Political dynamics behind different policy options: Long-term care insurance policymaking in Beijing and Shanghai from the perspective of local policymakers

Abstract Facing ageing challenges, long-term care (LTC) has become a focus of policymaking and policy analysis in China. However, the burgeoning literature obscures a lack of understanding of LTC policymaking, implying a linear and neutral process. This depoliticisation is unrealistic and contributes little to understanding the diversity of LTC policies and improving inclusive LTC provision. Focusing on the core institutional arrangement, LTC insurance (LTCI), this article explores a highly politicised policymaking process and reveals complex political deliberations behind different LTCI choices across regions. Underpinned by the multiple streams approach, which supports a systematic comparison of the policy process, this article identifies four key factors from a relational perspective that influence LTCI policymaking, including the tension between evidence and politics in the construction of LTC issues, the tension between policy effectiveness and stability in the assessment of policy options, strong or weak political will, and the presence or absence of municipal government.


Introduction
Population ageing is a global phenomenon, but ageing processes vary across regions and countries.In China, the fundamental problem of ageing lies in "poverty, disease and disability risks" (Dang, 2015, p. 7), so long-term care (LTC) provision has become one of the most critical social issues facing governments.However, LTC did not enter mainstream policymaking until 2016, when the National Committee on Ageing (NCA) released the Fourth Survey of the Disabled Elderly, showing that the number of disabled seniors exceeded 40.6 million and their financial and service conditions were worrying (Zhang, 2016).
The survey drew attention to disabled seniors from the central government, triggering a wave of LTC-related policies that instructed a joint LTC financing system combining three funding mechanismssocial welfare or social assistance, social insurance, and private insurance; and the provision of LTC services through health-social care integration (Yiyangjiehe).Accordingly, policy pilots on funding mechanisms, service organisation and delivery, and skilled workforce have been launched nationally and locally.Given LTC Social Insurance (LTCI) as the institutional infrastructure (Wu, 2020), the LTCI pilot launched by the National Health Commission (China NHC) in 15 sites in 2018 (the national pilot) has received the most attention, with an expansion to 49 cities in 2020.
With policy proliferation, the literature on LTC has increased substantially.Early research revolves around estimates of LTC needs and overseas experience of LTCI (Hu et al., 2015;Du et al., 2016;Ling and Dong, 2019) to prove the imperatives and the feasibility of establishing LTC institutions (Wang and Zeng, 2015;Zhao and Han, 2015;Dai, 2016), and LTC service delivery and related issues (Luo, 2014;Su et al., 2015;Peng et al., 2017).Driven by policy pilots, recent research has focused on LTCI system construction and diverse LTCI models; effects and problems; and policy instruments and proposals (Sun et al., 2019;Deng, 2021;Zhu et al., 2021;Gui, 2022;Hu, 2022).Among these, financial sustainability has received the most attention, for example, state fiscal capacity for LTCI and fiscal balance and deficit risk (Lu et al., 2019;Li, 2020;Chen et al., 2021).
The rich literature reflects the diversity of policy issues and practice, but is mainly descriptive in nature.It avoids serious debates about the political dynamics of LTC policy, leaving policy processes as a "black box."By emphasising LTC needs, policies, and their effects, the existing literature suggests a straightforward and neutral policymaking process that faithfully translates LTC needs into policies.This depoliticisation is illusory, as power is involved in defining the legitimate needs for policy intervention (Langan, 1998) and political will determines how the LTC burden is shared (Colombo et al., 2011;WHO, 2011).Thus, it hardly explains why LTC policies vary widely across China (the exception is Chan and Shi's (2022) work on social decentralisation in LTC reform).For example, Beijing and Shanghai differ considerably regarding LTCI pilot scope (district-or municipal-level), funding sources, coverage, insurance premiums and benefits, and services (Chen, 2021).
The very different local policy responses to central directives on LTC call for an investigation of locality-level policymaking.Furthermore, understanding the politics behind policy enables policymakers to use the levers of determinants to improve policy.In particular, LTCI is part of welfare arrangements characterised by stability (Øverbye, 1997), highlighting the importance of getting this high-stakes policy right in the first place rather than remediating it by lesson-learning.The research question is, therefore, Why do localities make significantly different choices under the same central directives?Answering this question requires an explanation of "how public decision-making works … why policy output and outcomes differ from place to place and across time" (John, 2013, p. 1).
This article reports findings from primary research underpinned by the multiple streams approach (MSA).It used the MSA as an analytical framework to examine and compare the LTCI policymaking process of two sample sites, Beijing and Shanghai.By uncovering this process, the article has generated insights into the politics underlying LTCI policy, with implications for policy analysis and development.Given the focus on local discretion, this article pays less attention to the impact of central steering.

Political determinants of policy choices
Policy studies and political science have produced a large body of literature on political behaviour and policy processes.Policy is usually regarded as influenced by actors/interests, institutions, and ideas, with power flowing through them.In addition, external events influence policy decisions or non-decisions, including media coverage, political, economic, and technological circumstances (e.g.finances), and others (e.g.new diseases) (Birkland, 1998;Pomey et al., 2010).

Power and actors, institutions, and ideas in policymaking
Power, defined by access to resources, for example, funds, capital assets, political connections and expertise (Shiffman, 2014;Ho, 2022), is most often identified as the key determinant of policy.However, power is dynamic and does not exist and operate in isolation.Despite different understandings of power through different theoretical orientations or frameworks, actors/interests, institutions, and ideas are useful concepts by which to understand the policy process and how power operates (Pomey et al., 2010;John, 2013;Sabastier and Weible, 2014).
Actors, who are thought to vary widely in terms of identification, rationality, and power (Howlett and Ramesh, 2003;Hill, 2005;Knoepfel et al., 2007;Sabastier and Weible, 2014), exercise power and make decisions.They can be regarded as self-interested utility-maximisers (Cairney, 2019), giving rise to competing, conflicting, and even contested interests in policymaking.As such, they are presented as "rational decision-makers" who critically and objectively review and choose between policy options (Griggs, 2006).However, in this process, powerful actors (e.g.social and political elites) can leverage other actors to advance their agenda (Shiffman, 2014;Gore and Parker, 2019).
Actors also need to be understood as acting within particular institutions, broadly defined as both formal organisations and informal but institutionalised norms, rules, precedents, and so on (Hall and Taylor, 1996;Howlett and Ramesh, 2003;Cairney, 2019).Institutions design power distribution, relationships, attention, and information flows (May, 2015;Peters and Zittoun, 2016;Dee et al., 2023), providing incentives or constraints for actors to act rationally and thus shaping actors' political behaviour (Béland and Waddan, 2015;Cairney, 2019).
Ideas, referring to knowledge or beliefs about what is and/or should be at three levelspolicies, programmes, and philosophiesplay a causal role in understanding policy change (Schmidt, 2008;Pomey et al., 2010).Beyond the exercise of power to pursue interests, actors, aided by ideas, deal with uncertainty in policymaking (Carstensen, 2011;Berman, 2013).Specifically, ideas shape actors' goals, motivations, and preferences, as well as the constraints and opportunities they face in particular contexts (Béland and Cox, 2010;Cairney, 2012;Berman, 2013).

Key determinants in Chinese policymaking
Turning now to look at how these determinants of policy play out in China.First, elitism has been the basic policymaking model in the post-1949 era (Wang and Hu, 2010;Liu, 2016); however, the original assumption that policymaking only occurs at the top has been challenged.Attention has thus shifted from the top leadership to bureaucrats (Hammond, 2013), recognising that political compromise within power elites characterises Chinese policymaking (Lampton, 1992;Wang and Hu, 2010;Duckett, 2019).
Third, rationality as an explanation of human behaviour is context-specific (Townley, 2008).In Chinese policymaking, rationality manifests itself as problem orientation and self-interest (Zi li xing).Policy problems are an important driver of Chinese policymaking, especially the logic of "problemoriented reform," that is, advancing reform by solving problems (Han, 2017;Wang and Ning, 2018), which critically influences Chinese politics.Problem orientation suggests rational policymaking in response to objective and identifiable problems (Xu, 2019).Since 1978, self-interest has become recognised as widely present in individual bureaucrats, functional departments, and the government (Dong, 2008).It is currently reified as maintaining stability, a multidimensional concept referring to social, political, and financial stability (Wu and Li, 2006;Zhang, 2006;Xing and Chen, 2007).
The emphasis on social and political stability stems from an increasingly heterogeneous Chinese society "full of anger, political consciousness, anxiety, and uncertainty" (Zheng, 2010, p. 28), and social unrest is considered a threat to social stability, economic growth, and the political regime.Thus, the primary task of cadre assessments is to maintain social stability (Lee and Zhang, 2013;Wang et al., 2018), making stability maintenance a concrete self-interest of Chinese bureaucrats and local governments.
Accordingly, attention to public opinion, which embodies the issues that matter to the masses (Kingdon and Stano, 2011), serves stability, as public opinion that has not been properly addressed causes "mass incidents (Qunti shijian)" or even social unrest and political instability (Zhao and Xue, 2017).
Maintaining fiscal stability often explains the constant misalignment between central and local governments in policy implementation.Finances constitute a basic orientation in Chinese policymaking (Lieberthal, 1992;Han and Kung, 2015).In particular, the asymmetric assignment of revenue and responsibilities creates tensions between unfunded central directives and stretched local budgets (Qian and Mok, 2016;Chan and Shi, 2022).Thus, when carrying out unfunded central mandates, local governments often strategically compromise the intentions of their superiors (Ngok, 2013;Guess and Ma, 2015;Qian and Mok, 2016).

Limitations of existing research on Chinese policymaking
To date, most studies of Chinese policymaking have either focused on describing the outcomes of a policy, leaving how policymakers' motives shape the policy remain underexplored (Duckett, 2019); or examined one component or aspect of the policymaking process, with little effort being made to systematically explore the interplay of multiple factors.Thus, the existing literature does not answer the question that requires tracing and explaining dissimilar LTC policymaking experiences in China: why do localities make significantly different choices under the same central directives?
While studies have focused on the role of political elites, studies on the role of other potential actors in the policymaking process, for example, activists, media, and professionals (Zhu, 2013;Cheng, 2014), remain scarce.This overlooks the increasing opportunities for non-state actors to participate in Chinese policymaking and extended government-society negotiations (Chan and Cabestan, 2001;Moore, 2014).Others have focused on the relative influence of central and local institutions (Nickum, 2010;Ahlers and Schubert, 2015;Zhang and Rasiah, 2016), which fail to explain divergent, inconsistent, or contradictory local responses to central directives within a unitary system.Moreover, they impede efforts to overcome institutional constraints to achieve equity and universalism (Lin, 2001;Saich, 2011;Ngok, 2013), a point highly valued by LTC.Furthermore, the role and influence of socially constructed realities, for example, policy problems and political interests, has received relatively little attention or recognition.A central political process of policymaking is to turn conditions into problems (Buse et al., 2012;Bacchi, 2016), while interest-based political behaviour involves interpreting interests (Campbell, 1998;Schmidt, 2008;Berman, 2013).
The MSA Drawing primarily on Kingdon's work, the MSA was used as the framework to inform and shape an investigation, overcoming the limitations of the existing evidence base noted above.Unlike univariate determinism, the MSA captures complex relationships between multiple causal processes, including interests, institutions, and ideas (John, 2003;Sabastier and Weible, 2014), providing a multifactorial explanation for different policymaking experiences (Cairney and Jones, 2016).
MSA states that a drastic policy change is the outcome of the convergence of three independent processes (so-called streams), at a critical moment (policy window) promoted by policy entrepreneurs.The problem stream is where certain conditions or issues are brought to the attention of policymakers and defined as policy problems.The policy stream is the process in which policy proposals are developed and reviewed against various criteria.Meeting these criteria significantly increases the probability of selection of a particular policy option (Kingdon and Stano, 2011).The political stream comprises the policy environment and decision opportunities it may afford (Ackrill and Kay, 2011).The intersection of two or all streams is the driving force for policy change (Ackrill et al., 2013), and it is vital to link a solution to the other two elements (Kingdon and Stano, 2011).
Convergence occurs when a policy window opens (Kingdon and Stano, 2011), referring to the opportunity that defines the context for policymaking (Zahariadis and Exadaktylos, 2016).Due to the time constraints of policymaking, its salient features are temporality and scarcity, requiring competent and resourceful policy entrepreneurs to seize the opportunity.Key to critical policy change, policy entrepreneurs have been broadly defined (Ackrill et al., 2013;Zohlnhöfer and Rüb, 2016); however, three qualities contribute to successful entrepreneurial action: expertise, the ability to speak for others, and an authoritative decision-making position (Kingdon and Stano, 2011).

Research design
This article reports part of a study on the implementation of national strategies for LTC provision in China.The study employed a qualitative approach to exploring the retrospective accounts of individuals directly involved in LTC policymaking in two Chinese localities.

The location of the research
Beijing and Shanghai were selected by their capacity to elucidate the policymaking process (Barbour, 2007) and analytic potential for comparison and generalisation.Their full-range responses to the central directives on LTC, including policies and pilots, reflect the richness of policy setting, meeting the criteria being a core research setting, "bounded, rich … microcosms" (Holliday, 2007, p. 36).
As municipalities, Beijing and Shanghai provide a wealth of information on the complex central-local dynamics of state policy implementation, as political debates between central and local governments occur most at the provincial level (Lieberthal, 1992).Furthermore, unlike other provincial-level governments, municipalities develop specific measures for district-level implementation, allowing research to explore their political deliberations.Beijing and Shanghai also have the potential to provide valuable insights into Chinese policymaking.Both are seen as role models of governance and pioneers of social policy, making them the primary source of policy learning for the rest of China.
However, Beijing and Shanghai show striking contrasts in LTC arrangements, making them wellmatched cases for instructive comparison.Compared to the rest of China, they are first-tier cities with more developed economies and greater fiscal capacity.Even compared with other Chinese metropolises, they face similar challenges from large ageing populations with high disability rates.However, Beijing opted for a system centred on targeted welfare and social care, while Shanghai established LTCI and an LTC-centric service system.
At the time of the fieldwork, Beijing and Shanghai were at different stages of piloting LTCI.Unlike Shanghai, which had joined the national pilot, Beijing was running a local pilot without discussing any form of expansion.However, this difference does not compromise the study, as the way of piloting LTCI itself, for example, whether or not to join the national pilot scheme, is part of the policymaking process.

Sampling and recruitment of participants
Purposive sampling was used.Given the research question, the samples must be insider view-holders, included by two criteria: participation in LTC policymaking and information richness (Bradley, 1993), which is more likely to be achieved when individuals hold key positions (Goldstein, 2002).Based on a five-stage policy cycle and first-hand experience, the sample was then narrowed down to those involved in the drafting of the policy text.Once released for interdepartmental discussion, this often means that the main body of the policy text has been finalised by the responsible department (or jointly with highly relevant departments), with minor or technical revisions to follow.This critical stage involves three types of policymakers.Theoretically and empirically, the government is a central actor in Chinese policymaking (Wong, 2004), especially (deputy) directors as policy drafters.Think tanks and researchers are increasingly involved in Chinese policymaking (Think Tank Research Centre of SASS, 2013).Invited by relevant government departments, service providers usually participate in policymaking as industry experts.
Using snowballing techniques, including reputation snowballing and interview tracing, 25 people were recruited, including central and local officials from four government departments (n = 17), researchers (n = 5), and service providers (n = 3).Self-identified researchers at government-affiliated institutions were divided into officials and researchers according to their specific roles, for example, those holding departmental directorships were classified as officials.Those from government-affiliated bodies, such as the Beijing Municipal Commission on Ageing and the Shanghai Research Centre on Ageing, were grouped as local government department participants.

Data collection
Informed by the topic and theoretical perspective of the study (Charmaz, 1990;Green et al., 2007), interviews explored the five topic areas presented in a broad and open-ended format in the conversation to avoid overdirecting, including (1) Informational questions: state policy development and main initiatives after 2013/2016; (2) Informational and reflective questions: significant policy changes in Beijing and Shanghai; (3) Reflective questions: core issues and tension points of LTC policies and practices; (4) Reflective and feeling questions: the policymaking process and their role; and (5) Conclusion question: future policy changes.
After obtaining ethical approval through the University ethics committee, face-to-face interviews took place in Beijing and Shanghai between June and December 2018.Interviews were audio-recorded by the researcher's encrypted recorder.Before the interview, informed consent was secured from participants.Most interviews lasted around 60 to 120 minutes (n = 21).Where participants had limited time (n = 5), the interviews focused on topics (3) and ( 4).

The analytical process
Audio-recordings were transcribed, and NVivo-12 used to support data analysis.Thematic analysis and comparative analysis were carried out by referring to MSA concepts (Figure 1).Thematic analysis was used to bring together the accounts of different participants, reconstructing the LTC policymaking process in Beijing and Shanghai.With careful reading through the data, coding, and interpretation, 8 themes emerged from 64 codes, explaining the core processes of LTC policymaking in each site.Comparative analysis is central to this research design.A constant comparison of thoughts and perspectives across purposively selected settings and participants, as well as sectors, was conducted to identify similarities, differences, or even contradictions, "using qualitative datasets to full advantage" (Barbour, 2007(Barbour, , p. 1116)).In particular, MSA allows comparisons at the theme level, overcoming a common flaw in thematic analysisfailing to show the connections between themes and gain insight into the phenomenon (Vaismoradi et al., 2016).

Findings
This section reports findings from the comparative analysis using the MSA, which are summarised in Table 1.Since the joint LTC funding mechanism instructed by central directives depends largely on how LTCI as a new element is established, this section focuses on the debates surrounding LTCI rather than on existing elements, for example, targeted welfare and private insurance.The tension in constructing LTC issues Comparing the factors that influenced the framing of LTC issues in Beijing and Shanghai highlighted the tension between political deliberations and evidence.Overall, the debates revolved around LTC needs, and local policymakers in both municipalities used similar factors to frame LTC issues, including political pressure, public opinion, research evidence,1 and budgetary/capacity constraints.However, they differed in the importance given to each factor.Despite its leading role in the LTCI pilot, Beijing HRSS showed little concern for LTC itself.It was the constant pressure from its superiors, the Beijing People's Congress and China HRSS, that pushed it to launch and lead the pilot.Another key actor in LTC policymaking, Beijing MOCA, relied solely on public opinion evidence and conceptualised the LTC problem as non-existent or viable.Strikingly, regular monitoring of public opinion did not capture voices about unmet or unaffordable LTC needs in Beijing.Therefore, Beijing MOCA ruled out LTC for policy intervention and thus the existence of the LTC problem.Importantly, this situation was not specific to Beijing.The account below shows how incomplete/biased public opinion has shifted the ageing policy system away from LTC in China.
Most policies aim for ordinary older people dealing with daily life needs … It is healthy older people that convince the top leadership that aged care [policy intervention] concerns over 200 million older persons (CBS01: official, China MOCA).
Therefore, neither Beijing HRSS nor MOCA used evidence to inform the conceptualisation of LTC issues.Note that little use of evidence did not mean that evidence was unavailable.Instead, research projects and the Ability and Needs Assessment System yielded ample evidence of LTC needs; however, it was either overshadowed or ignored in LTC policymaking.Researchers reported that Beijing MOCA shelved data from a ¥ 11 million project, and Beijing HRSS even rejected a proposal for a small-scale survey of LTC needs.
Unlike Beijing, Shanghai policymakers did not even refer to political pressure from superior authorities as important in influencing their views on LTC issues.Instead, they perceived central directives as an opportunity to make or legitimatise policy change, for example, incorporating the unauthorised pilot into the national pilot.Also, public opinion and evidence were both used to support the policymaking process.Rather than treating public opinion as comprehensive and reliable, Shanghai official participants believed in the duty of speaking for those whose voices were weak in public opinion and intentionally sought further evidence on LTC needs.Thus, they reported different methods of collecting evidence for policymaking, for example, bureaucrats' research, large-scale surveys, and field investigations.Relatively complete information revealed rapidly growing unmet LTC needs in Shanghai, forming a consensus on LTC issues as a societal risk.Unsurprisingly, as cited below, this view was consistent with all researcher and professional participants who also used evidence on ageing to frame LTC issues.
LTC is a kind of need that everyone may have at some point … most of them will face the problem that their families can neither provide nor afford [LTC] (CBS20: researcher, Beijing).
In addition to the three factors mentioned above, all Beijing official participants stressed LTC budget constraints, especially Beijing MOCA sought to use them to define a viable LTC problem, focusing only on seniors with severe disabilities.The following account explains the prevalence of this "reverse logic" in Chinese policymaking, which prioritises fiscal capacity over needs.
… starts with measuring fiscal capacity, then identifies the target population, and then decides to what extent to support them … the so-called inflexible needs [for policy intervention] … are defined … in terms of fiscal capacity … (CBS15: researcher, Beijing).
In Shanghai, fiscal and service capacity constraints overtook evidence on one occasion, ruling out cognitive disability-related needs from prior LTC policymaking; however, emerging evidence compelled Shanghai to reconsider them in 2018.
… we considered the resources at our disposal … Why should I expose the problem without solutions at hand? (CBS26: official, Shanghai HC).
For both sites, concerns about budget/capacity constraints were understandable.In Beijing, even with a severe disability rate of 3% (Tang, 2016), there were still an estimated 128,700 older people in need of intensive LTC, with a total cost of 3.24 billion per year estimated at the current level of LTCI benefits.In particular, the financial capacity of Shijingshan District, the LTCI pilot district, was weak.In 2021, its public revenue was 6.467 billion, the lowest among the six urban districts, less than one-tenth of the wealthiest district, and a third of the fifth-ranked district.In Shanghai, the LTCI fund expenditures were estimated to be 3.77-3.87billion in 2020 (Zhang and Yao, 2022).
Comparing the process of conceptualising LTC issues in Beijing and Shanghai showed that by using more evidence, LTC issues were more likely to be constructed as a challenge for government intervention, as seen in Shanghai.While responding to political pressure, public opinion or budgetary constraints tended to ignore or minimise the magnitude of LTC needs, thereby overlooking or underplaying LTC issues.As seen in Beijing, none of the functional departments considered LTC severe.At most, it was regarded as a sub-problem of home care, as reflected in the guiding policy, the Beijing Municipal Regulation on Home Care Services (the Regulation), which placed LTC under the agenda of home care and integration.

The tension in assessing policy options
Comparing the three main criteria used by local policymakers to assess policy options for financing LTC revealed a tension between stability maintenance and policy effectiveness.Alongside policy effectiveness and multidimensional stability, policy responsiveness in this article refers to the "creation and implementation of timely, intentional, and effective policy actions" (Sogie-Thomas et al., 2018, p. 367).Overall, Beijing policymakers gave conflicting accounts of the proposed funding solutions, reflecting their struggle to balance policy responsiveness and stability.Rarely bothered by this, Shanghai valued policy effectiveness and embraced reforms to the existing policy system.
Beijing official participants believed that instructions from higher authorities must be thoroughly implemented in the political centre of China.They, therefore, were keen to express their support in principle, agreeing on the imperative of LTCI.However, resistance to operationalising this new institution was evident, reflecting a deep-rooted preference for stability and great caution about change and risk within the Beijing government system.
… Beijing favours 'early rise and late action', meaning observing the situation until it becomes clear before any tangible action (CBS17: official, Beijing HRSS).
Accordingly, various constraints to radical solutions dominated the discussion, being attentive to handy and low-risk options.In particular, whether LTCI was financially sustainable was most debated.The following narrative stresses the overall fiscal challenges facing the Beijing government, which effectively divert attention away from LTCI.
… funding is the most critical issue [for LTCI] … Beijing is also facing difficulties … e.g.relocating the Beijing government … projected to cost hundreds of billions … [Additionally] the Central Special Administrative District will include Dongcheng District and Xicheng District … more than 100 billion in tax revenue flowing to it (CBS16: official, Beijing HRSS).
Then, the primary source of funding suggested by the central directives, the health insurance fund, was deemed unsustainable.Ironically, to seek financial stability, alternative funding sources were mentioned but were promptly abandoned due to political stability concerns, for example, collecting LTC insurance premiums in the name of health insurance, merging LTC-related social insurance funds, and reallocating welfare spending, all of which required reforms to the existing public finance and social security institutions, which entailed political risks.As a result, little data showed Beijing striving for more funding for LTC, especially dropping the opportunity to secure alternative funding sources for LTCI.
… Shanghai reduced the social insurance premium rate by one per cent and used it to finance LTCI … We proposed this route, but nobody supports it!(CBS16: official, Beijing HRSS).
Given the constraints, no forces opposed the proposal for a small-scale LTCI pilot, referring to LTCI mainly funded by health insurance funds to cover urban residents with severe disabilities in a single district with the smallest aged population.Although limited to the lowest administrative level and coverage, it ensured a minimum level of policy responsiveness to central directives, financial stability, and minimal political risk.
Anticipating its effectiveness in tackling the severe LTC problem, Shanghai policymakers unanimously supported the universal approach to LTC provision, valuing LTCI as an imperative institutional reform, as shown below.
It [LTCI] transforms [LTC] from an individual and family risk to a societal risk.The government should respond by institutional building.In this sense, it is the last safety-net … (CBS28: official, Shanghai HRSS).Thus, unlike their Beijing counterparts, who sought to preserve stability by maintaining the status quo, they were far less concerned about stability.Instead, they meant to achieve it through action, as inaction on rapidly growing LTC needs would cause social unrest.As such, Shanghai's positive attitude to policy change was clear.
Once decided, we create the conditions for it … You'd better do nothing following all the old rules.
[Policy] pilot is to abolish the old and establish the new (CBS22: official, Shanghai DRC).Moreover, Shanghai policymakers showed enthusiasm for seeking effective solutions regardless of political risks.Despite China HRSS's explicit rejection, Shanghai initially explored LTCI through the "Post-80s Healthcare Plan" in 2013, which was legitimised till the national LTCI pilot launched in 2018.Also, Shanghai policymakers strived to overcome budgetary constraints by obtaining alternative funding sources.Disregarding state policy of lowering the social insurance premium rate by 1% to ease social burdens, Shanghai retained and reallocated this portion of funding to LTCI, thereby ensuring fiscal stability without increasing the tax burden.
Comparing the criteria valued by local policymakers showed that local policymakers paying more attention to responsiveness to higher authorities and to political and financial stability tended to prioritise retaining the status quo with minimal change over establishing new institutions, while local policymakers with a greater focus on policy effectiveness were attentive to the latter.

Political environments
The comparison of the political environment for LTC policymaking in the two sites highlighted the role of political will at the municipal level in shaping adverse or supportive political contexts.In this study, "political will" refers to the "extent of committed support among key decision-makers for a particular policy solution to a particular problem" (Post et al., 2010, p. 659).In Beijing, the municipal leadership's weak political will fostered an unfriendly environment consisting of insurmountable obstacles, whereas in Shanghai, a firm political will constructed a favourable political context, drawing and sustaining the attention of relevant departments and eliminating constraints on LTC policymaking.
A lack of political will to drive LTC policy was evident in Beijing, starting with no intention of participating in the national LTCI pilot.The account below illustrates that active participation in other national pilots raised doubts about the municipal government's motives for being passive to the national LTCI pilot.
Beijing was not designated [as a national pilot area] … why is Beijing taking the lead in everything but this?Beijing has participated in many national pilots.We must succeed if we want to (CBS17: official, Beijing HRSS).
Lacking political will was reflected in the low priority of, and little attention to, LTC, for example, excluding LTCI from the mayor's regular meeting agenda and the precarious political support due to cadre turnover.This was particularly true in comparison with other cities.
… both the Shanghai government and the Shanghai CPC committee take it [LTCI] seriously … We visited Chengdu where the mayor and the CPC secretary chaired the meeting [on LTCI].Never happened in Beijing!(CBS16: official, Beijing HRSS).
A lack of political will made various barriers to LTC policymaking unbeatable.As noted above, even in terms of stability and political risk, which Beijing participants were most concerned about, the proposal to retain 1% of social insurance premiums for LTCI was achievable because it was a precedent in Shanghai; however, the municipal leadership did not take any action, declining the opportunity to remove financial barriers.
Consistent with the above quote, Shanghai policymakers attested to the intensity and continuity of the municipality's will to tackle LTC issues despite high cadre turnover, leading to a willingness across the government system to reform the existing ageing policy system and take political risks.
The municipal leadership rarely attaches such importance to it [LTCI] … Every deputy mayor pays great attention to it … The then deputy mayor held at least two [interdepartmental] meetings every month (CBS27: official, Shanghai HC).
In particular, a strong will transformed the old power structure unconducive to LTC.The data showed a relatively balanced power relation in Shanghai; however, it emerged not by natural evolution but by political intervention, for example, appointing Shanghai DRC and Shanghai HRSS to lead LTC policymaking successively and involving Shanghai HC.These appointments made more departments stakeholders and placed two LTCI-related departments at the centre, giving new impetus to LTC policy development.Note that losing dominance did not mean the marginalisation of Shanghai MOCA.Instead, with the continued support of the municipal leadership and its knowledge and experience of ageing policy, Shanghai MOCA remained a key player in influencing LTC, for example, developing new LTC services and negotiating services covered by LTCI.

Key actors
Comparing the different roles played by all the actors in the two sites revealed that the absence or presence of key players critically influenced LTC policymaking, as they brought together the problems, policies, and politics described above to achieve a profound policy change.Central directives created opportunities for LTC, but no forces in Beijing were dedicated to and/or capable of driving policy change, whereas the Shanghai municipal leadership played this central role.
At the municipal level, given the weak political will, there was little evidence that the Beijing municipal leadership participated in LTC policymaking in any form, for example, convening special meetings, coordinating relevant departments, listening to briefings, or giving direct instructions.Without higherlevel involvement, LTC policymaking was carried out by independent parallel departments, for example, Beijing HRSS and Beijing MOCA, along with loosely linked researchers (see below), demonstrating the inability to overcome political, financial, and institutional barriers.
Unlike Beijing, the Shanghai municipal leadership showed a strong presence in LTC policymaking, especially its coordinator role.The early involvement of Shanghai DRC highlighted the municipality's awareness and determination to coordinate departments to remove obstacles, mainly institutional barriers, as cited below.
… the municipal leadership instructed us [Shanghai DRC] to fully integrate health and social care … and funding … With various institutions, it is easy to say but difficult to achieve integration, highlighting the role of [Shanghai] DRC … usually to balance and coordinate multiple departments (CBS22: official, Shanghai DRC).
A dispute resolution mechanism was also established to address cross-departmental issues.Thus, the monthly meeting convened by the deputy mayor and attended by all relevant departments became "the main battleground to discuss various LTC policy ideas" (CBS27: official, Shanghai HC).Through robust steering, Shanghai reached high levels of agreement on LTC issues and solutions.
At the departmental level, Beijing MOCA dominated the ageing policy system, including LTC; however, it was the defender of the status quo.Its dominance stemmed from active involvement based on long experience regulating welfare and aged care.Moreover, Beijing MOCA, as the implementer of the Regulation, received support from Beijing PC and the municipal leadership, which significantly strengthened its position, as cited below.Inevitably, Beijing MOCA effectively pushed its targeted welfare and home care agenda while suppressing LTC-centric solutions.
Conversely, lacking experience in ageing and political support, Beijing HRSS, the implementer of the LTCI pilot, was weak and showed little intention and capacity to lead LTC policymaking.It held passive attitudes towards LTCI, with its chief director in particular giving specific instructions to keep a low profile.Rather than leading relevant departments, Beijing HRSS tended to exclude or distance itself from Beijing MOCA and Beijing HC, which were only informed of the policymaking outcomes, according to complaints uncovered in interviews.The following quote captures this exclusive or antagonistic relationship.As a result, all efforts to draw the attention of key post-holders to LTC failed, as mentioned above.In contrast to Beijing, where MOCA was dominant with a passive stance on LTC, Shanghai MOCA became an active contributor to LTC policy.As noted above, the strong presence of the municipal leadership changed the interdepartmental power relations by involving more departments and assigning a leading role to Shanghai DRC and HRSS, thus achieving relative balance in Shanghai.This clearly undermined the dominance of Shanghai MOCA.While in Beijing, none of these changes occurred, and the old power relations, dominated by Beijing MOCA, remained.

Beijing HRSS never acknowledges what
Both sites invited qualified researchers and industry professionals to participate in policymaking, for example, a good reputation in their field, familiarity with government officials, similar mindsets, and cooperation.Overall, mirroring the state-society relationship in China, they were disadvantaged, unwilling or unable to advance their proposals.This is particularly true of the Beijing researcher participants, who differed from most government officials by supporting new institutional arrangements to address serious LTC issues.More importantly, they lacked the willingness to stand up for their policy advocacy.As such, they tended to withdraw from policymaking when encountering disagreement, as cited below.
I suggested a survey of LTC needs, even if limited to a handful of streets.How can you conduct an LTCI pilot without data?… since they refused, I had no intention of staying (CBS19: researcher, Beijing).

Interactions of problems, policies, politics, and actors
The interaction of the above four factors determined the different outcomes of LTCI policymaking in Beijing and Shanghai.Beijing's LTCI policymaking was characterised by ignoring or denying LTC issues for policy intervention, an apparent preference for maintaining the existing ageing policy system, and the municipal leadership staying away without any forces committed to promoting LTC.As a result, various obstacles led to superficial policy change, a district-level LTCI.In contrast, Shanghai policymakers reached a consensus on a pressing LTC problem for governmental intervention, prompting them to seek effective solutions.To remove political, institutional, and financial barriers, the municipal leadership acted as the key actor, directing and coordinating concerted action across departments.Thus, Shanghai achieved drastic policy change, establishing a municipallevel LTCI.

Discussion
This article examined different experiences of LTC policymaking in China by applying MSA to a comparative study of Beijing and Shanghai.Findings on highly politicised LTC policymaking have four implications for policy analysis and development.
First, this study revealed a highly politicised LTC policymaking process in Beijing and Shanghai, which was full of ambiguity and political manipulation (Kingdon and Stano, 2011).These findings strongly challenge the assumption of a linear, neutral, and straightforward LTC policymaking process, bridging a critical knowledge gap in the current literature on LTC and Chinese policymaking.Unequivocally, findings on key influencing factors revealed that politics, for example, political actors, power relations, and political deliberations, run through the entire process of policymaking (Blackman et al., 2012), shaping the LTC problem and the discussion of policy options and determining whichever perspective won out on the agenda.
Complex central-local dynamics also manifested the politics in LTC policymaking.Consistent with the literature, the two cases showed that local governments had considerable discretionary power in deciding to what extent central directives could be enforced locally, for example, superficially and thoroughly in Beijing and Shanghai, respectively.Moreover, such significant variation suggests that merely acknowledging local autonomy or any general statements of self-interest in the current literature is insufficient to explain diverse LTC policymaking experiences and outcomes across China.For instance, a recent study argues that central coordination, regional competition, and local protectionism determine local LTCI models (Chan and Shi, 2022); however, the two cases in this study hardly fit this explanation, especially given their contrasting attitudes towards LTCI.Therefore, an in-depth analysis of policymakers' specific concerns, as done in this study, is crucial to generate accurate knowledge of policymaking.
As a result of policymaking, an incomplete translation of LTC needs into policies was identified in both sites, best exemplifying political engagement.Cross-case comparisons demonstrated that in Beijing, greater attention to political factors, compared to engagement with evidence and policy effectiveness, resulted in the least coverage of needs in LTC policy.Shanghai was more inclusive than Beijing in this regard; however, incompleteness still existed, for example, underplayed LTC needs related to cognitive impairment due to a lack of knowledge and capacity concerns.This finding suggests that the extent to which LTC needs were translated into policies, that is, the definition of the legitimate needs for policy intervention, is political (Langan, 1998), and knowledge use reduces political manipulation in policymaking (Kingdon and Stano, 2011).
Second, the four influencing factors revealed by this study encompass actors/interests, institutions, and ideas, providing a multi-factor explanation for LTC policymaking in China.The findings align well with the core argument of MSA: "policy outputs are neither exclusively rational nor solely a function of institutional design; rather, they depend heavily on a complex interaction between problems, solutions, and politics" (Zahariadis, 2008, p. 514).As such, this study challenges prior studies on Chinese policymaking, which, as reviewed previously, emphasises a single determinant, enriching the literature by examining social constructions and their impacts on policymaking outcomes.Empirically, this perspective offers an opportunity to change the policymaking course for more inclusive social policies.
A striking feature of LTC policymaking examined in this article was the diverse perspectives on LTC issues, policy options, and assessment criteria between policymakers and places.The LTC problem was socially and politically constructed by local policymakers differently, which challenged the problem orientation in Chinese policymaking that considers problems objective to be discovered.Furthermore, varying LTC policymaking processes in Beijing and Shanghai proved that problem recognition and construction critically affected policymaking outcomes (Barzelay and Gallego, 2006;Kingdon and Stano, 2011), especially the Shanghai case revealed a path-shaping process (Mahoney, 2000), starting with framing the need for institutional reform (Cox, 2001), namely, a severe LTC problem, followed by challenging the effectiveness of the existing ageing policy system and turning to a universal LTCI.Likewise, local policymakers varied widely in understanding key criteria, for example, stability and fiscal capacity, especially their change-or-not actions on fiscal capacity, confirming that some budgetary constraints are perceptual (Kingdon and Stano, 2011).
Third, despite the trend, pluralistic participation in Chinese policymaking was contingent on the government's acceptance.Defining participants in policymaking is a fundamental political issue (Baumgartner, 1989), and what role experts play in policymaking is "a consequence of the local political environment" (Lundin and Öberg, 2014, p. 25).This is particularly true in China, where a key political environment is a power imbalance between the state and society, leading to a governmentand CPC-centric policymaking model (Wang and Hu, 2010;Ren, 2018).In this study, only qualified researchers and service providers were invited to participate in LTC policymaking, and their impact on the outcomes was significantly weak, verifying the state-centric policymaking in China.
Fourth, the policy window is a core component of MSA; however, it did not emerge as playing a critical role in this study.One possible explanation is that a long duration of the policy window substantially reduces its significance to LTC policymaking.In this study, when the joint funding mechanism was formed by central directives in 2016, the policy window emerged in the political stream at both sites.Notably, it did not close as rapidly as Kingdon and Stano (2011) describes due to two contextual factors: the evolving LTC system and institutional ambiguity in LTC policymaking.As an emerging policy area, the LTC system is beginning to take shape in China, especially LTCI, which is still in the policy experimentation stage.Therefore, until finalising the national policy, the policy window remains open to local governments.For example, in 2020, Shijingshan District was included in the national LTCI pilot.Moreover, due to LTC policymaking encompassing sectoral and jurisdictional boundaries, institutional ambiguity expands the policy window through endogenous spillover (Ackrill and Kay, 2011).Namely, policy innovations in one department or one place may trigger policy changes in another department or other places.

Limitations and future research
First, this study collected retrospective accounts on LTC policymaking and focused on a specific time period from 2013 to 2018, which may omit some variables occurring over a long period, for example, the shifting national mood and public sentiment.It is therefore worth considering longitudinal studies of LTC during its formation.Second, imbalanced power between different policymakers is a key variable in LTC policymaking.Given the trend towards pluralistic participation in Chinese politics, future studies could explore changing power relations and how these changes shape LTC over time.Third, this study examined LTC policy in Beijing and Shanghai.Future studies could test the findings for applicability to other policy domains and locations.

Conclusion
Using the MSA as an analytical framework, this comparative study reveals a highly politicised process of LTC policymaking, from framing LTC issues and discussing policy options to the complex political environment and key players, identifying four influencing factors in a relational way that explain how and why local governments adopt different policies to finance LTC.
The article has theoretical and empirical contributions.Through the lens of local policymakers to detail the LTC policymaking process, this article complements and advances Chan and Shi's (2022) work, which explains the emergence of different local LTCI models from an institutional perspective, thereby enriching the literature on Chinese LTC policy.By recognising and understanding the politics of LTC, Chinese policymakers can utilise the knowledge generated in this study, for example, specific political processes and factors affecting them, to optimise policymaking for more inclusive LTC provision.At the very least, it helps policymakers minimise apparent exclusions in policymaking, for example, disabled seniors with weak voices.
To adapt to the Chinese context, this article makes multiple conceptual modifications to the MSA and has demonstrated the strengths of the comparative use of MSA to understand different policymaking experiences across China.Given the focus of this article, the analytical framework will be addressed in a future article.
Beijing MOCA promotes regarding LTCI, including what I'm working on [Unified Needs Assessment Standard] (CBS12: official, Beijing MOCA).

Table 1 .
Problems, policies, and politics and their interactions.