Ideas and Policy Response to the COVID-19 Crisis: Evidence from Jakarta, Indonesia

Coronaviruses have emerged as a potential disruptive force in policymaking. Using a comparative case study method, we examine two social policy responses in Jakarta, Indonesia: the Social Safety Nets (SSN) programme and the health policy. Such examples demonstrate an aggressive change in policy direction from means-tested systems and government-centred approaches to a total relaxation of conditions with the involvement of non-state actors in the provision of services. Our study analyses the ideational dimensions of the policy process that produces abrupt and radical change. From our analysis, the policy change may be explained by the emergence of a new policy paradigm created through the emulation-contextual process – an alternative model of policy learning. The theoretical implication of our research is that policy response in this study cannot be viewed in a completely path-dependent process. Instead, we propose a ‘ path-creation accelerator, ’ which represents an infrequent instance of policy change.


Introduction
COVID-19 is, first and foremost, a global humanitarian challenge. The World Health Organisation (WHO) declared that COVID-19 cases reached over five million in five months after the first outbreak (WHO, 2020). A widespread understanding of new norms to prevent virus transmission has stimulated a series of radical changes, which, for the vast majority of the population, boosted insecurity (Murphy, 2021;Stewart and Dodworth, 2021;Hornung and Bandelow, 2022). To mitigate the insecurity, many national and sub-national governments around the world implemented a range of generic welfare measures aimed at minimising health and socialeconomic damages (Irving, 2021). However, their responses often conflict with their traditional welfare system attachments (Ramia and Perrone, 2021;Hogan et al., 2022).
Traditional approaches to analysing policy change are divided into two camps: path dependence accounts and actor-oriented explanations (Carstensen, 2011a(Carstensen, , 2011bCarstensen and Schmidt, 2016). Rooted in historical institutionalism, the path dependence premise argues that change is influenced by past policy legacies. Such a focus often overlooks the potential role of policy actors as social agents with some control over their decisions. Another view suggests that the reversal of extensive policy reform is due to the presence of policy entrepreneurship (Lessenich, 2005). Policy entrepreneurship represents knowledgeable agents who take advantage of opportunities to act in ways other than those prescribed by the bounds of their traditional institutions (Bakir, 2009;Arnold, 2021). This view consequently downplays the importance of history and memory in influencing the policy making process.
The ideational approach then emerges, offering an alternative perspective that bridges these two opposite theoretical camps (Béland and Lecours, 2005;Béland, 2016;Béland and Powell, 2016;etc.). This approach allows us to understand that change would always be driven by reference sets inspired by layers of significant events that have inhabited the many past moments and actors' interpretations of social realities at the same time.
Jakarta, Indonesia's capital city and the city with the world's highest number of COVID-19 infection cases per million people, merits further consideration in the international context since it has been recognised for its effective COVID-19 response strategy. The two introduced social policiesnamely, the Social Safety Net (SSN) and health policiesare worth studying because they are pertinent to changes in institutional models in developing nations, particularly those experiencing emergency situations. These policies illustrate the various forms of change in the features of the original regulation, such as a system with proven capabilities and governmentcentred services, to regulations that are completely relaxed (discussed in the case of health), while continuing to involve the role of non-state actors in providing services (checked in the field of SSN). Although the change is meant to be temporary, this phenomenon signalled the magnitude of the shock caused by the COVID-19 crisis in stimulating change that crossed the bounds of traditional institutions.
Generally, the Indonesian welfare system is described in social policy literature as analogous to the productivist model (Wood and Gough, 2004). However, recent literature has challenged that claim, providing insight into Indonesia's classification of welfare systems in a comparative context. In the first strand of literature, it seems that Indonesia has been moving away from the productivist-dominated system, as have its counterparts in East Asia (Aspinall, 2014;Sumarto, 2017;Murphy, 2019); the arguments produced often draw on evidence from the expansion of social protection (cash transfers, in-kind benefits, and changes in health insurance to make it more inclusive). In the second strand of literature, however, the hypothesis has been rejected despite significant reforms in social policies (Rosser et al., 2016;London, 2018;Nurhayati, 2020;. Apart from the contentious debate over Indonesia's welfare system, welfare provision in Indonesia remains dependent on regional governments' political commitment and fiscal resources, producing a dichotomy among regions (Fossati, 2017). This provision reflects issues arising from the decentralised governance model adopted in 1998 and the institutional capacity of the government levels (Asmorowati et al., 2022). Jakarta's welfare system shares some characteristics with the city-states of East Asia which are characterised by productivist regimes (e.g. Hong Kong, Singapore). Examples of these characteristics are selective welfare, the rejection of interest as a social right, and a focus on social investment (see Yuda, 2020 for overview), where primary education assistance has been social policy focus.
Moreover, the case selected also represents initiatives related to drastic change, an articulation to challenge the 'stability' premise that is identically attached to institutional studies about welfare in the context of the advanced welfare state (Hills, 2004). In short, institutionalists would assume drastic changes are unlikely, arguing that policy is 'the outcome of a long process of development and represents distinct paths' (Hudson et al., 2008: 211). However, the scale of COVID-19 crisis, as our examination of the Jakarta case suggested, has provided a window of opportunity for major change and further institutionalisation.
The objective of this article is primarily to account for engagement of ideas in policy response during states of emergency and in renegotiating the roles between non-actors in the delivery of services and the provision of direct services. We found that ideas shaping examined cases in Jakarta during the crisis tend to be based on a conceptualisation of the causes of problems and solidarity value. Underlying assumptions of policy adoption that refer to a particular ideology as found in the West appear not to be dominant and tend to be discarded. This phenomenon occurs when ideas find fertile context, which, in turn, creates a favourable condition for policy adoption. We find that the policy adoption route that Jakarta has taken in dealing with the COVID-19 crisis may be viewed as a form of contextual emulation, described by Weyland (2004) as the process of modifying ideas from other settings to be applicable to a domestic context. The modification is made possible by the rising trend of using evidence-informed policy and practice in the policy process. And as often appeared in the case of health, commitment to evidencebased policy making has provided an incentive for the government to adopt policies in line with social contexts and real-condition considerations.
The examined cases also show us how multiple considerations and interests that originate from the government, parliaments and international organisations are negotiated and mediated by calling attention to the role of policy entrepreneurs (the Governor's Delivery Unit, TGUPP) in translating ideas into practice. The policy entrepreneur, in turn, produces path-switching strategies for a policy response. As an overview, TGUPP is a non-structural institution that serves the governor of Jakarta in policy making and implementation and also works with state agencies to align resources to the governor's strategic and visionary goals while also assuring delivery strategy and impact on society. TGUPP was involved in various crisis responses that reflected Jakarta's prompt policy responses and actions to prevent COVID-19 infections by intensifying testing, tracing, and treatment, along with recent efforts to give 'safe' vaccines to all residents.
For a theoretical implication, this article offers possible avenues for recalibrating the theoretical underpinning of institutional change during crisis.
Theorising policy responses Policy stability and change theories were created to explain very different developments from those observed during the crisis (Béland et al., 2021). Historical institutionalism, which is rooted in the path dependency premise, is arguably the most relevant theory for explaining the crisis. It is an extensive model used to explain the resilience and persistence of public policies in times of crisis and intentional reform (Steinmo and Thelen, 1992;Blyth et al., 2011;Lewis and Steinmo, 2012). East Asia, Scandinavian, Europe, the United States, and Canada might support this argument since there have been few major policy changes because of highly unstable economic conditions (cf. Aidukaite et al., 2021;Béland et al., 2021;Cantillon et al., 2021;Greve et al., 2021;Hick and Murphy, 2021;Moreira and Hick, 2021;Soon et al., 2021). The economies in these areas are becoming increasingly affected by globalisation, and it is not feasible for them to commit too much to policy reforms that may produce unknown economic risks and threaten the stability of markets.
Nonetheless, existing policy legacies may lose their effectiveness over time, especially if the crisis continues and uncertainty persists (Béland et al., 2021). Accordingly, the punctuated equilibrium theory provides a framework for understanding policy dynamics in crisis situations (Jones and Baumgartner, 2005). This theory views policy as moving along a trajectory or series of events but punctuated by short periods of shock called critical junctures (Capoccia and Kelemen, 2007). Recently, Collier and Munck (2022) provided a comprehensive assessment of the critical juncture concept, defining it as a 'change that produces an effect that persists over a long time, through mechanisms of reproduction, after the initial cause [exogeneous shock] has ceased to operate ' (2022: 4). This means that legacy matters when assessing whether a policy change can be classified as punctuated equilibrium or not. It is important to note that while the resulting change or innovation may have radical effects, it does not necessarily remove the existing institutional footprint.
Despite this, some scholars contend that exogenous factors became significant for policy change only when endogenous developments had begun to diverge from the old path in a cumulative manner (Pierson, 1994;Capoccia and Kelemen, 2007;Mahoney and Thelen, 2010; see also Varjonen, 2020). In other words, institutional change might emerge because of previous institutional routines for making minor or ad hoc changes continually, often called endogenous or gradual changes. The mechanisms of endogenous or gradual change have been studied further by providing examples of drift, conversion, layering, and displacement. In a drift mechanism, changes or recalibration are introduced while preserving institutional characteristics and adapting their features only to the political, economic, and social conditions surrounding the institution. In the conversion mechanism, no institutional features are altered. However, actors will convert the goals, functions, or purposes of existing institutions to respond to new social risks. The next is layering, which refers to adding a new element to an existing one without removing the old one. Layering is aimed at strengthening institutional capacity to respond to post-shock conditions. Last but not least, the displacement mode refers to the removal of existing rules and the introduction of new ones that result from institutional deterioration (Thelen, 2004;Streeck and Thelen, 2005;Mahoney and Thelen, 2010).
Two cases we analysed in the context of Jakarta early in the crisis demonstrate a sudden and massive economic shock which called for governments to enact aggressive emergency programmes to stabilise economies and prevent a severe health crisis. In contrast to traditional institutionalist approaches (Pierson, 1994;Capoccia and Kelemen, 2007;Mahoney and Thelen, 2010), the response generated in Jakarta was rapid 'without' previous policy changes. However, such a reform opportunity is not guaranteed to result in a permanent paradigm shift or a significant, path-departing change in policy institutions. It is, therefore, advisable to assign alternative theories to be relevant to this context. This would suggest that what happened in Jakarta represented a path-creation accelerator, an infrequent instance of policy change.
The term 'path-creation accelerator' derives from the original meaning of the term 'path creation.' A review of various literature Karnøe, 2001, 2005;Garud et al., 2010) has shown that the path creation process is characterised by a deviation of the policy from the established path and principle, while discontinuing the previous one. Recently, Hogan et al. (2022) defined it as 'novel or innovative actions creating new paths' (p. 43). The path creation process is characterised by an extensive modification and even a replacement of the main ideas and features of policy with a new one in a very short period of time, as opposed to displacement, path-breaking, and path-switching, which requires a slow-moving, long-term process. It would be appropriate to use the term 'path creation accelerator' in this context given the uncertainty regarding the future institutionalisation of the current configuration of SSN and health care.
In this article, we define path-creation accelerators as a variant of the policy process in which the policy diffusion mechanism allows a new "potential" policy trajectory to be accelerated once the established system rooted within policies is destroyed by innovative actions that create new paths.
Path creation-related concept is centred on the premise that policy entrepreneurs shape paths, rather than exogenous shocks, by initiating processes that actively shape emerging social practices (cf. Hogan et al., 2022). In our context, new 'potential' paths are not simply emerging as a result of shocks associated with COVID-19, but rather as a result of strategic agencies creatively modifying policy content based on embedded cultural values, previous practices, and policy learning (cf. Schienstock, 2011). Therefore, a pandemic can not necessarily be considered an exogenous attack commonly responded to by routines; it is an opportunity to adapt to the crisis with creative and innovative solutions.
Furthermore, a path-creation accelerator may lead to gradual change; or, in some cases, to completely new innovations, accompanied by new policies within a short period of time, challenging the core established institutions. To be sure, policy learning and the innovative actions taken by policy entrepreneurs that result in policy innovations are characteristic of path-creation-centred theories, which are different to established policy process theories in their emphasis on routines as the basis for formulating policy. In sum, we can argue that the path-creation accelerator occupies the intellectual territory between punctuated equilibrium theory and path creation, which determines whether the policy will remain largely within the historical continuum or not .
The focus of this article is to draw attention to the empirical discussion of crisis management as it relates to emergency measures. The emphasis on ideational analysis would be the main contribution of this study, bridging the relationship between actor and institution during emergency situations.

Ideas within policy responses
Ideational analysis in policy studies examines the role and embodiment of ideas in the interactive processes of policymaking (Béland, 2016), which includes agenda setting, framing and policy learning (Starke, 2006). The trend of ideational research has developed in conjunction with the decreasing ability of the historical institutionalist approach to explain policy change (Starke, 2006).
Many argue that historical institutionalism has placed minimal emphasis on the role of the actor (Carstensen, 2011a(Carstensen, , 2011bCarstensen and Schmidt, 2016). The influential literature on this subject, Pierson (1994), viewed actors' choices on policy change as being isolated by the institutional framework. Such frameworks are continually reproduced within the complexity of political engagement involving broad institutional anchors such as powerful constituencies, norms and ideologies (Moynihan and Soss, 2014). The research referred to this phenomenon as a 'self-reinforcing process'which, in turn, offered a 'locked-in effect' once the policy had been institutionalised for a certain period (Pierson, 1994). This locked-in effect enables 'policy to stay largely within the bounds of a historical continuum' (Ramia and Perrone, 2021: 3). Anchoring on this logic, it was argued, the crisis would not lead to policy changes that crossed the bounds of the previous choice. For this reason, moderation of the change will be negotiated only in gradual mode, which is commonly defined as piecemeal reform (Thelen, 2004;Streeck and Thelen, 2005;Mahoney and Thelen, 2010).
By highlighting diverse political dynamics in terms of policy innovation, Béland and Lecours (2005) offered insights for understanding the greater role of ideas in social policy development. Ideas can be defined as beliefs that are formed by individual and collective actors' perceptions of social reality (Béland, 2016). Moreover, 'ideas are also used as a strategic resource to influence other actors' normative and cognitive beliefs and to work around institutional constraints to bring about policy change' (Varjonen, 2020: 3). The scholarly work of Béland (Béland and Lecours, 2005;Béland, 2016;Béland and Powell, 2016;etc.) has inspired much research that follows similar lines (Schmidt, 2008(Schmidt, , 2010(Schmidt, , 2011Velázquez Leyer, 2020). However, there is a lack of empirical analyses of policy change resulting from ideas in the context of developing economies and the COVID-19 crisis. This research, therefore, contributes to filling this gap.
Taking seriously ideas like policy analysis has allowed us to explain the success or failure of how policy change goes beyond the deterministic perspective of path dependency, but includes path creation accelerators. Lessenich (2005) further elaborates that, to be able to explain how path creation stimulates change, a greater emphasis on the role of policy entrepreneurs is needed. Policy entrepreneurs are defined as 'knowledgeable agents with a capacity to reflect and act in ways other than those prescribed by existing social rules' Karnøe, 2001, cited in Lessenich, 2005: 349). Gunn (in Safuta, 2021: 1100) suggests a policy entrepreneur can be based 'within the state (politicians or civil servants), outside of it (in businesses, research organisations, interest groups or NGOs), or within international organisations.' This article considers that policy entrepreneurs here are the governor of Jakarta, Anies Baswedan and TGUPP. They negotiate and compromise in order to attract the support of stakeholders in the decision-making arena (Cohen, 2012;Frisch-Aviram et al., 2019).
Emergency situations such as the current COVID-19 outbreak can arguably be considered a 'path-clearing policy accelerator' (Hogan et al., 2022), accelerating pending policy proposals and spurring policy innovation. In Indonesia, for example, several local initiatives have sprung up all over the nation to respond to the crisis on many different scales (Asmorowati et al., 2022). Their innovation, especially associated with the SSN, has come to the fore on breaking the path that previously maintained patterns. And in this respect, Jakarta, through its policy entrepreneur community (TGUPP), produced the most innovative policy and was recognised as the best performing city in tackling the crisis compared with other districts or regions (Asmorowati et al., 2022). In the global context, many works also confirmed how the COVID-19 crisis has produced a critical juncture that has debilitated policy legacies (Béland et al., 2021). The crisis has also undoubtedly affected the present arrangements of state and non-state relationships with regard to meeting welfare responsibilities (Mok et al., 2021).
This case study of SSN describes how the COVID-19 crisis has disrupted the institutional framework of social policy embedded in the centralistic paradigm . Under the centralistic model, most social transfer schemes are organised at the governmental level and administered by social welfare agencies, while non-state-based schemes are placed as an alternative resort of support. Meanwhile, the case of health relates to the total relaxation of selective institutional arrangements embedded in previous policy practices.
In the broader discussion, ideas can be generated from processes of adoption (Kuhlmann et al., 2020). It is a process in which policy is channelled or spread through two major mechanisms, coercive and voluntary (Marsh and Sharman, 2008). The first occurs when policy stakeholders are forced to adopt certain initiatives by dominant actors. Donor countries or international or supranational organisations are considered a frequent cause of coercive policy adoption (Weyland, 2004). They compel the domestic government to change its policy orientation so that it is tailored to their political-economic interests. Loans, aids, or grants with conditionality and bilateral and multilateral agreements can be considered one effective instrument of coercion that is used by the aforementioned actors. Meanwhile, voluntary mechanisms occur when the flow of policy ideas is adopted voluntarily from other settings, including the best practice policy that has been made in another country (Meseguer, 2005). Marsh and Sharman (2008) divided voluntary mechanisms into two more types: emulation and political competition. The former is described as a situation where policy adoption is based on the logic of appropriateness or an ongoing trend. It can be understood as 'processes in which actors copy from others' (Kuhlmann et al., 2020: 82), reflecting the proportions of fatalistic decisions in the policy adoption process. The latter is the '[mechanism] shortcuts taken by domestic policy makers under pressure from electoral competition' (Velázquez Leyer, 2020: 135).
In the Jakartan context, we found that idea adoption in mitigating COVID-19 effects took the route of the contextual-emulation process. As appeared in the cases of health, we also argue that the fatalistic decision in policy adoption was successfully reduced and was becoming more contextual. This is mainly because Jakarta in its decision relied on evidence-informed policy and practice, although at certain points, boundaries between science and politics in the policy making process were unclear. To be sure, what is presented in the Jakarta case has been the process of convergence of ideas and interest between institutional entrepreneurs, local authority and international organisations, producing a response. The particular case also shows us how the concept of ideas and their components matters, particularly in (1) defining the social problem and policy design, (2) justifying the selected problem and public policies, and (3) prioritising particular problems and policy over other options.

Comparative case study
Using a comparative case study approach, our study aims to examine the production process of ideas within the social policy response to explain what made a path-shifting change in the SSN and health care possible. This approach involves two or more cases that are analytically comparable in topic and allow the systematic management of complexity within cases to provide mechanismbased explanations (George and Bennett, 2005;Obinger et al., 2013). It allows us to identify the key entities and activities of the mechanisms that are at work in a policy change and compare them with other policy processes (Kuhlmann and Nullmeier, 2021).

Case selection
Jakarta was selected as the case study because it was an epicentre of the crisis in 2020, and it had implemented aggressive social policies to mitigate the COVID-19 risks, making it feasible to study social policy changes during times of plague.
In April 2020, Jakarta implemented the Large-Scale Social Distancing policy (PSBB) ostensibly to stop the spread of COVID-19, resulting in substantial negative economic impacts. In this study, the importance of Jakarta is also evident from the fact that it contributed 37 per cent to Indonesia's GDP in 2019 and 10 per cent to its total employment. With this background, a wide range of welfare benefits during the COVID-19 crisis is aimed at stabilising national economies (DOC_1).
Moreover, the wide range of relief initiatives during the crisis made the selection of policy areas necessary to narrow down the scope of the study in order to identify cases that are both meaningful from a theoretical perspective and practicably feasible to study. We compare SSN and health care as they are social policies that experienced the path creation accelerator mode of change, diverging from their traditional welfare system, which was regarded as productivist; therefore, public funds have always been allocated primarily to supporting necessary social investments such as education.
The generosity provided by SSN and health care may have been temporary, but it should be understood in terms of both institutional changes and institutional traditions. This attempt might provide a path-clearing policy accelerator for policy institutionalisation.

Data collection and analysis
Our study uses multiple strategies to conduct data collection. In the first step, a total of five indepth interviews and informal discussions were conducted with influential actors who contributed to idea production on policy responses (see Table 1). The interviews assisted in reconstructing the sequence of events that led to the new direction of the policy process and in establishing the decisionmaker's opinions and thoughts regarding that path creation process.
In addition to the interviews, a secondary database was assembled, consisting of approximately nineteen official policy reports (see Table 2). The documents under consideration include government reports, programmes, memos, and preparatory documents that explain the reform programme regarding the policy design of SSN and health policy.
Moreover, we attempted to increase the accuracy of the data collected by supplementing the on-site visitation of relevant official meetings, with data collection taking place between September 2020 and May 2021. The purpose of the on-site visitation was to gain a better understanding of the  conditions under which policy processes and discussions take place in real-time. We can extract more meaningful data from this method since we can directly observe the process of decision making, as presented in our findings.
To hypothesise the validity of the inquiry, triangulation of sources and informants was undertaken whenever sensitive or conflicting information was provided by the informants. Several within-case and cross-case analyses were performed to elucidate the operative mechanisms of each case and achieve analytic generalisability (George and Bennett, 2005).

Case study one: Social Safety Nets programmes
Much remains uncertain about the economic damages of the virus. The adverse effects of COVID-19 on the economy have been more rampant in Jakarta. Motivated by the need to prevent capital outflows and the worsening of the economic crisis, the government immediately moved to strengthen its safety nets, especially during the period before the first lockdown measures were implemented. Some policy packages, e.g. unconditional cash transfer (BST) and food subsidies, were applied to poor families, the unemployed, and the self-employed as a strategy for improving their welfare. In addition, educational assistance (KJP Plus for elementary school students up to high school levels and KJMU for higher education students) was provided to the aforementioned groups. In total, 2.4 million households were targeted for the social safety nets, 1.1 million of which were from the Jakarta Social Welfare Agency and 1.3 million from the Ministry of Social Affairs (Doc_13).
The extensive review of official reports (Doc_3, Doc_6, Doc_7, Doc_12, Doc_13) and also of interviews confirmed that ideas adopted to craft social safety nets were drawn from crises experienced in the past instead of taking novel initiatives that might have created unknown policy risks. The unchanged feature also included the selective nature of coverage schemes that target vulnerable groups, with consideration given to economic status as a reference. A significant change occurred only in the implementation domain, where this process was performed by mobilising non-state actors, setting the current arrangement apart from its state-centred delivery model. The uniqueness of Jakarta's decision warrants an in-depth examination, especially of the ideational aspect that underlay such an abrupt decision.
At the onset of the virus outbreak, the TGUPP together with relevant parties developed a systematic understanding of the nature of the current crisis through several extensive meetings. In March 2021, the British government partnered with the Indonesian government in technically redefining the appropriate design. The partnership was then accepted, and in the subsequent meetings, the Jakarta government suggested that the British government's proposal, which stressed a more European welfare state model, could not be fully adopted (Interview_1). The governor of Jakarta, Anies Baswedan, added his preference to designing the contextualised implementation strategy, stressing the importance of combining state-led initiatives and citizen participation (Doc_1). Suggestions offered were finally considered for implementation.
It is very important to develop our policy scheme together with relevant national and international parties that mainly covers a benchmark process from across cities and countries, science and data driven, and evidence based (Interview 1) The economic shutdown and decreasing regional income fostered the idea of efficiency (Interview 3). While the number of people at economic risk increased, existing social safety nets were not sufficient to cover the increasing proportion of the population at risk. The radical change in the development agenda was also difficult to pursue. Faced with a trade-off between efficiency and redistribution, involving non-state actors came to be viewed as a better solution than centralising services or benefits.
The idea is the government can facilitate and collaborate with those who can donate and who are in need on one trusted platform that is reliable and updated in real time. (Interview 4) The importance of collaborative ideas and practice further lies in the burgeoning intention of the Jakarta governor to promote his ideas on urban solidarity among different political parties (Interview 4). In his idea, the initiation of a collaboration platform was formed by linking wide-ranging interactions and contributions among societies. The proposal regarding the ideas was studied by TGUPP and included broad policy advisory, and then forwarded to a wide array of stakeholders including government officials, private companies, and non-governmental organisations. Amid political resistance, Anies and TGUPP also succeeded in convincing the senator by framing the Large Scale Social Collaboration (KSBB) as a form of universal social solidarity (Interview_2). In addition, the KSBB programme promoted the idea that integrated schemes of formal and voluntary activity are an integral part of the Gotong Royong (working together toward a common goal) value that can be translated as 'collaboration' and was institutionalised as the Jakarta Development Collaboration Network (JDCN) (Doc_5; Doc_6). This value can be simply understood as undertaking collective action to address the collective problem. Here, we could imply that moral values are central to forming the ideas at work behind the policy.
In meetings we observed, the parties and citizen representatives involved exhibited a great collaborative spirit of resilience. After gaining attention from the public, the KSBB programme was finally introduced. The programme was unique, as the government acted only as a facilitator by setting up the real-time online platform that linked the contributors or donors and people in need. The platform called the KSSB web page allows potential collaborators to find information about the KSBB programme, map locations, and take part in donating to certain locations. The programme is currently in progress with more than 200 collaborators in the programme for food, SMEs, home learning, and kampung (village) improvement. The aforementioned collaborators include several major religiousbased non-governmental organisations, among others, Fast Action Response (ACT), the Amil Zakat National Agency (BAZNAS), the Indonesian Red Cross Society (PMI), Rumah Zakat, and Dompet Dhuafa Foundation Republika. Verified reports from neighbourhood and community units (RT/RW) to designated platforms are essential to support the success of this programme.
What was presented above represents institutional deviations, which, to a certain extent, resemble drift mechanism changes (Thelen, 2004;Streeck and Thelen, 2005;Mahoney and Thelen, 2010). During the case study, the recalibration was conducted on the implementation levels without altering the existing selective nature of the policy. Nevertheless, as these changes took place abruptly, without any guarantee of institutionalisation or wind-back, the argument on drift requires a longer period of observation.

Case study two: Health initiatives response General health measures
This section explores the inclusive policymaking process, which involves concerned groups in the response. Similar to the case above, the study found that the original design of health-related policies emerged from the extensive discussion of the TGUPP team in April 2020. From there, an exchange of ideas took place among stakeholders such as TGUPP, international organisations, an international consulting company, and the Faculty of Public Health at the University of Indonesia, which discussed the region's most urgent problems, as presented below.
Interviews show many actions were taken to avoid the spread of Covid-19 since Jakarta was announced as an epicentre. Several measures were taken, such as forming a twenty-four-hour call centre, creating a COVID-19 response team, and launching symmetric online information and one-data through the corona.Jakarta.go.id. On the belief that Covid-19 spread faster than bureaucratic processes, all resources were directed to crisis response and all on-site meetings and policy agendas were postponed. In dealing with pandemic situations, the joint team consisting of TGUPP, parliament, and Universitas Indonesia urged the Jakarta government to increase testing and health care capacities through collaboration and innovation to consolidate and integrate medical resources. The condition was even worse because the scarcity of medical equipment and lack of resource capacity triggered the government to act quickly.
Amid a precarious situation, the Jakarta government faced two-way pressure that came from the WHO, which urged the government to implement a lockdown on one hand, and the national government, which was dismissing the option of imposing a lockdown on the other hand. The main reason for this attitude was concern about the collapse of the national economy if the capital was locked down (Interview_5).
Instead of adopting the WHO's prescription or following the central government, Jakarta, benefiting from its privilege of autonomy, led the response through data-driven policy reflected in the implementation of large-scale social distancing, commonly known as PSBB, with great care based on 1) epidemiological data, in consultation with experts; 2) observation of indices related to public behaviour; and 3) preparedness of health facilities (factors: epidemiology, public health, health facilities). These three indicators were important for implementing or lifting restrictions based on the Governor Regulations Number 51 the Year 2020 (Doc_4; Doc_9).
The study found that produced response is made possible along with increased international organisation activities in policy making. Their involvement began at the onset of the crisis, in which TGUPP started to engage in partnership with them, pooling ideas to produce appropriate health responses (Interview_1; Interview_5). The next step in the process of pooling ideas was contextualisation. It began when the government engaged multiple actors to be involved to determine the health policy in different contexts and content. (Interview 2) In addition to the PSBB, international organisations, as our data showed, have other different specific functions when it comes to providing models for policy responses in Jakarta. International private consultant McKinsey, for example, provided the normative standard of PCR test procedures and other health-related protocols during the COVID-19 outbreak (Interview_3; Doc_7; Doc_8). Its prescriptions resulted from ideational exchange and mutual learning mechanisms from its global partners, who were also invited to join online meetings with TGUPP, high-ranking government officials, and other relevant parties in early April 2020. Meanwhile, another international private agency, Korn Ferry, took a different focus of cooperation (Interview_3). Its role was more concentrated on helping the government to develop a Standard Operational System for creating a COVID-19 referral system for hospitals across the city (Interview_3; Interview 4).
Below, we also demonstrate how ideas from various parties were applied. Table 3 constitutes the measurement matrix that became a guide for Jakarta's decision making. In this case, health issues were the root problem. This data was evaluated every week to track updated situations and used as a data baseline to create a concrete policy.
Besides, efforts to successfully control the spread were dependent upon the readiness of each region in developing and implementing the most appropriate strategies on a regional level. The Jakarta government, together with members of local parliament, initiated Local Government Regulation (Peraturan Daerah) Number 2 the Year 2020, ratified in November 2020. The regulation aims to provide public health protection and resilience, social protection, and economic recovery in the region as the impact of the Covid-19 crisis has been disrupting various aspects of the lives of people in Jakarta.

Responding to the second wave of transmission
To put it as a concern, the first wave of the pandemic can be traversed quite impressively before the second wave as the people commonly knew various spikes of the virus made the condition even worse (see Table 4). In addition, the precondition of unprepared infrastructure and people's behaviour aggravated the situation. While struggling to contain the outbreak and create health care reform, Jakarta was still able to respond quickly, consolidate responses among institutions and experts, and persist in innovating and scaling up the policy, as practiced in responding to the first wave of the outbreak.
Based on Table 5, the increase in the number of positive cases and the rise in the number of Bed Occupancy Rates (BOR) in Jakarta caused a crisis worse than the one in the first outbreak. Although health capacity had rapidly increased by about 80 per cent-90 per cent by May 2021 and the testing capacity of 104 laboratories exceeded 10,000 per day, a rise in positive cases was still inevitable.
As a precautionary measure, Jakarta's governor enforced the citizens of Jakarta to maintain health protocol, such as wearing double masks, washing hands, maintaining social distance, staying away from crowds, avoiding eating together, and reducing mobility.
Based on its experience with the first wave, Jakarta used the same policy patterns, maintained crisis responses, intensified public health based on data and science, and also assessed the workability and impact of its policies.

Vaccination strategy
The second wave of COVID-19 in Jakarta indeed caused a collapse in health care capacity and caused further economic turmoil. A report from the territorial apparatus retrieved from the COVID-19 website in Jakarta indicated that the number of red zones in Jakarta was rising  significantly (Doc_5). In this case, vaccination was one of the preventive efforts to hypothesise health protection and curb the spread of Covid-19 cases. This required the high participation of the community, especially in groups which qualified for the injection based on certain age and health conditions, to support the running of vaccination programmes and help achieve the goal of herd immunity: wherein 70 per cent of the population are vaccinated (Doc_3). Amid the lack of medical equipment and resources, the government instead encouraged ground-breaking policies by transforming a narrow paradigm of health services into one that was more universal. The new paradigm was based on a similar conceptualisation of how universal health risks impact the economic performance of cities. This conceptualisation led to the adoption of similar policy goalsnamely, the need to change the provisions and to provide everyone with access to vaccinations and free quarantine services (Doc_5).
The acceleration of the vaccination programme in Jakarta can be seen in its structured and well-managed efforts as the government committed to deploying more than 100,000 doses every day (Doc_2). The widely used operating model of vaccination determined the government programme and showed how the importance of this programme could become a movement or campaign. In terms of raising awareness, persuading people to get the injections and the responsibility of handling this crisis, top-down bottom-up initiatives should be aligned. Such a cooperation between all parties, whether it is the government or community elements, is required so that the increase in cases can continue to diminish (Interview_3; Interview_6; Interview_8).
Thus, as a metropolitan area with more than 10 million inhabitants, Jakarta needed to provide information and services that were fast and sufficient. During the pandemic, Jakarta construed the restrictions on in-person meetings as an opportunity to pursue digital alternatives and transform urban systems so that they could deliver excellent public services (Interview_2). At the moment, people in Jakarta can access all essential information in 'super apps' called JAKI. They even allow people to register for vaccination and see vaccination coverage and vaccination quotas in Jakarta's medical facilities. Thus, although many features can still be improved, the handling of the COVID-19 crisis by using technology has provided valuable lessons for enhancing the data-driven policymaking process (Interview_1). These policies must save the lives of people who tested positive, protect the lives of those who are fine, and help people recover from this crisis (Interview_3). Although the vaccination programme has been disrupted by alternative viewpoints, compliance with protocols, along with the operation of more than 700 vaccination centres in September 2021, allowed Jakarta to inject 10,221,172 (114.3 per cent) people, 65 per cent of whom were Jakarta residents based on ID (Doc_11). When dealing with the so-called 'delta variants,' some vaccinated people can still test positive, but the condition of health care facilities is much more controlled. This proves that vaccines were still engulfed by the virus but with lower rates of severe illness and death.
All in all, based on the health policy response used in the first and second 'waves,' Jakarta prioritised evidence-based decision-making and collaborative decision-making in every strategy and action they took to address the crisis.
The following is a summary of three principles applied in handling public health problems during outbreaks, with examples of real actions taken. The first is transparency. It provides factual information verified by experts, transparently without undue censorship for citizens, including daily Covid-19 updates via Jakarta's official social media, the corona.Jakarta.go.id website, and JAKI city 'super apps' related to Covid-19. The second is collaboration. It welcomes and actively engages the participation of various stakeholders, from private sectors, civic communities, international organisations, and academics in order to intensify 3T (Testing-Tracing-Treatment) strategy through innovation and partnerships. The last is science-driven action. This requires professional organisations, epidemiologists, behavioural scientists, and other experts from various fields to obtain innovative recommendations that are relatable on the ground to carry out a weekly evaluation of the handling of Covid-19 cases in Jakarta.
As far as policy changes are concerned, what was presented above appears to be different from SSN. Health care provision during the COVID-19 outbreak represented more drift, completed by changes in the policy paradigm as the selective nature of the policy was relaxed. Nonetheless, as already stated, the term 'drift' is inappropriate in this context, since it derives from the theory of gradual change. For this reason, the path creation accelerator is considered more appropriate.

Discussion
The purpose of this study is to explain what led to the path-shifting change in the way Jakarta's social policy was implemented by examining the production of ideas within SSN and health policy.
Furthermore, two mechanisms of policy response or change are provided. In the first mechanism, which we explored in SSN, formal rules on selection are deliberately held constant in the face of large shifts in the environment while mobilising resources to achieve goals. The second mechanism can be seen in the health sector when policy makers redirect previously selective provisions toward new purposes, resulting in universal services. However, such reconfiguration of a long-standing institution because of some exogenous shock is a rare empirical phenomenon. Together, these reconfigurations are often referred to as path creation accelerators: an intervention to revise established institutions and give way to new ones in a temporary capacity.
In SSN, we confirmed that ideas adopted to craft social safety nets were drawn from crises experienced in the past instead of taking novel initiatives that might have created unknown policy risks. The unchanged feature also includes the selective nature of coverage schemes directed to vulnerable groups, with consideration given to economic status as a reference. A significant change happened only in the implementation domain, where this process is performed by mobilising non-state actors, setting the current arrangement apart from its state-centred delivery model.
The uniqueness of Jakarta's decision warrants an in-depth examination, especially on the ideational aspect that underlay such an abrupt decision. In our study, Anies Baswedan and TGUPP succeeded in persuading the senator to adopt a new way of implementing SSN by framing it under the spirit of Large-Scale Social Collaboration (KSBB). The concept of collective action was inspired by the Gotong Royong value, which can be understood simply as addressing collective problems through collective action. In sum, the Gotong-Royong principle, along with risk and vulnerability frames, contributed to the proposal's being passed unanimously and with minimal discussion in the early stages of the outbreak. With this quick decision, transferring the benefit to SSN becomes faster, easier, and less stigmatising; it has allowed other partnered stakeholders to focus on clients who need social services. In this instance, we might suggest that the ideas behind the policy are formed by moral values.
Furthermore, we found different mechanisms for producing ideas in health care. The study indicates that this policy change model in health policy can be attributed to social learning processes from neighbouring countries. Previously, Singapore and Malaysia confirmed their first cases in January and February, respectively. Despite having a difficult time identifying its contextual policy, the TGUPP chose to look at the baseline policy used by neighbouring countries. What has been presented here can be thought of as an emulation model of policy learning (Marsh and Sharman, 2008), which takes place when ideas for policy are transferred voluntarily from other settings, including the best practices carried out in another country. Jakarta, benefiting from its privilege of autonomy, produced the response through data-driven policy instead of emulation per se. Data-driven ideas can be developed as a result of increased activity in international organisations and consultants in policy making (see findings for details). At the outset of the pandemic, a variety of organisations and consultants began to collaborate with TGUPP on building an appropriate response, starting with the standardisation of PCR test procedures until the development of a Standard Operational System for COVID-19 referrals to hospitals across the city. From a health case study, we learned that involving an array of stakeholders in policy formulation broadens the point of view and provides multiple perspectives. Generally, it attracts ideas from the outside and pushes innovation inside. Reconsidering the many inputs that have been collected, it is necessary to contextualise all of the information and to ensure that everything also meets standards, including practical usability. At this point, contextual emulation is proposed for framing the Jakarta case.
Furthermore, our analysis also pointed out that the presence of policy entrepreneurship during the COVID-19 crisis had a greater effect than during the Asian economic crisis. As presented in strands of social policy literature, citing Indonesia and Southeast Asia (for example, Aspinall, 2014;Yuda, 2019), showing that breakthroughs created by policy entrepreneurs will not always find sufficient support from the parliament unless the ideas are in line with political or ideological considerations.
What is presented here, however, differs from what is typically found in other developed nations which produce resilience responses. The resilience dimension can be explained further by anchoring on the threat rigidity hypothesis, meaning policymakers are inclined to 'stick to what they know best' (Starke et al., 2013:10, quoted in Moreira andHick, 2021: 262), and to prevent exacerbating the effect, policy adoption is pursued. And from this instance, we learn that in the less stable welfare regimes, path creation prevails over historically-driven institutions when it comes to reacting to an unknown emergency. It is conducted by considering advice from both international and domestic organisations, including adopting a broadly similar path of response as implemented by neighbouring countries through contextualised steps. Leveraged by evidence-informed policy and practice, the role of the policy entrepreneur becomes more significant in this last step.

Conclusion
Studying institutional change, including the triggers, drivers, and limitations of deliberate reform, is integral to the social and political sciences. While scholars have made numerous outstanding contributions to this topic, much work remains before we clearly understand COVID-19 in Indonesia in particular and in developing countries in general. The unexpected elements of the crisis have stimulated reforms of social policy institutions in response to new and unknown risks posed by this crisis.
There are two core features of the policy institutions that can be highlighted during the COVID-19 crisis in Jakarta. First, there was an increase in the centrality of risk-oriented consideration for welfare provision over means-tested determination (health case). Second, there was the mobilisation of non-state actors to undertake initiatives targeted at mitigating the social-economic risk of households affected during the lockdown period (SSN). The latter disrupted long periods of a government-centred social delivery system in Jakarta. Applying ideational approaches to this case, therefore, could result in a better understanding of the conditions under which social policy response in Jakarta experienced the path creation accelerator process. This includes capturing the important role of ideational exchange between international and domestic organisations and policy entrepreneurs to promote the responses necessary to develop more appropriate policies.
As a contested concept, the notion of continuity and change in policy dynamics, as presented in the Jakarta case, cannot fully be regarded as a path-dependent effect. The notoriously fast pace of policy response in Jakarta compared with that of other regions shows us that the exchange of ideas between international organisations and domestic agencies was key to the formulation of a response. While the ideas were continually evolving, TGUPP, a policy entrepreneur community, took the role of translating the ideas into action by managing the sequences and timing of responses and producing the path creation. As exemplified by the emphasis given to health policy responses, the principal necessity brought about by the COVID-19 crisis steered the response to be more generous and universal.
However, it is worth noting here that, though innovation prevails, the idea of path dependency does not necessarily lose its significance in keeping policy to 'stay partly' on the bounds of the legacies continuum. The case of SSN has demonstrated how the response produced was encased within the idea of a 'selective welfare paradigm' which previously shaped the social policy characteristics in Jakarta.
This article has attempted to enhance our understanding of what Pierson (1994) conceptualises about the 'lock-in' effect, which is that it is 'no longer fully' relevant to account for institutional dynamics in developing countries and perhaps in the COVID-19 crisis. In essence, what appears to be a 'frozen' institutional landscape could thaw out and result in policy change. Additionally, ideas themselves will have significant effects on policy change only when they find the 'window of opportunity' which is represented by the uncertainty around the COVID-19 outbreak.