. By examining policy changes in China’s fight against the pandemic, I show that pandemic as an exogenous shock invalidated the normal policy logics followed by multiple policy actors, resulting in policy inconsistency and chaos. A policy mix comprised traditional measures, i.e. strict community lockdown, cross-jurisdictional mobilization of resources and officials’ sanction contributed to the eventual effectiveness of China’s response to the pandemic. I argue that the policy mix during crises should conform with rooted national policy style to be consistent and effective.

of traditional measures has turned out to be effective in response to the unprecedented policy problem of Why and how have traditional measures been chosen and worked effectively in this moment of crisis? This essay argues that pandemic as an external shock manifests the inconsistency in policies adopted by multiple policy actors and traditional policy measures that are more compatible with the deep-rooted policy style (Richardson, 1982; Howlett & Tosun, 2018) can be more consistent with each other (Howlett & Rayner, 2007; Kydland & Prescott, 1977) ; and hence more effective in coping with crises. A policy style, similar to an individual's style, constrains and refines the policy choices and policy goals. Similar to stylish behaviors of an individual reveal themselves during special times, policy style as a collective of government's behaviors does not manifest itself in all scenarios; it becomes more salient during the crises. In the case of China's fight against COVID-19, the policy mix that has worked is not necessarily a set of best practices that could or should be transferred to other countries. Instead, it is a policy mix compatible with the policy style of China featuring a centralized leadership, bureaucratic mobilization and memories of the right policy mix of previous crises. The policies in this mix have been proven to be consistent with each other, which has resulted in its eventual effectiveness.

The paper proceeds as follows. I first briefly review the literature on the relation between the policy mix design and policy effectiveness to show why the compatibility with policy style of a policy mix is crucial for policy consistency and hence effectiveness. I then use the case of China to show both how inconsistency in the policy instruments deployed contributed to the chaos at the early stage and how consistency was achieved when a policy mix that conforms with the policy style was devised. I conclude with a discussion on what lessons can be learned from China's experience in fighting COVID-19.

The concept of policy mix, meaning a portfolio of policy tools (Howlett, 2009) , is often deemed self-explanatory (Flanagan, Uyarra, & Laranja, 2011) . Although a formalized account of policymaking may suggest a one-t0-one relation between policy tools and policy ends, more often we observe that a portfolio of policy tools is deployed to address one policy issue. Several reasons account for the use of a multiplicity of policy instruments. First, the complexity of a policy issue often entails attempting to pursue multiple policy ends at once. In the case of fighting pandemics, these multiple policy ends include, e.g. saving more lives, maintaining social stability, sustaining the economy, etc. One policy instrument could not address all these objectives at the same time. Second, the volatile relation between means and ends (Lindblom, 1959 (Lindblom, , 1979 rationalizes the adoption of a policy mix. Without prior information about the effectiveness of certain policy means, adopting a mix of policy instruments is a rational move. Third, the multilevel governance context complicates the policy space, as various policy actors have multiple and often different policy instrument preferences or even policy ends. A 'complex multilevel policy mix' (Howlett & Del Rio, 2015) is then the result.

The adoption of a portfolio of policy tools, despite aiming to achieve the policy ends effectively, itself brings about new problems. Multiple policy tools may work in conflict with each other or even cancel out each other's effects. How to achieve synergies and complementarities in the policy mix, or the so-called 1 + 1 ≥ 2 goal, is then a crucial question for policy designers. Howlett and Del Rio (2015) have suggested three guiding principles in designing a policy mix: to avoid over-and under-designing, to avoid conflict and promote complementarities and synergies among policy tools, and to promote patching as well as packing in portfolio design.

However, principles for policy design during a crisis should be different from those during normal times (Yang, 2020) . The imminence of the challenge brought by a crisis leaves a short time window for a policy designer to respond and adjust. Effectiveness should dominate efficiency in crises as the top priority. Hence, the second principle, i.e. to avoid conflict among policy tools, stands out as a more important principle than the other two under such circumstances. Facing a pressing policy issue, the immediate effectiveness of the policy mix should be a top concern for policymakers. Divergent policy instruments may aggravate policy problems by causing conflicts and confusion. In contrast, overdesigning is less of a concern in a crisis scenario, as costs are much less of a concern; and using a patching strategy in a layering (Thelen, 2004) fashion might be an excessively mild response to cope with the problems.

Conflicts among policy instruments may take various forms, e.g. policy incoherence (May, Sapotichne, & Workman, 2006) or policy inconsistency (Howlett & Rayner, 2007) , both of which refer to a conflict among policies in their ideas, objectives, effects, etc. I argue that policy inconsistency, which has a temporal dimension, is a better term for conflicts in the policy mix during a crisis, when the policy problem is fast-evolving. As defined by Kydland and Prescott (1977) in their seminal work, a policy being consistent means that a policy at a certain time period can maximize the agreed-upon social objective function taking as given a previous decision. In a time of crisis, policymakers are pressed to adopt in sequence multiple, and often excessive (Maor, 2019) , policy instruments in response to the fast-evolving situation. If policy instruments adopted later contradict those adopted earlier, policy inconsistency occurs, and ineffectiveness follows.

In a crisis moment, consistency of the policy mix is a hard goal to achieve. The pandemic, as an exogenous shock, is likely to reveal more inconsistency which may have been tolerated during the normal time. The emerging literature on government responses to the COVID-19 summarizing various policy practices around the world has documented how politicians, local governments, agencies have made policies undermining each other (Maor, 2020; Carter & May, 2020; etc.) . The key challenge is that multiple policy actors are required to take policy actions toward a new and pressing policy issue on the ground of multiple ideas and interests formed in the normal time. Conflicts among these policy actions therefore cannot be neglected in a 'if it ain't broke don't fix it' fashion.

The distinctive policy style of a national government could serve as an anchor to reconcile conflicts among policy actions of multiple policy actors in crises. By definition, policy style refers to 'a typical way of doing business which is both institutionally and psychologically rooted' and 'a set of political and administrative routines and behaviors heavily influenced by the rules and structures of the civil service and political system in which is located' (Howlett & Tosun, 2018) . In a policy system, multiple policy actors making policies at multiple levels may also have their own 'typical ways' or 'routines and behaviors' in accordance with their own interests and ideas, which is dubbed as policy logics here. The relation between various policy logics and a distinctive policy style in a policy system is twofold. On the one hand, various policy logics of policy actors are not identical with the policy style due to the multiplicity of ideas and interests. On the other hand, the policy style is the common denominator of various policy logics as the former are shaped by the most rooted characteristics defining a policy and political system. When various policy logics have generated irreconcilable conflicts and policy problems persist, to rally behind the unified policy style is then a feasible and acceptable option for all policy actors. It is feasible in the sense that the institutional arrangement in which the policy style is rooted allows certain policy instruments to be chosen and implemented; it is acceptable in the sense that various policy actors may all perceive and expect that certain policy instruments should be adopted.

In a comparative perspective, China's policy style distinguishes itself from other countries. Defined by the two dimensions of the inclusiveness of decision-making and key policy actors, China's policy style is a type of 'closed-centralist', in contrast to those of the 'representative democracies, participatory or consultative democracy, and competitive electoral authorities' (Howlett & Tosun, 2018) . By looking into policy-politics connection in China, Qian (2018) figures out that the strict top-down bureaucratic management stands out as the most important institutional factors affecting China's policy style. While both capturing salient institutional characteristics of China's political system, it should be admitted that these characteristics does not exclude the existence of other opposite policy logics like decentralization (Montinola, Qian, & Weingast, 1995; Xu, 2011) and professionalism (Rothstein, 2015) in China's policy system. However, these institutional characteristics may manifest themselves during crises. In addition, a psychological aspect of policy style should also be noted. Heavily stricken by SARS in 2003, China's policy style has been affected by the memories of policy practice in dealing with a similar pandemic in the past. The institutional and psychological aspects together constitute China's policy style manifested in its coping with the COVID-19 pandemic. In the next three sessions, I show exactly how China has rallied on its policy style and provided a consistent policy mix after the early chaos.

China's early response to the pandemic has been widely criticized. The focal point of the critique is that the government has hidden information from the public and has not responded to the imminent threat. A close look shows that such a critique may have mistakenly been applied to the Chinese government as a whole and may have confused mishandling with inaction. When looking at the actions of local government of Wuhan and the National Health Commission (NHC) separately, one may find that each offered quick responses to then-unknown virus. The real problem, however, is that they both responded with their own policy logic nested within the multilevel authoritarian governance structure of China, which is characterized by decentralization arrangement in a centralized authority. In short, although both policy actors responded, the inconsistency in their responses led to chaos.

The local government of Wuhan took prompt action to deal with the unknown virus but with a dual goal of containing the virus and maintaining the normal order. Thanks to the epidemic alert system constructed after the 2003 SARS pandemic, the Hubei and Wuhan government received early warnings of 4 abnormal pneumonia cases on 26 December 2019 (The State Council Information Office, 2020). An epidemiological survey was then conducted on 29 December. On 2 January 2020, the government quickly shut down a large local wet market that was related to most of the earliest cases and was believed to be the origin of the unknown virus transmitted from animals to humans. From 31 December 2019, the Wuhan Health Commission held daily briefings about the new cases of 'unknown pneumonia'. Together with all these prompt responses, the local government also tried its best to maintain the normal political and societal order, especially because of the important upcoming 'two sessions' (lianghui) political events, that is, the annual sessions of the local People's Congress and the local People's Political Consultative Conference, held at the provincial and municipal level. No new cases were reported in the daily briefings in the period of 6-10 January and 11-17 January, when the 'two sessions' of Wuhan and the 'two sessions' of Hubei were held.

The response of the National Health Commission was also quite prompt. The NHC dispatched an expert team to Wuhan on 31 December 2019, 2 days after Wuhan's epidemiological survey. Starting on 3 January, the National Center for Disease Control (NCDC) started to exchange information about the virus and epidemic with its counterparts in the US and WHO. Inconsistency arose, however, between the local government and NHC in their assessments of the epidemic. On the first day of Wuhan's 'two sessions', 6 January 2020, the NCDC issued a level-two response to the public health emergency. On 15 January, during the Hubei provincial 'two sessions', the response level was upgraded to level one, while the local government announced no new confirmed cases. An even more salient example of the inconsistency between the NHC and the local government was observed on the crucial question of whether the new epidemic could be transmitted person-to-person. It seemed that the health commission intentionally avoided giving a definite answer but repeated the scientific jargon that 'no solid evidence has been found to prove person-to-person transmission' 8 before January 2o. On that day, Zhong Nanshan, a legendary doctor who earned his reputation fighting SARS in 2003, made a public statement on a national TV program based on his own judgment that it was most likely that the novel coronavirus was transmitted person-to-person.

In retrospect, both the Wuhan government and the NHC responded to the pandemic quickly in accordance with their normal policy logics based on their roles in China's multilevel authoritarian system. A defining characteristic of the Chinese authoritarian China's governance system is its de facto decentralization in a centralized authoritarian system (Chung, 2000; Montinola et al., 1995; Xu, 2011) . That is, although the central government can wield authority over the local governments and functionary departments, the authority itself is fragmented and disjointed, dubbed by Lieberthal and Lampton (1992) as fragmented authoritarianism. In the central-local authority relation, the center has largely devolved policymaking on local affairs to local governments while managing them through ex ante policy targets and ex post appraisals (Mei & Pearson, 2017; Tsui & Wang, 2004) . At the same time, local governments do not have a relationship of direct accountability with their local constituencies. In such an institutional arrangement, meeting the multiple goals set by the central government by any means becomes the primary policy logic of local governments (Zhou, 2010) . With this logic in mind, it is no surprise to observe that the local governments of Hubei and Wuhan scrambled both to achieve the goals of pandemic control and to maintain the normal order at the early stages of COVID-19.

The policy logic of the NHC as a functionary department is also deeply rooted in the fragmented authoritarian arrangement. As the central government relies heavily on horizontal local governments for fulfilling policy goals, the latter have a higher status than the vertical functionary departments in China's policy system. Nominally, the NHC is the highest authority on health-related policy issues in a unitary system. However, the decentralization arrangement generates dual leadership on local health policy issues. On the one hand, the NHC guides the local health department in professional affairs, for example, national standards and regulation policies; on the other hand, the local health department is a component of local governments that manages the NHC's local human resources and budgets. As 'direct management' obviously has a higher weight than 'professional guidance' (Mertha, 2005) , the professionalism of the NHC is often observed to be compromised by the interests of local governments. Buck passing and blame avoidance (Ran, 2017) turn out to be a more reasonable policy logic for functionary departments. At the early stages of China's COVID-19 outbreak, the NHC's prompt response and the reluctance to speak out against the local governments embodied this policy logic.

For the general public, the game changed on 22 January, when Wuhan announced an unprecedented and strict lockdown of the whole city. All public transportation inside Wuhan was suspended. Airports, railway stations and highway exits were closed to shut down all outbound traffic from Wuhan. Clearly, such an extreme and costly policy move by the Wuhan government did not conform to the aforementioned policy logic of local government. What happened?

It became clear 2 days later that the central government had intervened with a heavy hand. On 25 January, New Year's Day in China's lunar year, the standing committee of the politburo convened to discuss the pandemic, which signified that the central government had taken the lead in fighting COVID-19. Later, the central government disclosed that it had stepped in from 20 January, when Zhong Nanshan confirmed person-toperson transmission on national TV (The State Council Information Office of the People's Republic of China, 2020). Considering how tightly propaganda machines are controlled in China, Zhong's public statement was most likely a move approved by the highest authority to signal the center's intervention.

Several major moves in fighting COVID-19 were announced after the 25 January politburo meeting. All these moves signified that the central government aimed to break the existing policy logics of both local governments and the NHC. In terms of the guiding principle of policymaking, the meeting clearly stressed the centralized and unified leadership of the party center in fighting COVID-19. One quite unusual move showing the determination of the center was that the paramount leader Xi Jinping openly announced several times after the politburo meeting that 'I by myself command and steer' (qinzi zhihui, qinzi bushu) the fight against COVID-19. For central-local coordination, the central government dispatched a central leading group headed by a vice premier to Wuhan on 27 January, which has since been stationed in Wuhan for three months to lead local governments in handling the epidemic. For the functionary departments, the State Council Joint Prevention and Control Mechanism (hereafter referred to as the 'joint mechanism') was created on 20 January and headed by the same vice premier who headed the central leading group to Wuhan. The NHC served as the coordinating department for the joint mechanism.

These moves together shifted the policy logics of both the local governments and the NHC, which were aligned with fragmented authoritarianism. For the local government, the policy logic changed from meeting multiple policy goals by any means possible to meeting the sole policy goal of containing the pandemic by any means necessary. In fact, the role of the Hubei and Wuhan local governments shifted to become pure implementers as decision power was shifted upward. Several extreme policy measures have been implemented in Wuhan since then. After the city lockdown, Wuhan immediately launched construction of two temporary hospitals with 1,000 and 1,600 beds, respectively. Both hospitals were built in 10 days from scratch and opened on 4 February and 8 February, respectively, eventually treating 5,060 patients before both were closed on 15 April. Starting from 4 February, the Wuhan government followed experts' advice and started to build fangcang hospitals (temporary treatment centers). The novel idea of the fangcang is to 'convert existing public venues to temporary health-care facility which could isolate patients with mild to moderate symptoms from their families and communities while providing medical care, disease monitoring, food, shelter and social activities' . The Wuhan government quickly appropriated 16 sports arenas or convention centers and converted them. The first fangcang opened on 5 February. In the end, over 13,000 patients were admitted to and treated in fangcang before all 16 were closed on 11 March as the pandemic dissipated .

For the NHC, its policy logic also shifted with the backing of the powerful joint mechanism. The NHC is no longer a disadvantaged professional leader in a system dominated by horizontal local governments. A cross-jurisdictional collaboration mechanism was soon built to move resources from other provinces to Hubei and Wuhan. Policy consistency was observed between the local government and the NHC. While Wuhan was building temporary hospitals at an extraordinary speed, the NHC was sending requests to all provinces and the military to send medical personnel to support Wuhan's hospitals. On Jan 24, the first batch of medical personnel from Shanghai and Guangdong arrived in Wuhan, each sending approximately 135 medical workers as requested by the NHC. 9 By Jan 28, over 30 medial teams of 4,130 medical workers from other provinces and the military started to work in the hospitals of Wuhan and Hubei. 10 As the epidemic evolved, on 7 February, the joint mechanism announced a program named 'one province for one city' (yi sheng bao yi shi). This crossprovincial collaboration mechanism program involved 16 provinces, each of which helped one Hubei city with medical personnel and equipment (Hu, Zhang, Kapucu, & Chen, 2020) . For example, the province of Jiangsu was designated to help Huangshi, a city with 1,015 confirmed cases. For a period of 22 days, 362 medical workers from Jiangsu worked in Huangshi's hospitals. 11 According to the latest statistics, the NHC was able to mobilize 346 medical teams with over 43 thousand medical personnel from other provinces and the military to help in Hubei's hospitals and inspection facilities. 9 http://news.cctv.com/2020/01/28/ARTIEnQtOuPRYZJqMAUUd7qe200128.shtml?spm = C94212.P4YnMod9m2uD.

ENPMkWvfnaiV.115, accessed on 20 June 2020. 10 http://www.chinanews.com/gn/2020/01-28/9071516.shtml, accessed on 20 June 2020. 11 http://news.xhby.net/index/202,003/t202,00328_6580091.shtml, accessed on 20 June 2020.

From the early chaos to the onset of consistency, the shift from various policy logics to the policy style of centralized leadership marked the crucial turning point. For the local governments, the policy logic changed from that of a trustee (Hood & Lodge, 2006) overseeing multiple policy goals to that of an agent with simple implementation tasks. For the NHC, the policy logic changed from that of a disadvantaged bureaucratic department to that of a professional organization tasked with allocating resources effectively. Neither new policy logic would have been possible in the fragmented authoritarian system. The fast recentralization after the early chaos was decisive for the shift in policy logic and the eventual consistency achieved between the policy instruments adopted by the local governments and the NHC.

The difficulty faced in making the shift from multiple policy logics to the unified centralized leadership, however, should not be taken for granted even in an authoritarian state such as China. Even after the center's clear declaration of recentralization, some localities still stuck to their normal policy logic. For example, Dali, a southwest city of Yunnan Province, was reported to seize imported medical supplies headed to Hubei Province for its own use on 2 February 2020. The Dali leader was soon dismissed by a top-down order on 6 February. 12 It is therefore worth noting that recentralization worked in China because it conforms with policy style nested in China's institutional context, in the case here meaning the preexisting authority of the central government in managing its local leaders.

Many have noticed that government capacity is a necessary condition in crisis management and specifically in the COVID-19 pandemic (Christensen & Laegreid, 2020 ). However, it should be noted that government capacity fit for normal management is by no means equivalent to capacity during a crisis. A crucial capacity for crisis management is the ability to mobilize a large amount of resources in a short time to the most needed place. In China, the government adopted strict measures to lock down communities in early February, which aimed to cut the transmission channel and was consistent with the previous policy of admitting all confirmed cases to the recently constructed temporary hospitals and treatment centers. However, the strict community lockdowns required a large corps of community workers to be mobilized and dispatched. It was the China's policy style of effective bureaucratic mobilization that made the strict community lockdowns possible.

Compared with that of other countries, the Chinese version of community lockdown is prompt and strict. From 1 February, several major cities issued administrative orders for community lockdown. All provinces and autonomous regions except Xizang followed suit by mid-February. The strictness level of the community lockdowns varied in China. The strictest version was adopted in Hubei Province from 16 February, after all temporary hospitals started to operate. All residents were forbidden from leaving their residential communities, and all medical and daily necessities were delivered by community workers. 13 12 http://www.chinanews.com/gn/2020/01-28/9071516.shtml, accessed on 20 June 2020. 13 The strictest community lockdown was implemented in Wuhan for 52 days until 8 April.

Milder versions of community lockdown did not forbid all residents from going out, but registration and temperature checks were strictly implemented. Observers have already commented on why these strict lockdown measures were possible in China. From a culturalist perspective, some cited the obedience rooted in Chinese people to explain their compliance and endurance even when necessary errands were forbidden. 14 Others cited the paramount authority of government and claimed that 'people will do whatever the government told them to do'. 15 Both accounts may underestimate the difficulty of enforcing the community lockdowns in China. As authoritarian as China is, the reach of the state also has its limits. Indeed, China has extended state to society at the grassroots level via Communist party branches in communities or villages in urban and rural areas. In the absence of autonomous civil society at the grassroots level, the government also organizes semiautonomous community organizations called residential/villager committees through free elections to manage community/village affairs. In total, there were over 100 thousand such semiautonomous community organizations in urban areas and over 550 thousand in rural areas. 16 In the urban area, the government even went further to divide communities into smaller grids for the purpose of close monitoring and management. However, a large body of literature has documented the weakness of community management in China, especially in urban areas (Wei, 2003; Wu, 2019) . As extension of the state apparatus, urban community committees have still served as 'the last stop' for the delivery of some public services like subsistence allowance and unemployment benefits; However, the key challenge to the authority of community committees is that the majority of urban residents do not depend on them for more crucial public services such as education, health and retirement benefits in a marketized and liberalized authoritarian system. In fact, even during the planning economy period, most residents received their welfare benefits from their work unit instead of their community (Walder, 1986) . In the reform era, most people live in apartments bought or rented from the market instead of allotted through their work units, which further attenuates the role of community organization in public-service delivery. With the gradual socialization of public-service delivery and deregulation (e.g. the revocation of the birth control policy) of societal control, the connection between government-controlled community organizations and residents has become even weaker (Wu, 2019) .

Capacity-wise, government-controlled community organizations are not strong, either. In a Beijing township with 10 communities with household counts ranging from 1,000 to 4,000, the number of staff members working in each community organization is less than 20 and ranges from 11 to 19. 17 This understaffing problem is even more severe in localities outside of Beijing. In a Wuhan community with over 8,000 households and 20,000 residents, the community organization has only 23 staff members. 18 A clear capacity gap manifested itself during the community lockdowns. As the extension of the state apparatus, the semiautonomous community organizations were the only reliable implementers of the strict measures of the community lockdowns. Multiple tasks were assigned to the community organizations. In a Beijing community that I interviewed, the job list of community workers included the following: completing a household survey of all residents, contacting close-contact cases, responding to residents' requests from the government hotline 12,345, responding to frequent visits and checks by government officials, guarding entrances and conducting temperature checks, disinfecting public areas, providing meals service to residents with special needs, surveying nearby businesses, conducting public education on the epidemic, and completing normal daily works. 19 In Wuhan, the job lists for community organizations were even longer. At the early stage of the outbreak, when hospitals quickly became squeezed, community organizations in Wuhan served as the pivotal hub to transport severe cases to hospitals. 20 After the adoption of the strictest community lockdown measures, community organizations in Wuhan had to take over the task of meeting the daily needs of thousands of residents who were forbidden from leaving the community.

Clearly, the community organizations could not assume such daunting task lists by themselves. A special arrangement was made by the government to temporarily designate officials from various government organs and public institutes to communities. In Wuhan, the number of 'squatting officials', or Xiachen Ganbu, meaning officials sent from upper-level governments to the communities, amounted to 44,500 by 28 February. 21 For the aforementioned Wuhan community with 23 community workers serving over 20,000 residents, over 100 squatting officials were designated. 22 In an urban district in Beijing with a population of over 3.8 million, the government was able to mobilize 2,500 ~ 3,000 squatting officials daily to help the 3,200 normal community workers from the end of January to the middle of April. 23 The policy measure of designating squatting officials to communities was crucial to the effectiveness of the community lockdown policy measure in China. Not only did squatting officials directly strengthen the manpower in the community; they also encouraged volunteerism in the community. Community participation during crises is a typical assurance game of collective action. Having sufficient participants at the early stage is key to reaching the collective optimal (Chong, 1991) . Squatting officials, as external support, greatly increased the perceived possibility of the collective optimum being reached and hence encouraged more participatory behaviors. In the aforementioned Wuhan community, the 23 community workers together with over 100 squatting officials eventually organized a team of over 400 volunteers and secured adequate delivery of public services to the community, which was fully locked down for approximately 2 months. 24 It would be naïve to assume that squatting officials went to communities because of obligations or voluntarism. The Chinese government's capacity to mobilize such a large group of officials for temporary tasks has been a rooted element of Chinese policy style and was the backbone of the consistency of the community lockdown measures. Following the Leninist party tradition, the Communist Party manages its elites through a unified cadre management system (Burns, 1994) . The idea of the system is to cultivate cadres' loyalty towards the party with strong incentives of rewards and punishment (Mei & Pearson, 2014 , 2017 Rothstein, 2015) . While conflicts between professionalism and loyalty do emerge in China's current policy system (Rothstein, 2015) , loyalty prevails over professionalism during crisis management. For the squatting officials, their primary motivation in joining the teams of community workers was to fulfill a duty assigned by their work units, no matter what their daily tasks were. As put by the head of a Beijing community organization, 'the squatting officials have been really helpful and we did not have to take pains to organize or manage them; they came here just like they went to their own offices'. 25

Many have noticed the role of policy memory in countries' fight against COVID-19 (Moon, 2020) . One telling piece of evidence is that East Asian governments all have responded promptly to COVID-19, and citizens have also shown higher compliance with government policy measures compared to their counterparts in other regions. 26 A reasonable guess is that learning from the 2003 SARS outbreak in East Asia facilitated the prompt response and compliance. I, however, contend policy imitation from past experience, i.e. 'copying the actions of another in order to look like that other', instead of policy learning, i.e. learning based upon 'evidence of success of the policy' (Shipan & Volden, 2008) , is a better depiction of the actual practice in China's fight against COVID-19. Memory of past policy practices constitutes a psychological aspect of policy style and defines 'what is the right thing to do' and hence may bring about effectiveness in a selfrealizing fashion. To imitate policy instruments conforming with such psychological elements of policy style makes the measures more likely to be accepted by multiple policy actors and results in more consistent compliance.

A telling case in China's response to COVID-19 is the removal of provincial leaders in Hubei Province and Wuhan City on 13 February and the ensuing escalation of the response measures. On 13 February, the center dismissed the party secretaries of Hubei and Wuhan, Jiang Chaoliang and Ma Guoqiang, respectively, and replaced them with former Shanghai mayor Ying Yong and former Jinan party secretary Wang Zhonglin. Holding leading officials accountable for undesirable policy outcomes has been a common practice in China since the 2003 SARS crisis, in which the Beijing mayor and health minister were both removed amid intensive criticism of a sluggish response to the epidemic. Other senior officials were removed in later years because of workplace safety accidents, food and drug safety problems, overheated investment, social unrest, etc. The primary goals to sanction these leading officials, however, are not to punish them for their personal wrongdoings but rather to stimulate other officials to work harder to achieve better outcomes and to appease the general public (Mei & Pearson, 2014) .

The replacement of the leaders of Hubei and Wuhan had an immediate effect. The aforementioned strictest community lockdown measures were adopted on 16 February, 3 25 Interview, April. 22. Head Z of a community organization in Beijing. 26 Kathrin Hille & Edward White. 2020.3.16. Containing Coronavirus: Lessons from Asia. Financial Times days after the leaders' replacement. From 17 February, another rigorous measure was adopted to screen all residents in Wuhan to achieve the goal of 'five 100%', meaning admitting 100% of confirmed cases to hospitals, administering nucleic acid tests to 100% of suspected cases, testing 100% of febrile patients, isolating 100% of close contacts, and locking down 100% of communities (The State Council Information Office of the People's Republic of China, 2020). The 'five 100%' measure was a daunting move considering that the population of Wuhan amounts to over 10 million and the time frame for implementation was set to 3 days. However, it was later proven that these measures were implemented effectively and were crucial to the final containment of the pandemic.

The question, however, is what was the real mechanism connecting the sanctions of officials with policy effectiveness. It should be noted that the sanction of the Hubei and Wuhan leaders during COVID-19 departed from previous cases of holding officials to account, in which sanctions of officials were used to promote compliance. For example, in the case of the 2003 SARS crisis, the sanction of the Beijing mayor and health minister occurred at the outset and was used a signal that the government would respond to the epidemic with rigorous measures. In contrast, the sanction of the Hubei and Wuhan leaders occurred over 20 days after rigorous measures were taken at both the central and local levels. By the time of their removal on 13 February, the local leaders had managed to accomplish several daunting tasks, such as implementing city lockdowns and constructing two temporary hospitals and a dozen temporary treatment centers in a very short time period. A counterfactual rejoinder could be that the epidemic would have been contained even if the local leaders had not been sanctioned. Why did the central government still sanction them after rigorous measures had been undertaken?

Given the murkiness of Chinese politics, the actual deliberations over whether to sanction the two local leaders may never be known. A reasonable guess, however, is that policy memory of sanctions of officials in previous crises played a pivotal role in the sanction decision this time. In fact, both the general public and government officials were expecting the sanction to come as complaints and grudges against the local governments of Hubei and Wuhan started to accumulate at the outset of the outbreak of COVID-19. On 25 January, a local journalist even publicly called for the replacement of the Hubei and Wuhan leaders on social media. 27 On 27 January, the mayor of Wuhan also publicly admitted on a TV program that he was willing to resign if it would help contain the epidemic. 28 Given such expectations rooted in policy memories as part of policy style, the sanctions were probably the most reasonable choice for the central government to make.

What can be learned from China's fight against COVID-19? This paper does not suggest that a specific policy mix is the most effective and hence should be copied by other countries. Instead, I argue that having a policy mix conforming to the national policy style is crucial for the consistency and hence the effectiveness of multiple policy instruments chosen during times of crises. A real challenge of policymaking posed by a crisis is that multiple policy instruments chosen by multiple policy actors based upon normal policy logics tend to be inconsistent with each other. China's early chaos in response to the pandemic embodied such inconsistency and ineffectiveness. However, the outbreak of crisis did not leave sufficient time for policy actors to experiment and determine new and sufficiently stable policy logics to yield a consistent policy mix. In response to the crisis, policy actors should look to the rooted policy style to design their policy mix. In China's case here, I find that a traditional policy mix has worked consistently and effectively because it conforms with China's policy style characterized by a centralized leadership, bureaucratic mobilization capacity, and policy memories about what constitutes the right measures to handle a pandemic.

This finding has theoretical implication on the relation among policy style, policy logic and policy mix. A policy mix, as 'a long-lasting arrangement of policies that have accumulated over time' (Howlett & Tosun, 2018) , is often framed by a policy logic of a specific policy actor. Conflicts between multiple specific policy logics and the underpinning policy style exist in all policy systems and are often tolerated during normal times. Crises requires such conflicts to be reconciled, even just temporarily, in the short time window. As a result, to have the underpinning policy style replace the multiple contradictory policy logics is conducive to policy mix consistency. A remaining question, however, is that whether such replacement will persist after crises. Further study should look into the possible changes in policy logics of multiple policy actors after the pandemic. In retrospect, the 2003 SARS crisis in China has witnessed similar replacement of underpinning policy style for multiple policy logics (Thornton, 2009 ). Its long-term effect on policy logics of multiple actors is at best inconclusive. On the one hand, some recentralization arrangements, e.g. holding local leaders accountable for undesirable policy outcomes, have become part of the policy style of China and have shifted policy logics of local players, which has also contributed to the effective containment of COVID-19 this time. On the other hand, the policy logics formed in the decentralized decision-making such as inter-agency bargaining, fragmented authority still persist.

Two caveats are in order. First, an interesting discrepancy in assessments of China's response to COVID-19 is between the positive appraisal of China's effectiveness in containing the pandemic and the blame heaped on China for using excessively heavyhanded policy instruments. This paper suggests that such discrepancies should be reconciled, as effectiveness and heavy-handedness were two sides of the same coin in China's policy style. For an authoritarian system with emerging liberalization, a more feasible and effective way to respond to a sudden crisis such as COVID-19 is probably to return to the traditional measures that conform with its rooted policy style instead of moving forward in uncertainty. To have a policy mix consistent with the authoritarian context, overreaction is probably a necessary evil. Second, effectiveness should not be equated with efficiency. The effectiveness of China's response to COVID-19 was achieved at huge economic and social costs, as the state has dominated the market and society. In no way does this essay suggest that the predominance of the state, though rooted in China's policy style, is an optimal solution for policy problems in normal situations. organized by Jae Moon at Yonsei University. My sincere appreciation also goes to interviewees for this research, most of whom have been fighting against the pandemic at various fronts since January.

Strengthening central CCP control of leadership selection: The 1990 Nomenklatura. The China Quarterly

Disciplining local officials in China: The case of conflict management

Making sense of the US COVID-19 pandemic response: A policy regime perspective

A policy mix for environmentally sustainable development-learning from the Dutch experience

Fangcang shelter hospitals: A novel concept for responding to public health emergencies

Coproducing responses to COVID-19 with community-based organizations: Lessons from Zhejiang Province, China

Collective action and the civil rights movement

Balancing governance capacity and legitimacy -how the Norwegian government handled the COVID-19 crisis as a high performer

Central control and local discretion in China

Chronicle of Events for COVID-19

Reconceptualizing the 'policy mix' for innovation

Scenarios for a robust policy mix: The final report of the German study commission on sustainable energy supply

The politics of public service bargains: Reward, competency, loyaltyand blame

Governance modes, policy regimes and operational plans: A multi-level nested model of policy instrument choice and policy design

The parameters of policy portfolios: Verticality and horizontality in design spaces and their consequences for policy mix formulation

Design principles for policy mixes: Cohesion and coherence in 'new governance arrangements

Policy styles and policy-making, exploring the linkages

Hybrid coordination for coping with the medical surge from the COVID-19 pandemic: Paired-Assistance programs in China

Rules rather than discretion: The inconsistency of optimal plans

Bureaucracy, politics, and decision making in post-Mao China

The science of" muddling through

Still muddling, not yet through

Technology empowerment, digital drive and accurate pandemic control. Social science weekly. 1695

Strategic policy overreaction as a risky policy investment

Policy over-and under-design: An information quality perspective

Policy coherence and policy domains

Killing a chicken to scare the monkeys? Deterrence failure and local defiance in China

The dilemma of "managing for results" in China: Won't let go

China's "Soft" centralization: Shifting Tiao/Kuai authority relations

Federalism, Chinese Style: The political basis for economic success in China

Fighting against COVID-19 with agility, transparency, and participation: Wicked policy problems and new governance challenges

Policy styles in China: How to control and motivate bureaucracy

Understanding blame politics in China's decentralized system of environmental governance: Actors, strategies and context. The China Quarterly

The Chinese paradox of high growth and low quality of government: The cadre organization meets Max Weber

The mechanisms of policy diffusion

Big data contributes to prevention and control of coronavirus in Zhejiang

The State Council Information Office of the People's Republic of China

How institutions evolve

Crisis and governance: SARS and the resilience of the Chinese body politic

Between separate stoves and a single menu: Fiscal decentralization in China

Communist neo-traditionalism: Work and authority in Chinese industry

The mode of our urban community administration: development and system innovation

Integration of governance power, gravity down of services and selective participation: The form of "policommunity complexity" of urban community governance in China

The fundamental institutions of China's reforms and development

Unprecedented challenges, familiar paradoxes: COVID-19 and governance in a new normal state of risks

Is information technology effective in controlling the pandemic? Beijing: The Beijing News

Exploring the administrative mechanism of China's paired assistance to disaster affected areas program

The institutional logic of collusion among local governments in China

The author is extremely grateful for suggestions on earlier versions of this paper from Michael Howlett and Ramesh M., the editors, the two very helpful reviewers and participants in the webinar

No potential conflict of interest was reported by the author.