November 2020 Driving the COVID-19 response from the center Institutional mechanisms to ensure whole-of-government coordination Jana Kunicova World Bank Governance Global Practice Acknowledgments nance and Institutions COVID-19 Rapid This working paper is part of the Gover­ Response series spearheaded by the Governance Global Practice, World Bank. The cova, Senior Public Sector Specialist and Chair paper was developed by Jana Kuni­ of the Center of Government Community of Practice, between May and October 2020. The paper presents initial thinking and uses immediately available data, as it aims to document the emerging knowledge, provide the grounds for just- in-time policy and institutional advice to governments, and suggest avenues for further research. The author worked closely with Robert Beschel and Jim Brumby, and under the overall guidance of Ed Olowo-Okere. She is grateful to Ali Halawi, Jens Kristensen, Peter Ladegaard, Tracey Lane, Alberto Leyton, Roland Lomme, and Ismail Radwan for helpful discussions and comments. All questions should be directed to jkunicova@worldbank.org. 2 Photo by Tai’s Captures on Unsplash Executive Summary Facing the COVID-19 pandemic, governments around the world have set up various coordination mechanisms at the center of government (COG) to facilitate pandemic response. Based on what we know from the COG literature and World Bank’s ex- perience, the paper distills a set of options for whole-of-government (WOG) coor- dination at COG when responding to a pandemic. Recognizing that “good fit” is as important as “good practice”, it explores a range of institutional mechanisms that could be employed for coordination at COG, from simple to sophisticated. Different COG mechanisms will be effective and fit-for-purpose in different country con- texts, such as developed or developing countries, governments with high capacity or low capacity, or federal or unitary constitutional systems. The broader COG literature provides some important lessons for the coordination during COVID-19 pandemic. The choice of the COG mechanisms for pandemic re- sponse will interact with a complex set of other institutions and processes that in- 3 fluence coordination. Therefore, simplicity and flexibility of the institutional design for a COG coordination mechanism for pandemic response are critically important, as long as they are underpinned by high-level leadership and aligned incentives. Also, during a pandemic crisis, best-performing COGs react and deliver results fast, as lives and livelihoods are at stake. In this sense, whatever institutional form COGs adopt for the pandemic response, it will fulfil functions that may be similar to those of a delivery unit (DU). These lessons can be further distilled to the following:  Secure strong backing and involvement of the top leadership. Success depends critically on the government’s chief executive being committed to evidence- based policy making and implementation. This requires that she be directly involved in its routines. Continued support from the chief executive is of course conditional on the delivery unit’s performance.  Create a focused and granular results platform. Successful delivery systems generally focus on a limited number of well-defined and operationalized strategic priorities. It is also critically important to prevent a mission creep and firefighting, which means that a COG mechanism should have a well-defined mandate.  Create institutional interface with MDAs. Units at COG do not implement the policy prerogatives by themselves – the MDAs do. COG’s role as a driver of MDAs’ performance is greatly facilitated if they have dedicated counterparts within the implementing MDAs (either designated units or staff). Because MDAs are the implementers of government’s top priorities, the real action takes place on their turf, not at the Prime Minister’s office. A delivery system strengthens the link through introducing the routine of reporting and regular problem-solving meetings where unresolved issues are progressively escalated.  Create simple dashboards, reporting routines, and optimize the amount of reporting by MDAs. Whenever possible, build KPIs around the existing indicators that MDAs already report on. Work across the performance ecosystem to create synergies with existing reporting structures, such as existing performance-based budgeting or national development plan reporting.  Create a simple problem-solving mechanism. A successful delivery system not only tracks progress in implementation but is also actively involved in clearing up bottlenecks between MDAs. These incentives work only with the regular involvement of the Prime Minister or President through 4 routines, such as the problem-solving meetings. To create incentives to solve problems at the lowest levels, it is important to create a progressive escalation mechanism, so only persistent bottlenecks that cannot be solved at the technical level reach the policy maker.  Communicate, communicate, communicate. Communication becomes a daily routine across multiple platforms. Based on the stylized facts about the government response to the pandemic and the reviewed COG literature, the essential functions of successful COGs could be categorized as follows:  Policy-setting and decision-making (e.g. establishing Command Centers). A Center of Command (CoC) is a “nerve center” of the pandemic response and works best when it is close to the chief executive. It should be tasked with high-level oversight of the whole-of-government response.  Operational coordination (e.g. oversight of action plans, mobilization, financing and interagency coordination). This includes oversight of action plans, underpinned by budgets, as well as inter-agency coordination systems. Response plans with concrete targets should be activated. Consequently, detailed delivery maps with granular implementation responsibilities, accountabilities, timelines can be drawn. Systemic approach could also include sub-committees reporting to CoC that are responsible for particular aspects of operational coordination.  Information gathering and M&E (e.g. dashboards and monitoring routines). This is achieved through dashboards and monitoring routines, whereby responsible MDAs feed the data into dashboards at regular intervals. This should be linked to the dedicated interface within implementing MDAs. Problem-solving and de-bottlenecking mechanisms should accompany the monitoring, as should internal communication mechanisms to keep various parts of government on the same page.  External communication (e.g., press briefings, media campaigns). This can take various forms, from daily press briefings with high-level officials through diligent communication campaign on multiple platforms promoting 5 handwashing, social distancing, and face covering. Transparency about infections, hospitalizations, and mortality is also important to reinforce trust and therefore compliance. Successful COGs will seek to fulfill these four basic functions to handle the pan- demic regardless of the country context. However, the institutional forms will adapt across different contexts. The larger the country, the higher the relative importance of vertical coordination vis a vis horizontal coordination. For gov- ernments with lower capacity, the simpler the mechanism the better. Thus, small countries with low government capacity would do best with simple CoG coordi- nation mechanisms focusing on horizontal coordination, while large low-capacity countries can be advised to focus on vertical coordination. Small high-capacity countries are well-advised to adopt more complex mechanisms emphasizing hor- izontal coordination, while larger high-capacity countries will be best served by a matrix structure. Other context variables, such as urbanization or inequality, will further affect the type of coordination required. Finally, various additional fluid factors, such as trust in government and quality of leadership, will have an impact on how CoG coordination will work and to what extent it can succeed. Some initial conditions that are relevant for the success of COG coordination can change in response to how the government handles the spread of the virus. These include government’s credibility and legitimacy, which are related to trust in gov- ernment. While initial credibility and trust in government can help fight the virus, policy and institutional responses to the virus can also influence how much trust and legitimacy the government accrues (or loses) during the pandemic. Leadership is another condition that has profound effect on the success of the chosen CoG coor- dination mechanism and can change in the medium term depending on the success of the pandemic response. This is especially true in democracies, where elections are likely to be seen as referenda on the incumbent’s handling of the pandemic. On balance, some initial conditions can have a stronger effect on the success of the pandemic response than CoG coordination; however, well-chosen CoG mech- anisms can in turn improve some of these conditions. In those countries where trust and legitimacy of the government are low, it is particularly important to adopt CoG policy and institutional responses that emphasize transparency and communication, thus increasing credibility of the government. 6 Table of Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Toward a framework: Stylized facts and key questions. . . . . . . . . . . . . . . . . . . . . . . . . 9 Form and function of COG for pandemic response. . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Lessons from the literature on coordination and driving performance from COG. . . . . . . . 16 Scanning the horizon: Emerging COG mechanisms to coordinate COVID-19 response. . . 21 Distilling COG functions relevant for coordinating pandemic response. . . . . . . . . . . . . . . . . 22 Relevant features of the country context for COG coordination choices. . . . . . . . . 25 Sticky initial conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Fluid initial conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Annexes 1. Establishing Center of Command (CoC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 2. Systemic Approach: Planning, Financing, and Inter-Agency Coordination . . . . . . . . . 44 3. Evidence-based response: Data and M&E tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 7 4. Communication tools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Introduction COVID-19 pandemic has brought into sharp focus the need for coordination of government response at the center of government (COG). The pandemic response is an ultimate test for the ability of government machinery to work in sync to fight the rapidly spreading invisible pathogen. Depending on constitutional and political factors, COGs around the world vary in terms of their own authority to command and control policy. However, virtually all COGs have a function to steer collaboration and communication among different ministries, departments, and agencies (MDAs), as well as across national and subnational levels of government. Different countries have adopted different coordination mechanisms at the center, with varying success. The purpose of this note is to guide the World Bank advice to governments on how to set up coordination mechanisms at COG to facilitate pandemic response. This can be just-in-time advice, or future technical assistance, as our member govern- ments think through their preparedness plans for the future. The primary audi- ence comprises country management units (CMUs) and Governance Global Prac- tice (GGP) task team leaders (TTLs) looking to offer to their counterparts a menu of options for Bank support. The note may also be useful more widely to policy- makers and their advisors around the world and researchers in related fields. This note focuses on institutional options for COG mechanisms that are fit for purpose in particular country contexts, rather than rigid prescriptions drawing from best practices. Based on what we know from the COG literature and Bank’s experience, the note distills a set of options for whole-of-government (WOG) co- ordination at COG when responding to a pandemic. Recognizing that “good fit” is as important as “good practice”, it explores a range of institutional mechanisms that could be employed for coordination at COG, from simple to sophisticated. It asks what could work, and what should be avoided, in different country contexts. 8 Photo by Hello I’m Nik on Unsplash TOWARD A FRAMEWORK Stylized facts and key questions Pandemic response has been compared to a natural disaster response. However, while disaster response management literature offers some useful lessons, there are also limits to its applicability to a pandemic. The key differences include:  Overlapping phases of indeterminate length. The three phases of disaster management – response, recovery, and rebuilding – are much less distinct in a pandemic. They may overlap significantly and can become protracted but with no clear ending. Moreover, the path may not be linear, as multiple outbreak phases over time are possible and even likely.  Imperfect information. The disease of a pandemic – definitionally – is not well understood, so the policy response needs to adapt as more is known and understood about the contours of the disease. Scaling up testing is complex and expensive. If there is insufficient testing, there is limited information available to determine whether the country is in the response or recovery phase.  Global nature. Unlike natural disasters that are localized, pandemics are global. Because virtually all countries are vulnerable and many are dealing with the challenge simultaneously, there are distinct challenges to mutual aid. When earthquakes or tsunamis strike, international community comes together to 9 help the affected countries. However, in the current pandemic, nations may not be in a position to aid other nations as they themselves struggle with shortages of medical personnel, equipment, and overextended healthcare systems.1 Also, unlike a disaster response where the roles and responsibilities of major global players can easily be identified (UNOCHA, UNHCR, HCRC, UNHCHR, UNICEF, etc), major multilaterals other than WHO are not easily apparent.  Preparedness vs coping. Given the scale of the COVID-19 pandemic, even the best prepared governments did not have blueprints comprehensive enough to deal with the crisis of such magnitude. It is therefore important to recognize that there are limits to preparedness. In addition to being prepared, governments need to be able to cope in real time. This means government systems that are flexible, agile and limber, so they can steepen their learning curve and quickly come up with contingency measures that take the future recovery and rebuilding with resilience into account. Given the differences between a pandemic and a natural disaster, a useful starting point for the analysis is a set of stylized facts regarding key government actions that help stave off the spread of the virus. Governments that successfully con- tained the pandemic to date focused on the tasks that included the following: closing borders to prevent new spreaders of the virus from entering the coun- try; mandating spatial distancing and facial coverings to slow down community spread; closing down non-essential public and private sector activities once the community spread took hold; scaling up testing; contact tracing for positive cases and isolating those exposed; procuring emergency medical equipment (e.g. venti- lators) and personal protective equipment (PPE) for healthcare workers; ensuring sufficient number of ICU beds for COVID-19 patients; providing economic support for enterprises affected by the virus, either by sector or by enterprise size; main- taining social safety nets and basic income protection; and eventually implement- ing a phased reopening of the economy to support recovery. These key government actions are undertaken by various ministries, departments and agencies (MDAs) across government, as well as at different levels of govern- ment. A more unified response is likely to be a more effective response. The min- 1  Because COVID-19 pandemic hit many developed nations early, the world has seen a reverse flow of medical aid. In a particu- larly poignant example, Somalia and Albania sent medical personnel to hard-hit Italy in March-April 2020. 10 See https://bit.ly/3iZYgDk; https://reut.rs/2Fqzpec istry of interior will do best to work together with the health ministry on border closures and quarantine measures. Local authorities will be best advised to work with the central agencies to ensure contact tracing and isolation of exposed individ- uals. Education ministries will need to work closely with health ministries on school closures. The plans for phased reopening will be most effective if they involve rep- resentatives across all involved MDAs as well as private sector and civil society. The multi-sector nature of the crisis leads to the competing and conflicting prior- ities within government, putting additional pressure on CoG coordination mecha- nisms. The crisis, which was originated as a public health emergency, has evolved into a multifaceted and unprecedented emergency with long lasting economic and social consequences affecting most areas of public policy and public services. This situation demands from government complex and multi-sector responses. These responses need to provide solutions in multiple fronts at the same time, as well as triage among competing and conflicting objectives and priorities. For example, as the crisis evolves, requirements to address economic recovery objectives or the re-opening of schools start clashing with the need to continue to contain con- tagion through social distancing and other similar restrictions driven by health priorities. Ministers of education or economy bring their reopening plans backed by their own evidence, which in many cases will confront priorities set by health authorities. Similar confrontations occur at sub-national levels with local author- ities setting their own priorities based on demands driven by different composi- tion of their constituencies, or even just by the stance of their political affiliation. All of this puts additional pressure on CoG structures and processes. Although CoG is usually designed to tend to such kind of tensions in normal circumstances, the stakes during the crisis are much higher. This underscores the need to put in place appropriate coordination and decision-making processes at CoG. Country context matters in determining the best choice of coordination mecha- nism. The coordination of this response is a gargantuan task that is best delivered by the COG. However, the appropriate COG mechanism will differ by country con- text: different coordination mechanisms will be fit-for-purpose in developed or developing countries, for governments with high capacity or low capacity, or for constitutional systems that are federal or unitary. 11 Against the backdrop of these departure points, a possible framework for thinking about COG institutional responses to the pandemic includes addressing the fol- lowing three sets of questions:  Form and function of CoG for pandemic response: What is the role of COG to ensure that the key tasks for pandemic response are delivered effectively? What can we glean from the existing COG literature on coordination about what may work and what may not in terms of institutional mechanisms? What have governments been doing so far? What are the key functions that COG perform to coordinate the pandemic response horizontally across MDAs and vertically across levels of government? What corresponding forms may they take?  Relevant features of country contexts. Not all COG coordinating mechanisms will work equally well everywhere. What aspects of the country’s economic, political, and institutional context make some COG mechanisms more suitable than others? In other words, what are the relevant features of the country context that make COG mechanisms “fit for purpose”? Some features of country context may be “sticky,” or slow to change; these include the level of development, overall government capacity, constitutional structures, and administrative or cultural norms such as individualism or collectivism. Other context features – such as level of trust in government, or quality of leadership – are more fluid and can change in short to medium term. To make matters more complicated, these fluid factors may also be affected by the success of pandemic response.  Different CoG options for different country contexts: Are there typologies that may allow for narrowing down COG options best suited for the given context? What are examples of some of the well-functioning COG coordination mechanisms that contributed to success in containing the pandemic in different contexts? Although each country case is unique, what might work and what should be avoided? 12 The remainder of this paper explores these interrelated questions. As Figure 1 suggests, the success of pandemic response depends in part on how well the cen- ter of government can coordinate. However, the particular COG forms, and to some extent its functions, will be determined by the country context. The paper discusses both in turn, and then matches different types of COG activities most suitable to types of country contexts. Figure 1. Toward a framework: Interdependencies between country characteristics, CoG coordination options, and pandemic response Pandemic Response CoG Institutions Country Characteristics 13 Box 1 Defining Center of Government (COG) and its functions Conceptually, COG refers to the institution or group of institutions that provides direct management support to the chief executive (James and Ben- Gera 2004; World Bank 2010a). Technically, there is a COG at both national and subnational levels, at least in federations. Unlike service delivery-oriented ministries and agencies, COG institutions deal with the strategic management, coordination, monitoring, and communication of government decisions. COG refers to institutions that are placed at the apex of executive power. The term “institutions” is used broadly, to include government bodies, but also rules, roles, people and organizations. Thus, COG comprises institutions such as the Office of the President or Prime Minister, as well as cabinet offices, sub- cabinet committees, and other central coordinating mechanisms, including the ministries of finance and development planning. If the COG performs its tasks well, collective expertise from across the public sector is mobilized and brought to bear on the most pressing decisions confronting the country. MDAs with a stake in a particular issue are consulted, and their views and technical knowledge are fully integrated into the decision process. Senior officials have the opportunity to thoroughly weigh and review various options and to fully understand their legal, financial, and policy implications, resulting in evidence-based decision-making. Once decisions are taken, ministries move forward with a clear set of directives, and adequate resources to implement them effectively. Incentives for implementation, such as systematic monitoring and evaluation (M&E) and public accountability to the chief executive, can also be put in place by the COG. Source: Adapted from World Bank 2018. Improving Public Sector Performance Through Innovation and Inter- 14 Agency Coordination. Photo by Camila Perez on Unsplash Form and function of COG for pandemic response This section draws on the COG literature and empirical exploration of the emerg- ing data to define the key functions of COG for pandemic response. It begins by discussing lessons for the pandemic response from the existing literature on COG, coordination, and driving performance from the center. It then scans the horizon to see what form various COG coordination mechanisms of pandemic response have taken around the world. Against the backdrop of these various forms, the next section distills the key functions that COG should fulfil to coordinate the re- 15 sponse to the pandemic and matches them to the observed forms. Lessons from the literature on coordination and driving performance from COG Coordination is challenging for most countries. Box 1 provides a working definition of COG as well as its functions, which focus on coordination at various stages of the policy process. These functions are directly relevant for the pandemic response, given its inter-ministerial and inter-governmental nature. However, as Box 1 points out, the reality in most developing countries stops short of the intended ideal. This is why deriving lessons from the existing literature is particularly important for de- ciding which coordination mechanisms to adopt for the pandemic response. There are three key lessons from the broader COG literature of relevance to COVID-19 pandemic. First, the choice of the COG mechanisms for pandemic re- sponse will interact with a complex set of other institutions and processes that in- fluence coordination. Second, simplicity and flexibility of the institutional design for a COG coordination mechanism for pandemic response are critically important, as long as they are underpinned by high-level leadership and aligned incentives. Third, during a pandemic crisis, best-performing COGs react and deliver results fast, as lives and livelihoods are at stake. In this sense, whatever institutional form COGs adopt for the pandemic response, it will fulfil functions that may be similar to those of a delivery unit (DU). We now explore each in more detail. First, the choice of the COG mechanisms for pandemic response will interact with a complex set of other institutions and processes that influence coordination. The existing literature distinguishes among various formal and informal mechanisms of COG coordination that build on a broader environment, determined by political and social factors. Table 1 provides a “conceptual map” of coordination. It outlines three levels of coordination: whole-of-government policy coordination, inter-agency horizontal coordination (across MDAs), and vertical coordination (across level of governments). It also distinguishes between formal mechanisms whose primary purpose is coordination (e.g., delivery units), and processes that have been creat- ed for a different primary purpose and yet strongly influence coordination (e.g., budget process). The COG mechanisms analyzed in this note fall mainly into two highlighted cells in Table 1. However, they cannot be considered apart from other structures and processes that already exist, as the latter influence both the choice 16 and effectiveness of COG coordination mechanisms for the pandemic response. Table 1. Government coordination: A conceptual map Broader Environment Political and Constitutional Factors: Social and Cultural Factors Government structure and fragmentation Leadership style (collaborative vs. hierarchical) Single-party state versus multi-party state Legitimacy, values, and vision Political cycle; period expected until next election Shadow of the future (iterative versus one-off (where relevant) engagement) Coalition versus single party governments Political party composition of the national and sub-national governments (e.g. “cohabitation” when Parliamentary versus presidential systems the leadership comes from different parties at the national and subnational level) Coordination Mechanisms Bilateral and Multilateral Inter- with Sub-national Whole of Government Agency Mechanisms Governments (Primary Focus is Policy (Primary Focus is Operational (Both Policy and Operational Coordination) Coordination) Coordination) Cabinet Coordination Ministries Regulatory practices and Formal Coordination Mechanisms standard setting; league tables Sub-cabinet committees Ministerial clusters with senior ministers Voluntary and involuntary Central agencies (President, PM and mandates Cabinet Office, Chancelleries) Formal and informal inter-agency working groups, task forces, etc. Inter-governmental councils Delivery units Dedicated liaisons and contact National and regional Expert panels and advisory boards points associations Established protocols for communications and information- sharing (working level) The Budget process Reorganization, mergers Legislative or parliamentary Practices that Influence Coordination bargaining Government-wide M&E Systems Staff secondments and rotations Joint training and preparation IT Systems Joint distribution lists, exercises conferences and retreats Generalist/Executive Service Cadres Use of properties, facilities and Combined training and staff Transparency equipment development Advisory services & counseling Professional networks and associations Joint messaging Media and social media networks Source: Adapted from World Bank 2018. Improving Public Sector Performance Through Innovation and Inter- Agency Coordination. 17 Second, simplicity and flexibility of the institutional design of a COG coordina- tion mechanism for pandemic response are critically important, insofar they are underpinned by high-level leadership and aligned incentives. Table 2 distills the World Bank experience in advising governments that have attempted to improve coordination at the COG. High-level leadership and well-structured incentives were preconditions for success, as is a degree of flexibility and focus only on stra- tegic issues rather than nitty-gritty implementation. On the other hand, complex- ity works against success, both in designing the mechanisms and in overlaying the existing structures. Moreover, overeager institutional borrowing from other contexts leads to isomorphic mimicry, so using existing structures is often more productive. Table 2. Lessons from global experience to strengthen coordination through reforms What has worked? What has not worked? Leadership. High-level political backing is Complex designs. Simple mechanisms often work important for any reform to enhance coordination, best in low-income countries and FCV contexts, where as is the quality of leadership. The person at the capabilities are often more limited. helm of the reform, if not the PM him/herself, should be technically skilled and politically savvy, as well as close to the chief executive. Incentives. The reforms that anticipated Adding new structures with overlapping functions. resistance and invested in buy-in were most likely There can be value in policy contestability, but the unless to succeed. well-defined, overlapping functions can blur accountability and make coordination difficult, both in sectoral coordination and in government-wide coordination. Flexible arrangements. Flexible and adaptive Relying solely on “best-practice” institutions. coordination mechanisms work better than Institutional solutions (e.g. delivery units) uncritically rigid and prescriptive ones, as they have a transferred from one context to another rarely better chance to be sustained and become self- produce the desired outcome, and have been reinforcing even as leaders change. criticized for ‘isomorphic mimicry’. Before introducing new institutional coordination mechanisms, it is important to take stock of what already exists. Building on the existing institutions tends to work better. COG focus on strategic issues. COG functions best when focused on strategic coordination and leaves the granular upstream and downstream coordination tasks to the MDAs. Source: Adapted from World Bank 2018. Improving Public Sector Performance Through Innovation and Inter- Agency Coordination. 18 Third, during a pandemic crisis, COG’s role is to predict, react, and deliver results fast, as lives and livelihoods are at stake. In this sense, whatever institutional form the COG adopts for the pandemic response, it may need to fulfil similar functions as a delivery unit (DU) does in relation to one sector or in relation to one specific delivery problem. While this paper does not advocate establishing a formal unit for a pandemic response, the literature on driving performance from the center of government offers important insights about the functions within COG that deliver tangible results relatively fast. Box 2 summarizes these lessons. These lessons have implications for the nuts and bolts of the selected COG coor- dination mechanisms. For example, if the government decides to form an in- ter-ministerial task force to coordinate its pandemic response, then the following needs to be ensured:   The terms of reference for the task force focus on the key issues, avoiding mission creep and firefighting;  Its monitoring capabilities are established (including data collection/ reporting, its frequency, data visualization, dashboard)  An institutional mechanism for regular interface with the implementing MDAs is established;  Troubleshooting mechanisms are well-defined and functional (both technical and policy level/escalation) when solutions are not moving, or virus flare-ups occur;  A communication strategy and its implementation (e.g., regular briefings) are established. The best institutional arrangements in the world still require a leader capable of steer- ing them. The personal competence of the leader, and in fact the entire leadership cadre, will certainly play an important role in galvanizing support and commitment. 19 Box 2 Lessons for pandemic response coordination structures derived from successful delivery units Secure strong backing and involvement of the top leadership. Success depends critically on the government’s chief executive being committed to evidence- based policy making and implementation. This requires that she be directly involved in its routines. Continued support from the chief executive is of course conditional on the delivery unit’s performance. Create a focused and granular results platform. Successful delivery systems generally focus on a limited number of well-defined and operationalized strategic priorities. It is also critically important to prevent a mission creep and firefighting, which means that a COG mechanism should have a well-defined mandate. Create institutional interface with MDAs. Units at COG do not implement the policy prerogatives by themselves – the MDAs do. COG’s role as a driver of MDAs’ performance is greatly facilitated if they have dedicated counterparts within the implementing MDAs (either designated units or staff). Because MDAs are the implementers of government’s top priorities, the real action takes place on their turf, not at the Prime Minister’s office. A delivery system strengthens the link through introducing the routine of reporting and regular problem-solving meetings where unresolved issues are progressively escalated. Create simple dashboards, reporting routines, and optimize the amount of reporting by MDAs. Whenever possible, build KPIs around the existing indicators that MDAs already report on. Work across the performance ecosystem to create synergies with existing reporting structures, such as existing performance-based budgeting or national development plan reporting. Create a simple problem-solving mechanism. A successful delivery system not only tracks progress in implementation but is also actively involved in clearing up bottlenecks between MDAs. These incentives work only with the regular involvement of the Prime Minister or President through routines, such as the problem-solving meetings. To create incentives to solve problems at the lowest levels, it is important to create a progressive escalation mechanism, so only persistent bottlenecks that cannot be solved at the technical level reach the policy maker. 20 Communicate, communicate, communicate. Communication becomes a daily routine across multiple platforms. Source: Adapted from World Bank 2017. Driving Results from the Center: Malaysia’s Experience with PEMANDU. Scanning the horizon: Emerging COG mechanisms to co- ordinate COVID-19 response As COVID-19 pandemic crisis spread around the world, different countries have opted for different institutional coordination mechanisms for pandemic response at COG. The World Bank Governance Global Practice (GGP) has recently launched its COVID-19 Response Tracking Portal (RTP) that focuses on institutional and policy responses as reported by Bank staff around the world. RTP codes all re- sponses along GGP business lines, including Public Institutions Reform, and fur- ther breaks them down by “action categories” under each business line. One of the action categories under PIR is “center of government, leadership, and com- munication,” which allows extracting the relevant actions from the database. As of June 4, 2020, the database included 50 COG institutional and policy responses in 32 countries across all world regions.2 An initial analysis of this qualitative dataset yields the results discussed in this section. Many of the recorded government actions amounted to creating special structures at COG, often led by the chief executive, that coordinate the pandemic response. For example, in Bolivia, a Crisis Management Team was set up at the President’s office. Similarly, Central African Republic created a Crisis Committee led by the President. Cambodia set up a National Response Team led by Prime Minister (PM), while the Kyrgyz Republic created Operational Headquarters chaired by PM. Viet- nam created a National Steering Committee chaired by Deputy PM. Lao PDR set up an inter-ministerial task force chaired by Deputy PM and Minister of Finance. Some governments further established operational sub-committees addressing specific dimensions of the challenge. For example, in Uruguay, the Departmental Directions of the Health Department and the Departmental Centers for Coordina- tion of Emergency are coordinating actions at the local level, for instance, moni- toring people quarantined after coming back from high-risk countries. Cameroon established a consultation framework between the Ministry of Finance and the Ministry of Economy and Planning, with the main economic players, in order to mitigate the effects of the crisis and encourage a rapid resumption of activity. 2  The database is continuously updated. It is also illustrative rather than comprehensive, as the type and number of entries depend on the assiduousness and background of the particular contributing staff. However, the current entries provide a 21 broad sweep across the world documenting institutional responses. Yet others created a secretariat supporting the government’s response, or a des- ignated ministry providing technical support. The Government of Grenada’s COVID-19 Economic Support Secretariat, (CESS) has officially been established to implement the economic stimulus package. Relatedly, some COGs devised a mech- anism for reaching out to external medical and scientific expertise, and for chan- neling this expertise into government. In the Dominican Republic, a high-level commission led by health experts and with representation of key line ministries has been set-up to coordinate response to COVID-19. Some COGs also established structures to raise and supervise the distribution of funding for emergency response. Djibouti established a COVID-19 Emergency and Solidarity Fund to ensure transparency in the collection of domestic resources and the execution of the expenditure related to the COVID-19. Kyrgyz Republic created two commissions on the distribution of donations from individuals to fight coro- navirus infection and individuals and legal entities to provide food for low-income families. India leads in establishing a COVID 19 Emergency Fund for South Asia. Distilling COG functions relevant for coordinating pan- demic response Based on the stylized facts about the government response to the pandemic and the reviewed CoG literature, the essential functions of successful COGs could be categorized as follows:  Policy-setting and decision-making (e.g. establishing Command Centers). A Center of Command (CoC) is a “nerve center” of the pandemic response and works best when it is close to the chief executive. It should be tasked with high-level oversight of the whole-of-government response.  Operational coordination (e.g. oversight of action plans, mobilization, financing and interagency coordination). This includes oversight of action plans, underpinned by budgets, as well as inter-agency coordination systems. Response plans with concrete targets should be activated. Consequently, detailed delivery maps with granular implementation responsibilities, accountabilities, timelines can be drawn. Systemic approach could also 22 include sub-committees reporting to CoC that are responsible for particular aspects of operational coordination.  Information gathering and M&E (e.g. dashboards and monitoring routines). This is achieved through dashboards and monitoring routines, whereby responsible MDAs feed the data into dashboards at regular intervals.3 This should be linked to the dedicated interface within implementing MDAs. Problem-solving and de-bottlenecking mechanisms should accompany the monitoring, as should internal communication mechanisms to keep various parts of government on the same page.  External communication (e.g., press briefings, media campaigns). This can take various forms, from daily press briefings with high-level officials through diligent communication campaign on multiple platforms promoting handwashing, social distancing, and face covering. Transparency about infections, hospitalizations, and mortality is also important to reinforce trust and therefore compliance. The observable COG responses to the pandemic around the world can be easily cat- egorized as fulfilling one of the above four functions. Table 3 matches these four functions with various forms that the institutions have taken and gives examples of simple or more complex institutional mechanisms that may be fit-for-context. In addition, Annexes 1-4 provide the categorization along these four functions of the qualitative data on COG mechanisms from the World Bank COVID-19 RTP. Even though RTP is illustrative rather than comprehensive, Annexes 1-4 provide an interesting picture. In particular, it is notable that most recorded mechanisms refer to either establishing a CoC (21/50) or some element of a systemic approach (22/50). There are only three evidence-based/monitoring-related mechanisms re- corded in the database, and only four communications-related mechanisms. This may be a function of what is recorded rather than a reflection of the prevalence of these systems on the ground, but it does point to the necessity to emphasize these two crucial COG functions to both policymakers and development professionals. 3  The pandemic brings also the uncertainty and the need to plan for evolving scenarios. Effective use of data an evidence is only part of the puzzle, given the uncertainty about how the situation may evolve over time. Under such circumstances, CoG are called to develop alternative plans for alternative scenarios by using scenario planning methods and other similar tools. For example, in Romania, WB governance team is currently advising the government on scenario planning for school 23 reopening, through a process managed from the CoG. Table 3. Form follows function: Examples of various mechanisms corresponding to key COG functions Function Policy-setting and decision-making (e.g. establishing CoC, high-level oversight, and WOG response) Forms Cambodia. National Response Committee (NRC), chaired by the Prime Minister. The NRC is responsible for: (a) Identifying national policy and strategy in response to COVID-19; (b) Leading the implementation plan to prevent, protect and control COVID-19; (c) Minimizing political and socio-economic impacts of COVID-19; and (d) Leading and facilitating the implementation of multi-sectoral and inter-ministerial measures at national and sub-national levels. Central African Republic: A crisis committee led by the President has been established to: (i) ensure coordination of COVID-19 response actions throughout the country; (ii) approve the preparedness and response plan to COVID-19; (iii) mobilize financial resources to fight against COVID-19; and (iv) ensure sound implementation of strategies related to COVID-19. Kyrgyz Republic. Operational Headquarters chaired by the Prime Minister is established to combat the spread of coronavirus infection and eliminate its consequences on the territory of the Kyrgyz Republic. Morocco. An inter-ministerial monitoring committee is coordinating the government’s social and economic response. On the health side, the Ministry of Health has set up a Technical Committee and scientific advisory body to track progress of the pandemic and develop standards and guidelines for health professionals. Republic of Korea. Establishment of the Central Disaster and Safety Countermeasures Headquarters (CDSCHQ), headed by the PM. Vietnam. Establishment of a National Steering Committee (NSC) headed by a Deputy PM. The steering committee includes Ministry of Health, Ministries of Communications, Finance, Transportation, Public Security, Industry and Trade, and Defense. See Annex 1 for more examples Function Operational coordination (e.g. oversight of action plans, mobilization, financing and interagency coordination) Forms Cameroon: Development of a comprehensive multi-sectoral response plan to COVID-19 (Plan de Riposte sanitaire, d’adaptation et de soutien socio-economique) coordinated by Ministry of Economy and Planning. Djibouti: Establishment of a COVID-19 Emergency and Solidarity Fund to ensure transparency in the collection of domestic resources and the execution of the expenditure related to the COVID-19. Kiribati: The Government has set up a taskforce with subcommittees responsible for vulnerable sectors that could be affected by the disease outbreak. Uruguay: Activation of the Coronavirus National Plan: The Plan defines a national response framework to respond to the outbreak, promotes interagency coordination to respond, maintains a flow of trusted and expedient information, and regulates actions and procedures to be adopted at every outbreak stage. See Annex 2 for more examples Function Information gathering and M&E (e.g. dashboards and monitoring routines) Forms Brazil: An “Inter-secretariat Data Committee” was established and developed a detailed Risk Matrix, a public Data Dashboard with live updates, and a comprehensive and evidence-based strategy for reopening economic activities. See Annex 3 for more examples Function External communication (e.g. press briefings, media campaigns) Forms Montenegro: A dedicated government portal with real-time data on COVID-19 outbreak, relevant news, social distancing guidelines and government measures taken, and enabling collection of donations. 24 See Annex 4 for more examples Relevant features of the country context for COG coordination choices COG institutional form and function will be strongly influenced by country con- text, which also has an effect on how easy or difficult it is to respond to a pan- demic. So far, this paper focused on both form and function of COG in the face of a pandemic. However, these institutional choices and functionalities do not emerge in a vacuum. The opportunity set for these choices is different depending on the country’s historical and institutional legacies, constitutional structure, country size, as well as the level of development and general government effec- tiveness. These in turn affect the success of the pandemic response, both through the COG channel and independently. Most would agree that is objectively easier to contain an epidemic in a small island state than in a large federal country that is home to hundreds of millions of people. At the same time, large and diverse countries will require different COG coordination mechanisms for their response to the virus than the coordination mechanisms that are appropriate for small and homogeneous countries. More complex, larger governments with more layers and institutions will also require different kinds of coordination mechanisms than simpler, smaller ones. These initial conditions influence the pandemic response not only through the ef- fect on the choice of a COG coordination mechanism, but also through other chan- nels. For example, geographically large countries tend to have more points of en- try across borders, making the spread of the virus harder to control even initially. Indeed, empirically, many have observed that large countries have a harder time to contain the virus than smaller countries, regardless of the level of development. Brazil, India, South Africa, or the US have struggled much more so than Georgia, Mauritius, New Zealand, or Slovakia. Institutionally, COGs in large countries have a more complex coordination task, battling potentially simultaneous outbreaks in different parts of the vast territory and involving large numbers of patients, and possibly across different decentralized or devolved jurisdictions. Other contextual factors, such as government credibility, legitimacy, and quality of leadership can enhance or hamper the success of CoG coordination efforts. 25 Some features of the relevant country context are sticky, while others are fluid. Sticky conditions can be given, such as country size, or can be slow to change, such as government capacity or level of development. In general, geographic and institutional factors are sticky. In contrast, fluid conditions are quick to change, including in response to the government’s handing of the pandemic. For example, trust in government may plummet if the government mishandled the pandemic response. The government can quickly lose legitimacy, politicians can be voted out of office in democracies, or top bureaucrats can lose their jobs if they fail to contain the virus. The interaction between fluid contextual factors and pandemic response is complex, as their mutual effect can become self-reinforcing: for ex- ample, trust in government can affect the success of the pandemic response, but also whether the pandemic response is successful or not can affect trust in gov- ernment. This can result in virtuous or vicious cycles, and the resulting hysteresis after the pandemic. Figure 2 depicts the direction of causation between the pan- demic response and sticky and fluid initial conditions. Figure 2. Country context and pandemic response Sticky initial conditions Level of development Internal governance Urban/rural (federal/unitary) Government capacity Inequality Population density Country size Pandemic Response Fluid initial conditions Trust in government Government legitimacy Leadership 26 Box 3 New Zealand: COG mechanisms to respond to COVID-19 with an emphasis on horizontal coordination The success to date of New Zealand’s dealing with coronavirus is largely a story of foresight, planning, and practice. New Zealand’s healthcare infrastructure has experienced similar challenges in recent years: measles outbreak (2019), H1N1 (2009), SARS (2003). Building on the revision of the 2002 Influenza Pandemic Strategy update in 2017, all relevant ministries and agencies have been planning their structured roles for such a pandemic and developing/testing communication structures between agencies. By and large, it seems that they were ready to move rapidly when the situation required it because there was a plan in place and each institution understood its role. New Zealand is responding to the COVID-19 threat through an application of its all-of-government approach to disasters and its National Influenza Pandemic Strategic Plan. Coordination on any issue is delivered by the Department of the Prime Minister and Cabinet (DPMC), the chief executive of which chairs the Officials’ Committee for Domestic and External Security Coordination (ODESC). ODESC operates like a National Security Council. ODESC reports to the Cabinet through the Cabinet External Relations and Security Committee. When ODESC identifies an issue requiring an All-of-Government approach, it establishes relevant committees and identifies a ministerial lead on the issue. 27 ODESC can also appoint an All-of-Government Controller with executive control across ministries, cognizant that the lead minister is often ill-positioned to make executive decisions related to other portfolios or to understand internal processes within other ministries. The position of All-of-Government Controller also allows for the Prime Minister and his/her office to focus on the larger machinery of government, with the Controller focused on the nuanced details of the All-of-Government issue at hand. When the COVID-19 pandemic broke out, ODESC oversaw the initial risk assessment, and appointed the Director-General of Health as the lead ministry for the pandemic response. It established a border security committee in January. Other sectoral workstreams began meeting to ensure a coordinated approach. ODESC appointed John Ombler as the All-of-Government Controller to oversee coordination between sectoral workstream pillars. Under the Influenza Pandemic National Strategic Plan, there are 10 sectoral pillars or workstreams. Each of these are led by ministry, with contributions from other relevant ministries, national agencies, local agencies and – in some cases – private sector actors and non-governmental organizations. Coordination of the response was conducted through the National Crisis Management Centre (NCMC), which is responsible for workstream coordination, information sharing, planning and resource allocation. Headed by an All of Government Controller, the NCMC leadership team included the Director-General of Health, the head of National Strategy and Policy, the Director of Civil Defense Emergency Management (CDEM), and Strategic Operations Oversight (former police commissioner). The CDEM set up an Operational Command Center (OCC), to provide oversight and day-to-day management of the response. The OCC oversees the coordination of COVID-19 response pillars, specific sectoral task forces, and regional and local CDEMs and District Health Boards. Such coordinated response has enabled the country to address the COVID-19 threat fairly quickly. Businesses and citizens have been provided some important protections. Within a month, New Zealand moved from an initial steep rise in cases to near elimination and has been able to return to more normal economic behavior sooner than other countries. 28 Source: Brookings Doha Center 2020a: New Zealand COVID Institutional Responses Sticky initial conditions Most of the initial conditions that determine the relevant country context for the type of the required COG coordination are sticky: either given or very slow to change. These include government capacity, which closely correlates with the lev- el of development as well as with the contested role specificity of the government machinery. Country size, both in area and population, is another given condition that is somewhat correlated with unitary/federal constitutional structures, as fur- ther discussed below. Government Capacity Low-income countries with lower government capacity can be very successful in battling a public health crisis when they rely on simple COG coordination mech- anisms. A good example is Liberia during the peak of Ebola crisis, where the government instituted a simple but highly effective weekly meeting on Monday mornings among government agencies, donors, and NGOs to coordinate its re- sponse.  The participants would meet to resolve any outstanding issues for the week, and most relevant decisions were taken in the context of that meeting. This simple coordination mechanism proved fit-for-purpose not despite, but perhaps because of being “low-tech.” An institutionally heavier mechanism could have proven counter-productive, as it would detract attention and scarce capacity from the substance of the matter and toward the process. The needs of more developed countries may be best addressed by a more institu- tionalized center of command, such as a structured task force with appropriate hierarchy of subcommittees. A country with higher institutional capacity may not only be able to handle a more complex coordination mechanism, but it may also require a more complex mechanism that will span its more developed institutional ecosystem within and possibly outside government. 29 Country Size and Systems of Internal Governance Size is somewhat correlated with federal or unitary system of internal governance, as large countries are more likely to be federal. Most of the world’s 200+ countries are unitary; only about 25 are true federations. However, of the world’s ten most populous countries, only three are unitary (Ban- gladesh, China, and Indonesia, even though they have some similarities with fed- erations) and seven are federations (Brazil, India, Mexico, Nigeria, Pakistan, Rus- sia, and USA). The correlation is even stronger among 10 largest countries by area: only two are unitary (China and Kazakhstan) and eight are federal (Argentina, Australia, Brazil, Canada, India, Russia, Sudan, and USA). However, it is worth noting that that federations are not always large, nor are the unitary states always small. For example, Vietnam and Myanmar are large unitary states with devolved responsibilities; on the other hand, Belgium and St Kitts and Nevis are examples of small federations. Technically, there are COGs in every state of a federation, with the state’s governor being the chief executive with their own COG. It is safe to assume that regardless of the system of internal governance, the larger the country, the more complex the coordination role that its COG will have to play. Both large and small countries require coordination horizontally across MDAs and vertically across geographic jurisdictions. However, the larger the country, the more complex the vertical coordination becomes, regardless of whether subna- tional units are decentralized, devolved, or deconcentrated. The nature of the pan- demic is such that large country size poses a unique challenge. Large population means potentially millions of infections, and consequently a large number of hos- pitals and health workers who need to be supplied with equipment and medicines. On the other hand, large land mass implies that there could be numerous simul- taneous outbreaks spread across the country, requiring a coordinated response. Another complicating factor is that as the crisis unfolds, the pandemic hits in ev- ery location at a different timing and intensity; the larger the country, the more challenging the ability to respond. This characteristic calls for a more complex and combined approach of centralized and localized responses at the same time, with pressures coming from every location in different ways. Some regions, lo- calities or communities could be going through intense hardship demanding ex- 30 traordinary support from CoG, while others may not be hit as hard and may be claiming the right to open up and be less exposed to restrictions. The situation changes over time, making it very difficult for central agencies to manage their response, and making the vertical coordination effort much more complex as well. This effect is more pronounced as the size of the country increases. Although both vertical and horizontal coordination is required in all settings, larger countries are likely to focus primarily on coordinating across jurisdictions, while smaller countries will emphasize coordinating across MDAs. To illustrate why, consider Australia and New Zealand, two countries that have both seen early successes in containing the pandemic. Australia is a geographically large fed- eration, with service delivery overwhelmingly in the hands of the states, while New Zealand a relatively small unitary (island) state. The two countries adopted very different COG mechanisms for their pandemic response. New Zealand focused on horizontal coordination across MDAs, while Australia brought in national and state officials together in a hybrid “national cabinet” that emphasized vertical coordination. Boxes 3 and 4 flesh out the details of these two cases. Photo by Ashkan Forouzani on Unsplash 31 Box 4 Australia: COG mechanisms to respond to COVID-19 with an emphasis on vertical coordination A National Cabinet, likened to the unique War Cabinet established in 1939 to lead Australia’s WWII war effort, was created to coordinate responses on the federal and state level. The Cabinet consists of the premiers and chief ministers of the Australian states and territories. The same group of officials have been the primary members of the Council of Australian Governments (COAG), which meets on a quarterly basis to manage governmental relations within Australia’s federal system. The National Cabinet currently meets on a weekly basis to assess the pandemic and the need for further or eased restrictions, chaired by the Prime Minister. The Australian governments were so pleased with the success of the co-ordination through the National Cabinet that the Prime Minister announced in May 2020 that they scrapped COAG and replaced it with a permanent National 32 Cabinet that would meet monthly. Within the Department of the Prime Minister and Cabinet, Deputy Secretary of Social Policy chairs the COVID-19 taskforce. Sub-committees include COVID-19 Health Preparedness and Response, COVID-19 Planning, COVID-19 Intergovernmental Relations, Disaster Preparedness and Response, and COVID-19 Data. Coordination concentrated in a few key intergovernmental agencies has provided a swift response across the country when dealing with COVID-19. While state governments retain significant decision rights to make their own determinations with respect to lockdown arrangements, in general there has been a high degree of coordination between federal and state governments; and among the state governments aided by strong coordination mechanisms ‘from the top’ through the instituting of the National Cabinet of political leaders. The creation of a National Cabinet of leaders meeting weekly and speaking on an almost daily basis has been seen as a cornerstone of Australia’s response and as an effective crisis management innovation. While leaders have not always agreed, it has ensured a high level of communication and a generally strong level of alignment on policies and actions. The resilience of the Australian approach is now being put to the test, with a large increase in cases in Melbourne, and some in Sydney. The States are introducing differentiated quarantining rules across the country depending on whether the visitor has been in Sydney (or some other designated areas in New South wales), in Victoria or elsewhere. Source: Brookings Doha Center 2020a: Australia COVID Institutional Responses 33 Bringing it all together: Different COG coordination options for different country contexts Based on the analysis above, we can offer advice to countries on how to tailor their COG coordination mechanisms to their specific initial conditions. Of course, there are many possible additional types of country characteristics that will shape unique challenges with the pandemic and hence affect what coordination mecha- nisms are required, some of which are discussed in the subsequent sections. How- ever, the level of government capacity (correlated with the level of development) and country size (weakly correlated with the federal/unitary form of government) seem to be the most fundamental “sticky” initial conditions. Classifying countries along two dimensions — from low capacity to high capacity and from small to large size (in both area and/or population) — yields four quadrants with different prototypes of COG coordination mechanisms suited for each. Table 4 depicts this stylized analysis. Photo by Julian Wan on Unsplash 34 Table 4. A simple typology of COG coordination types based on country context Low Gov Capacity High Gov Capacity Emphasis on horizontal coordination with Emphasis on horizontal coordination Small countries simple mechanisms (e.g. New Zealand, Slovakia, Uruguay) (e.g. Cambodia, CAR, Kyrgyz Republic, Lao PDR) A CoC close to head of government, with subcommittees Basic CoC close to head of government as needed Simple operational coordination Institutionalized operational coordination (e.g., weekly meeting of key ministers for (e.g., sector-specific working groups; oversight of pandemic response) response plans) Key data collection and spreadsheet monitoring Detailed data collection and dashboard monitoring Regular press briefings; outreach to citizens Regular press briefings; media campaigns (e.g., radio and SMS) Emphasis on vertical coordination with Likely to involve complex vertical and horizontal simple mechanisms coordination, requiring matrix mechanisms (e.g. Australia, Brazil, US) (e.g., India, Myanmar, Pakistan) A CoC close to head of government, with subcommittees Basic CoC close to head of government as needed Operational coordination involving national Institutionalized operational coordination bringing and local leaders (e.g., weekly conference call); together national and local leaders (e.g., “National basic horizontal coordination across MDAs Cabinet”; weekly call between president and governors; (e.g. weekly ministerial meeting) in addition, MDA coordinating body that involves both Large countries Key data collection and spreadsheet/ national institutions and their devolved, deconcentrated, dashboard monitoring or decentralized local organs) Regular press briefings; outreach to citizens Detailed data collection and dashboard monitoring (e.g., radio and SMS) Regular press briefings; media campaigns As is clear from the table, successful COGs will seek to fulfill their four basic func- tions discussed in the previous section regardless of the country context. Howev- er, the institutional forms will adapt across different contexts. Small low-capacity states can be successful in tackling the pandemic with simple, low-tech coor- dination mechanisms focused on horizontal coordination across MDAs. In large low-capacity countries, the “less is more” principle also holds, but the focus of operational coordination will shift to bringing together national and local leaders to coordinate response across the country, while maintaining a simple mechanism of coordinating across MDAs. In high-capacity settings, more institutionalized coordination mechanisms may be required to bring together a larger, more so- phisticated government apparatus in charge of the pandemic response. The differ- ence will be again in emphasis: horizontal coordination will feature prominently 35 in small countries, while the focus will shift to vertical coordination in large high capacity countries, where the coordination forms will be most complex. Of course, any specific country-level advice by operational teams would need to take into account the specifics of the country context, while this typology points in a par- ticular direction where the country context specifics are expected to be relevant. Other sticky conditions Aside from government capacity and country size, there are many other possible sticky conditions that may affect the type of coordination required in a pandemic, including population density, the share of urban vs rural population, or inequali- ty. For example, population density and urbanization affects how difficult it is to fight the pandemic as virus spreads more easily in densely populated cities. This poses unique challenges for densely populated small city states, such as Singapore or Hong Kong, requiring particularly fast response across multiple MDAs. This suggests the need for seamless and particularly efficient horizontal coordination in densely populated cities. Inequality is another compounding factor in such cas- es, especially the migrant workers who tend to live in cramped conditions where the virus spreads more easily. This may require COG coordination with a particular focus on working with such populations, e.g. coordinating not only across MDAs, but also including NGOs. Command-and-control versus contested form of government is another contex- tual condition that affects the type of the CoG response. One could argue that in high-capacity command-and-control systems, such as China, coordination is less important because CoG can simply order lockdowns in subnational jurisdictions and impose discipline. However, this is not borne out in the early available data. In a recent research paper, Frey, Chen, and Presidente (2020) find that although au- tocratic regimes imposed more stringent lockdowns, there is no evidence that au- tocratic governments were more effective in reducing travel. The paper also finds that “while countries with democratically accountable governments introduced less stringent lockdowns, they were approximately 20 percent more effective in reducing geographic mobility at the same level of policy stringency.” In addition, countries with more collectivist traits in their culture were able to reduce travel more relative to more individualistic countries. Thus, democratic and collectivist 36 countries seem to respond best during the pandemic.4 This, in turn, gives even more importance to different CoG institutions in democracies that coordinate such response. When western democracies such as Italy and Spain were able to impose long-lasting and generally well-observed stringent lockdowns to stem the spread of the virus, it required resolute action and coordination from CoG. Arguably, government credibility and ubiquitous communication contributed to the cooper- ation among citizens. These are features associated with fluid initial conditions, to which we turn next. Fluid initial conditions Some initial conditions that are relevant for the success of COG coordination can change in response to how the government handles the spread of the virus. These include government’s credibility and legitimacy, which are related to trust in gov- ernment. While initial credibility and trust in government can help fight the virus, policy and institutional responses to the virus can also influence how much trust and legitimacy the government accrues (or loses) during the pandemic. Leadership is another condition that has profound effect on the success of the chosen CoG coor- dination mechanism and can change in the medium term depending on the success of the pandemic response. This is especially true in democracies, where elections are likely to be seen as referenda on the incumbent’s handling of the pandemic. CoG coordination only contributes to successful pandemic response insofar the so- ciety at large cooperates. CoG can put out a consistent and coordinated communi- cations campaign about physical distancing, mask wearing, and hand washing, but these actions have to take place at the individual level among citizens. The success in stopping the spread of the virus depends on the actions of society as a whole, not only the government. It has been argued elsewhere that trust in government may help countries to mount a successful response because it helps overcome the collective action problem within societies (Davenport, Kallaur and Kunicova 2020). However, the COVID-19 crisis may also be an opportunity to build trust in low- trust environments. Like any major catastrophe or war, the pandemic may initially 37 4  See https://bit.ly/2SVrCbq produce a “rally-around-the flag” effect, when the trust in government spikes.5 Khemani (2020) argues governments can take this “windfall of legitimacy” and either seize it to facilitate a shift toward authoritarianism, or take advantage of it to build trust and social cohesion more sustainably. The latter would require using communication of credible, nonideological, and nonpolitical knowledge strategi- cally to shift norms. On the other hand, if the pandemic response falters, trust in government may sharply decline and this effect may persist.6 Moreover, COG coordination mechanisms are only useful if political leaders take advantage of them. Political leaders can take the data and advice they receive and respond appropriately, or they can ignore it, or even go against it. A well- tuned and functioning CoG policy process and institutions are only as strong as the leaders who steer them and civil servants who inhabit them. Some observed that countries with populist leaders seem to have fared particularly badly with the virus. Anderson (2020) argues that populists tend to construe themselves as representing the popular will, which does not allow for diversity of opinion or checks on authority, and leads to inability to take on new or contradictory information.7 In other words, contestability and diversity of opinion that are a hallmark of a good CoG policy coordination mechanism do not work with such leadership style. Of course, in democracies, elections are precisely what make the quality of leadership “fluid,” and presumably punish poor leadership at the times of COVID-19 at the ballot box. At the other end of the spectrum, countries with female leaders seem to have done better in containing the virus. Kristof (2020) argues that the leadership qualities that allowed female leaders reach the top – notably, effective communication in low-key way and inclusiveness – may be the same as those required to bring the country together. Yet another pos- sibility is that the electorate that chooses a female president or prime minister may be more inclined to listen to epidemiologists.8 On balance, some initial conditions can have a stronger effect on the success of 5  New York Times. 2020. “Coronavirus Has Lifted Leaders Everywhere. Don’t Expect That to Last.” April 15. https://www. nytimes.com/2020/04/15/world/europe/coronavirus-presidents.html 6  Washington Post. 2020. “Coronavirus will undermine trust in government, ‘scarring body and mind’ for decades, research finds.” July 5, 2020. https://www.washingtonpost.com/business/2020/07/05/coronavirus-pandemic-trust-government/ 7  https://www.newyorker.com/news/news-desk/populists-inflame-the-coronavirus-outbreak-across-latin-america 38 8  https://www.nytimes.com/2020/06/13/opinion/sunday/women-leaders-coronavirus.html the pandemic response than CoG coordination; however, well-chosen CoG mech- anisms can in turn improve some of these conditions. In those countries where trust and legitimacy of the government are low, it is particularly important to adopt CoG policy and institutional responses that emphasize transparency and communication, thus increasing credibility of the government. At the same time, even institutionally well-functioning coordination mechanisms will be of little use if the leadership fails to take advantage of them; yet in democracies, elections can result in a change in leadership in the aftermath of the poor COVID-19 response. 39 Photo by Kay Lau on Unsplash Conclusions This paper explored the lessons from the CoG literature and accumulated World Bank experience for coordination mechanisms required during the pandemic re- sponse. It derived the essential functions that CoGs perform in a pandemic and mapped them into various institutional forms observed around the world. Because our advice to governments about which coordination mechanism to adopt must be fit for context, the paper then discussed various types of contextual factors that affect the choice of the appropriate coordination mechanisms. It isolated two fundamental “sticky” initial conditions – government capacity and country size – that determine the emphasis of the coordination functions, as well as a relative complexity of their form. Finally, it discussed additional “fluid” factors that may further influence both the choice of a CoG mechanism and success in fighting the 40 spread of the virus, including trust in government and the quality of leadership. The main takeaways include pointers to governments on how to design CoG coordi- nation mechanisms based on the country context. The larger the country, the higher the relative importance of vertical coordination vis a vis horizontal coordination. For governments with lower capacity, the simpler the mechanism the better. Thus, small countries with low government capacity would do best with simple CoG coor- dination mechanisms focusing on horizontal coordination, while large low-capac- ity countries can be advised to focus on vertical coordination. Small high-capacity countries are well-advised to adopt more complex mechanisms emphasizing hor- izontal coordination, while larger high-capacity countries will be best served by a matrix structure. Other context variables, such as urbanization or inequality, will further affect the type of coordination required. Finally, various additional fluid factors, such as trust in government and quality of leadership, will have an impact on how CoG coordination will work and to what extent it can succeed. The paper also offers possible testable hypotheses for future inquiry about how CoG coordination affects success in fighting the spread of the virus. Because the pandemic is far from over worldwide, it is too early to declare victories.  Some fail- ures are readily apparent, though by no means can they be ascribed to the failures of CoG institutions alone.  At the same time, if the effort is to explain the degree of success or failure, then it will have to wait until a mature dependent variable can be constructed in the aftermath of the COVID-19 pandemic.  This paper suggests a hypothesis that could be tested at that time: The type and quality of CoG coordi- nation matters for how successful country is in battling this virus, but only after a number of other intervening variables and complex causal mechanisms are taken into account.  Those include leadership, legitimacy, and trust in government. The paper also provides a conceptual framework to guide operational teams work- ing with governments on the ground. An operationalization of this framework will be developed in a companion note to empower the teams with specific tools to provide just-in-time advice to clients grappling with institutional choices for CoG coordination during a pandemic. 41 ANNEX 1 Establishing Center of Command (CoC) Policy Action Description Bolivia A crisis management team at the President’s office has been created included LCR representatives with decision making power from each government agency to facilitate an immediate response to the pandemic. Cambodia National Response Committee (NRC), chaired by the Prime Minister. The NRC EAP is responsible for: (a) Identifying national policy and strategy in response to COVID-19; (b) Leading the implementation plan to prevent, protect and control COVID-19; (c) Minimizing political and socio-economic impacts of COVID-19; and (d) Leading and facilitating the implementation of multi-sectoral and inter- ministerial measures at national and sub-national levels. Central African A crisis committee led by the President has been established to: (i) ensure Republic coordination of COVID-19 response actions throughout the country; (ii) AFR approve the preparedness and response plan to COVID-19; (iii) mobilize financial resources to fight against COVID-19; and (iv) ensure sound implementation of strategies related to COVID-19. Djibouti Set up a high-level Committee to coordinate the Government’s response to the MNA COVID-19 pandemic Gabon An interdepartmental committee has been set up to supervise the government AFR response plan. Grenada The Government of Grenada’s COVID-19 Economic Support Secretariat, (CESS) LCR has officially been established to implement the economic stimulus package announced on 20 March 2020. Korea, Republic of Establishment of the Central Disaster and Safety Countermeasures Headquarters EAP (CDSCHQ), headed by the Prime Minister. Kyrgyz Republic Operational Headquarters chaired by the Prime Minister is established to combat ECA the spread of coronavirus infection and eliminate its consequences on the territory of the Kyrgyz Republic: Lao People’s Establishment of a Task Force Committee led by the Deputy Prime-Minister and Democratic Republic Minister for Finance. The Committee serves as focal point for coordination with EAP all concerned agencies in leading, monitoring and evaluating the outcome of the implementation of the prevention, protection and control measures of the outbreak, as well as the treatment provided to infected persons and preparedness in all aspects that ensure a quick, sufficient and highly effective response. 42 Policy Action Description Morocco An inter-ministerial monitoring committee is coordinating the government’s MNA social and economic response. On the health side, the Ministry of Health has set up a Technical Committee and scientific advisory body to track progress of the pandemic and develop standards and guidelines for health professionals. Papua New Guinea Establishment of the COVID-19 National Operations Centre (CNOC), with a multi EAP sector agency task forces. Peru Central Government created a High-Level Multisector Commission to follow up on LCR the most relevant measurements that should be taken to control the COVID-19. Philippines Establishment of the Inter-agency Task Force for the Management of Emerging EAP Infectious Diseases (IATF-EID) which serves as the policy-making body for the National Action Plan for COVID-19 (NAP). The NAP is designed to contain and mitigate the spread of the virus by coordinating response actions. Saint Lucia Activated the National Emergency Management Advisory Committee (NEMAC) in LCR February, the implementing agency in times of national disaster Samoa Activation on March 21, 2020 of the National Emergency Operations Centre EAP (NEOC), as a focal point providing collaboration and operational coordination amongst all stakeholders, with advice and technical guidance from the Ministry of Health, following the declaration of a state of emergency. The activation of NEOC is part of Samoa’s National Emergency Response Plan as per the National Disaster Management Plan under the National Disaster and Emergency Act 2007. Somalia Establishment of a national coordination committee to respond to COVID-19 AFR Tanzania Government has established a National Task Force to coordinate the response AFR against COVID-19 headed by the Prime Minister. Two other committees established for the response include the committee of permanent secretaries chaired by the Chief Secretary and a committee of technical specialists. Vietnam Establishment of a National Steering Committee (NSC) headed by a Deputy EAP Prime Minister. The steering committee includes Ministry of Health, Ministries of Communications, Finance, Transportation, Public Security, Industry and Trade, and Defense. Zimbabwe Establish central coordination and data aggregation mechanisms to drive AFR COVID-19 response Congo, Republic of Establishment of a National Committee in charge of defining, coordinating and AFR implementing the policies and measures/ Dominican Republic A high-level commission led by health experts and with representation of key line LCR ministries has been set-up to coordinate response to COVID-19 43 Source: World Bank 2020. COVID-19 Response Tracking Portal (RTP) ANNEX 2 Systemic Approach: Planning, Financing, and Inter-Agency Coordination INTER-AGENCY COORDINATION Policy Action Description Kiribati The Government has set up a taskforce with subcommittees responsible for vulnerable EAP sectors that could be affected by the disease outbreak. Tuvalu A COVID-19 Health Taskforce was convened in late January. EAP Bhutan District Rapid Response Team for respective districts (Dzongkhags) formed SAR Cambodia Establishment of response committees in international border check points . It is led by the EAP border check point office. Cameroon Establishment of a consultation framework between the Ministry of Finance and the Ministry AFR of Economy and Planning, with the main economic players, in order to mitigate the effects of the crisis and encourage a rapid resumption of activity. Uruguay The Departmental Directions of the Health Department and the Departmental Centers for LCR Coordination of Emergency (Cecoed) are coordinating actions at the local level, for instance, monitoring people quarantines after coming back from at risk countries such as Italy or Spain. Bolivia The government has created a National Council for Economic Reactivation and Employment LCR tasked with coordinating and implementing a national employment program to address the effects of the COVID-19 pandemic. The Council is chaired by the Ministry of Productive Development. 44 FUNDRAISING, BUDGETING, AND EXPENDITURE OVERSIGHT Policy Action Description Benin Under the MEF instructions, a focus group has been set up to ensure speedy disbursement, AFR proper management of resources and quality of expenditures that will be carried out Kyrgyz Expert working groups are established to attract external aid to support: i) the budget and Republic restructuring of the state external debt; ii) the health care system in emergency situations ECA and to provide social assistance to citizens; iii) entrepreneurial activity, restoration and development of economic activity; iv) to ensure food security. Angola To reduce costs and enhance government efficiency, on April 06th, the government apparatus AFR has been reduced from 28 to 21 ministries and from 24 to 18 secretaries of State Djibouti Establishment of a COVID-19 Emergency and Solidarity Fund to ensure transparency in the MNA collection of domestic resources and the execution of the expenditure related to the COVID-19. India India leads in establishing a COVID-19 Emergency Fund for South Asia SAR Kyrgyz Establishment of two Commissions on the distribution of donations from i) individuals to fight Republic coronavirus infection and ii) individuals and legal entities to provide food for low-income families, ECA which include representatives of state bodies, members of parliament and the civil society. Papua New The Emergency (General Provisions) (COVID-19) Act 2020 establishes the functions of the Guinea Emergency Controller. The Commissioner of the Royal Papua New Guinea Constabulary was EAP appointed to be the Emergency Controller on March 26, 2020. Peru Nearly US$30m to the Health Ministry to reinforce the prevention, control and response LCR mechanisms. 45 PLANNING Policy Action Description Angola Contingency Plan AFR Cameroon Development of a more comprehensive multi-sectoral response plan to COVID-19 (Plan de AFR Riposte sanitaire, d’adaptation et de soutien socio-economique) coordinated by Ministry of Economy and Planning. Cameroon Government has developed a health sector-led emergency preparation and response plan to AFR COVID-19 budgeted at ~ US$10 million for 6 months of implementation. Central The government has prepared a Global Preparedness and Response Plan Against Coronavirus African in CAR, with an estimated cost of 27 billion of FCFA (USD 44 million) Republic AFR Congo, On Coordination: Response Plan estimated at US$170 million (100 billion XAF)/ Republic of AFR Kazakhstan Special governmental commission to prepare action plan to protection population ECA Uruguay Activation of the Coronavirus National Plan: this Plan defines a national response framework LCR to respond to the outbreak, promote interagency coordination to respond, maintain a flow of trusted and expedient information, and regulate actions and procedures to be adopted at every outbreak stage. Source: World Bank 2020. COVID-19 Response Tracking Portal (RTP) 46 ANNEX 3 Evidence-based response: Data and M&E tools Policy Action Description Brazil Establish an “Inter-secretariat Data Committee” that has developed a detailed Risk Matrix, a LCR public Data Dashboard with live updates, and a comprehensive and evidence-based strategy for reopening economic activities Uruguay Per National Law N° 19.355 (2016), SINAE is heading information management for the LCR Presidency Zimbabwe Establish data aggregation mechanisms to drive COVID-19 response AFR Source: World Bank 2020. COVID-19 Response Tracking Portal (RTP) ANNEX 4 Communication tools Policy Action Description Kazakhstan Single COVID-19 related information resource is established that publishes information on ECA a single call-center number; the number of registered cases, recovered patients and deaths by region; prevention measures; news update; official announcements and decisions; an interactive GIS map with registered cases and their contacts; frequently asked questions. Montenegro A dedicated government portal established providing real time data on COVID-19 outbreak, ECA relevant news and social distancing guidelines and government measures taken, and enabling collection of donations. Uruguay Per National Law N° 19.355 (2016), SINAE is heading public communications for the LCR Presidency and on behalf of Health Ministry as part of the Coronavirus National Plan, issuing regular situation updates on cases etc. Vietnam New online information and services for workers and employers impacted by COVID-19 EAP Source: World Bank 2020. COVID-19 Response Tracking Portal (RTP) 47