Political Will for Anti-Corruption Reform Communicative pathways to collective action in Ukraine 20 23 Working paper series 2023:3 Oleksandra Keudel Marcia Grimes Oksana Huss Political Will for Anti-Corruption Reform Communicative pathways to collective action in Ukraine Oleksandra Keudel Marcia Grimes Oksana Huss WORKING PAPER SERIES 2023:3 QoG THE QUALITY OF GOVERNMENT INSTITUTE Department of Political Science University of Gothenburg Box 711, SE 405 30 GÖTEBORG March 2023 ISSN 1653-8919 © 2023 by Oleksandra Keudel, Marcia Grimes, and Oksana Huss. All rights reserved. Political Will for Anti-Corruption Reform: Communicative pathways to collective action in Ukraine Oleksandra Keudel Marcia Grimes Oksana Huss QoG Working Paper Series 2023:3 March 2023 ISSN 1653-8919 Abstract The importance of political will to bring about institutional reform to reduce corruption is universally rec- ognized, but little is known about when and how it emerges. This paper examines six local settings which exhibit positive change in terms of anti-corruption political will. Theoretically, we develop a definition of political will suited to anti-corruption, and then, drawing on scholarly work on policy formation, norm change, and collective action, identify key steps involved in shifting and sustaining local actors’ preferences to support anti-corruption work. The analyses build on data from 70 interviews with local public authorities in six cities in Ukraine. The processes varied considerably across the six cases in terms of the actors involved and the organization of anti-corruption work. The commonalities, identified inductively, instead relate to the functions of communication. Those functions enabled local actors to establish, continually reaffirm and mutually recognize commitments to reform, operationalize those commitments into actionable plans, and continually persuade relevant stakeholders to sustain reform momentum. Oleksandra Keudel Marcia Grimes Department of Public Policy and Governance The Quality of Government Institute Kyiv School of Economics Department of Political Science okeudel@kse.org.ua University of Gothenburg Marcia.grimes@pol.gu.se Oksana Huss Department of Political and Social Sciences Bologna University oksana.huss@unibo.it 2 Acknowledgments We would like to thank the research team from the Kyiv School of Economics (KSE) and the Anti-Cor- ruption Research and Education Centre at the National University of Kyiv-Mohyla Academy (ACREC) for their contribution to the research project, especially, Dmytro Iarovyi, Kristina Khambekova, Alina Los, Ksenia Ivanyshyn, Anna Shcherbiak, and Khrystyna Holynska. We thank Anastasiya Romanyuk for quick support at the final stages of this report-writing. We are also grateful to the interview partners in Kyiv, Chernivtsi, Ivano-Frankivsk, Lviv, Poltava, Vinnytsia, Zhytomyr, and Uzhhorod for sharing their experiences and views with us. Funding This study was conducted within a research project “Opening the black box of political will: Local public authorities and anti-corruption efforts in Ukraine”, funded by the Swedish International Centre for Local Democracy in January 2020 – March 2022. 3 Introduction Failure to implement effective laws or institutions to reduce corruption is often attributed to a lack of po- litical will. Absent political will, anti-corruption bodies and legal frameworks, even once put in place, can suffer budget cuts or political interference. Studies of both positive developmental trajectories (Quah 2015; Jones 2017; Jongen 2021) and policy failure (Batory 2012) mention political will as decisive, as do strategic policy documents on the anti-corruption agenda (USAID Office for Democracy and Governance 2006; Chêne 2010). Yet, little is known about what conditions and factors might contribute to the development of anti-corruption political will. Historical analyses of political development in Europe suggest that major events such as threat of war or resounding military defeat often preceded major reform efforts to mitigate corruption (Tilly 1985; Mungiu- Pippidi 2015, 71). Contemporary success cases such as Botswana, Hong Kong, Rwanda, Singapore, like many of their historical predecessors, occurred under autocratic leaders, when political will is primarily a function of an individual ruler’s preferences (Post et a. 2010; Quah 2017). While interesting, these examples do not serve as useful sources of policy strategies (autocracy and war), and to date little research exists on the conditions and processes that might promote positive change in democratic settings. This paper examines six municipalities in Ukraine, an electoral democracy, which all exhibited positive de- velopment with respect to political will to tackle corruption between 2018 and 2021. Corruption exists in many forms. The focus here is on municipal level grand corruption, i.e. when politicians or upper-level civil servants use their positions of power for private gain. While paying bribes to access public services is im- mediately noticeable to citizens, grand corruption can have more far-reaching implications for the quality and efficiency of government services, as for example when officials in procurement processes award con- tracts to firms who offer the largest kickbacks rather than the best services. Political will to combat grand corruption is arguably the most difficult to mobilize and sustain, as the actors involved most commit to refraining from rent-seeking and policing themselves. Corruption as a policy issue is exceptional in the gap between talk and action: all officials can gain from publicly promising to take action, but bringing about change requires extensive effort by large set of actors, and therefore also trust among those actors, to have any hope of effecting meaningful change. The analyses did not reveal commonalities across the cases in terms of the main drivers of change or which actors were involved, nor how reform efforts were organized. Instead, the interviews identified processes – which we call communicative functions – that transpired among local politicians, civil servants, civil society activists and policy experts which combined to promote and sustain positive change. The analyses identified three main communicative processes related to the mobilization and deepening of political will. The first related to stating, strengthening and sustaining commitments. This required that actors a) communicated to identify and make visible the prospective benefits of mitigating corruption, b) made public declarations to back reform efforts, and, crucially, c) continuously engaged in mutual account- holding, invoking and reminding each other of commitments to work to mitigate corruption. The second set of communicative functions related to deepening and operationalizing anti-corruption political will, which entailed communication d) to map the dimensions and strains of corruption and to prioritize areas 4 most in need of attention; e) to translate abstract policy goals such as transparency, impartiality and effi- ciency into operative policy solutions; and f) to vet policy solutions among relevant stakeholders once they began to take shape. The third overarching function was a constant throughout: actors continuously engaged in persuasion, efforts to influence relevant stakeholders and especially veto-players of the necessity for pol- icy action. We begin by briefly examining political will as it relates to anti-corruption efforts, arguing that anti-corrup- tion is somewhat distinct from other policy issues in a way that places different, and arguably higher, re- quirements on the communication among key actors in order to induce political will. The paper then ex- plains the Ukrainian social and institutional context for anti-corruption reform and the selection of cases, and overviews the selected cases. Then, drawing on data from 70 interviews and research findings on norm change, policy formation, and collective action, the paper lists and illustrates the crucial communicative functions identified as crucial to developing political will. This section theorizes each function of commu- nication, and then provides empirical illustrations from the cases. Together, the communicative functions promoted and sustained key actors’ commitments to anti-corruption efforts, as well as those actors’ trust in one another’s commitments to do the same. Political will as it relates to anti-corruption efforts What is political will? Conceptualizations span from seeing political will as an individual-level phenomenon arising from disso- nance between leaders’ ideology or preferences on the one hand, and political and societal reality on the other (e.g. Blickle et al., 2018); Šmigoc, 2015), to much more encompassing definitions. Brinkerhoff (2000, 242) recognizes that a leader’s intent is unlikely by itself to predict change and therefore includes leaders’ power and capability, as well as support from key stakeholders, institutional resources and capacity, the epistemic foundations of reform proposal, development of effective sanctioning mechanisms to uphold reform efforts, and the durability of these cumulative factors. This conceptualization sees political will as a “…latent quality; it is not visible separate from some sort of action. Measuring it can only be done indirectly” (Brinkerhoff 2000, 241). Such broad conceptualizations render it difficult to examine the explanatory power of political will, or the conditions under which it is likely to emerge, as both are folded into the concept itself. Post et al. (2010) offer a more operationalizable definition: “political will is the extent of committed support among key decision makers for a particular policy solution to a particular problem” (2010, 659), and suggest veto player theory (Tsebelis 2002) as a tool to identify “key” actors.i While a useful starting point, the nature of corruption as a policy challenge requires two modifications to this conceptualization. Actors face a col- lective action dilemma when choosing to abstain from corruption in favor of rule-following behavior. De- veloping rules and enforcement mechanisms to mitigate corruption is a second order dilemma. An individ- ual political leader, unless that person is omniscient and all-powerful, cannot single-handedly develop and enforce rules to change behaviors. Developing effective rules and norms must thus be a collective effort, and the best short-term strategy for political actors is to talk the talk without “walking the walk” – i.e. to profess to work against corruption while in actuality contributing little to efforts to develop needed rules and monitoring systems, (or even continuing to engage in corruption, thus free-riding in both first and second order dilemmas). 5 Key actors thus both need to have a commitment to promote change, but, crucially, also need to trust one another’s commitments to do the same. A reform-minded political leader, however well-intentioned, may abstain from vesting scarce time and resources if other key actors are not prepared to do the same, not least as concerted efforts to develop rules, and detect and sanction corruption may entail considerable political cost if other key actors are only nominally committed. If only a handful of actors make genuine efforts to pro- mote reform, those efforts will have little or only short-lived effects, and reform promoters will incur costs in the form of effort with no returns on their investments, and potentially even retaliatory actions from others (Karklins 2005). Second, political will to reduce corruption requires not only that government actors commit to a change of rules or praxis, but non-government actors as well. Post et al. (2010) and many others note that actors with decision-making power but also those with the power to implement and enforce a reform must commit to reform efforts. Corruption, to a greater degree than other policy issues, involves interactions between gov- ernment and economic actors. Many regulatory efforts imply restrictions on business, and thus benefit from quasi-voluntary compliance, but corruption in most cases consist of actions already deemed illegal, and business actors therefore have as much power to subvert anti-corruption regulations as government offi- cials. We offer the following conceptualization, which can be observed independent of broader contextual con- ditions and actual corruption mitigation: anti-corruption political will entails that a critical mass (either in terms of numbers or positions of influence) of government, both political and administrative, and business leaders hold commitments for corruption mitigation, and they also trust that others in this set of actors also hold such commitments. We thus propose that political will can be observed in two indicators: 1) the ob- servable commitments of key actors, and 2) expressed trust among key actors in one another’s commit- ments. These mutually trusted commitments constitute a minimal, binary component of political will (Post et al., 2010). A number of additional conditions can determine whether political will is weak or strong, a continuous component. Two such conditions are whether key actors share a common understanding of a particular problem and agree upon effective policy solutions (Post et al., 2010). Agreement among key actors on a problem definition and appropriate responses represents a stronger political will to bring about change. Finally, while outside the realm of power and thus not strictly involved in enacting and enforcing policies, we theorized that the existence of civil society actors with strong anti-corruption commitments further strengthens political will. Communication and the development of political will Lessons from historical trajectories of today’s best performers yield insights which, though important, hold limited contemporary relevance to supporting and promoting the emergence of political will in contempo- rary settings. And as Mungiu-Pippidi notes, “the challenge remains in building control of corruption by democratic means, therefore by solving collective action dilemmas” (Mungiu-Pippidi 2015, 132). Absent a 6 benevolent dictator, solving collective action dilemmas is replete with pitfalls and contingencies, as the mas- sive literature on social dilemmas and common pool resources attests.1 In a review of 35 years of game theoretical research, David Sally (1995) identified conversation as the strongest predictor of individuals’ ability to commit to a collectively beneficial payoff. Compared to other factors such as players’ characteristics, the instructions of the game, game iterations, the payoff structure, player anonymity or group identity “the experimental evidence shows quite clearly that discussion has an extremely positive effect on subjects’ willingness to cooperate” (Sally 1995, 61). The “…act of communi- cating itself influences people’s preferences: Conversation is a form of social contract that may create em- pathy among participants, even those belonging to groups with different social identities” (Sally 1995, 69; Ostrom 1998). While in line with our own argument, the finding that communication is important, does not add much to saying that political will is important. What do players need to do when they communicate such that anti-corruption political will might emerge? We study communication processes through which information and knowledge are transferred and dissem- inated, actors describe their positions, and persuasion and learning occur. In all of this, actors can be more – or less – open and truthful, strategic and manipulative, or clear and skillful. By looking at cases where the outcome is known, and by asking actors to describe the processes that moved the anti-corruption agenda forward, we distill communicative functions that enable the collective outcome. Context, methodology and overview of cases Anti-corruption in 2015-2021 Ukraine: high salience, sustained public pressure, and progressive national-level reforms. Ukraine at the time studied was an instructive albeit not a typical case for examining anti-corruption political will at the local level. Corruption was among the rallying issues of the 2013-2014 Revolution of Dignity and in the years following, civil society coalitions for transparent and accountable government operations mo- bilized at both national and local levels (Bader et al. 2019; Keudel 2022, 447–48). The Revolution created a window of opportunity and numerous anti-corruption reforms to increase transparency, accountability and public participation followed (OECD 2017). First, Ukraine introduced administrative and criminal liability for all major corruption-related offenses as per UNCAC. 2 Independent anti-corruption institutions were created, covering corruption prevention (NAPC), investigation (NABU and SAP) and justice (HACC). The law on access to public information introduced far-reaching transparency regulation, obliging officials to publish their assets, and generating all 1 Game theory research tells us that communication/conversation is important, but is silent with respect to what people do when they communicate. In all likelihood, in the reduced and controlled settings of laboratory games, that conversation entails express- ing a willingness to pursue a collaborative strategy and perhaps affirming to one another the potential gains of doing so. This would only represent a highly preliminary starting point in real world collaborative efforts of the scale and complexity that policy reform entails. 2 In addition, the Law on Corruption Prevention was extended to cover local public authorities, which inter alia, required asset declarations of public officials and responsibility for false declaration. Some of them were rolled back, however, at the end of 2020, including illegal enrichment and liability for false statements in asset declaration. 7 information on public procurement – one of the most corrupt spheres of policymaking – in machine read- able open data format. This provided both law enforcement agencies and grassroots activists with legal foundations to develop instruments for counteracting corruption. Second, numerous digital tools have been created and civil society networks mobilized to make use of open data and to exercise accountability through monitoring, investigating, publishing, and reporting of corrup- tion. The Prozorro platform for transparent public procurement, developed and utilized by civil society, is internationally recognized as a leading anticorruption technology, and a country-wide community of activists uses artificial intelligence to detect corruption. Similar initiatives have arisen around asset declarations, pub- lic budget expenditures, decision-making process (e.g. Bihus.info). Finally, citizens and authorities developed and utilize numerous tools to enhance citizen engagement in policymaking: e-petitions, citizen consultations, participatory budgeting and social audit tools. Against the historic background of a massive oligarchic influence on political parties and politics, this brought a major shift in the social contract, which increased citizen agency to shape the public sphere on an everyday basis. In tandem with these anti-corruption reforms, the 2014 Decentralization reform (Cabinet of Ministers of Ukraine 2014) created fiscally independent and politically empowered municipalities, which became the main providers of public services (education, healthcare, utilities) and economic development (Roberts and Fisun 2014; Romanova and Umland 2019). Decentralization increased opportunities for corruption locally but also gave rise to new demands for accountability (Huss et al. 2020). Exercising their relative autonomy to shape anti-corruption policies and tools within the national framework, local authorities enacted trans- parency and accountability regulations, though to varying degrees (Fedynchuk et al. 2018; Khutkyy and Avramchenko 2019; Lukerya et al. 2016; TI Ukraine 2018). Data and Methods: case selection and background Measurement of political will and case selection This paper analyzes the processes involved in the emergence of anti-corruption policies and tools in 2015- 2021 in Chernivtsi, Ivano-Frankivsk, Lviv, Poltava, Vinnytsia and Zhytomyr – all regional centers (Figure 1).ii In comparison to other cities, regional centers have more resources in terms of material assets and human capital (Dudley 2019). The six selected cases exhibited a positive change in political will for anti- corruption. Our measure of political will includes two components: 1) commitment by key actors to legal and institutional reforms limiting their power (Brinkerhoff 2000, 243) and 2) trust of one set of relevant stakeholders (civil society) that others’ commitments to anti-corruption are genuine. 8 FIGURE 1. UKRAINIAN MUNICIPALITIES AND CASE STUDY CITIES Zhytomyr Poltava Lviv Vinnytsia Ivano- Frankivsk Chernivtsi We measure the first component as progress on at least one index of local authorities’ transparency: TI Transparent Cities Indexiii or Local Government Publicity Index.iv These indices reflect the existence of a register of municipal property, open data on places for municipal advertising, or having supervisory boards for municipal enterprises, all of which limit discretion among local authorities when managing resources. The second component – trust among key actors that anti-corruption commitments are genuine – is cap- tured through the assessments of anti-corruption activists. The measure builds on two original surveys of CSOs: the first face-to-face with 242 CSOs in 2018 (Huss et al. 2020) and the second conducted online in March-April 2021 of 172, with some organizations included in both rounds. Both surveys asked CSO ac- tivists to assess anti-corruption political will in their local government, and whether local anti-corruption efforts were moving “in the right direction”, allowing us to detect positive change.3 We selected six cases with indications of comparatively high political will or positive development over time. We include all cases that exhibited higher than average political will in 2021 or positive change between 2018 and 2021; cities where activist assessments in 2021 suggested that political will had diminished over time were then excluded second (Figure 2). Then we used progress on at least one index of local authorities’ transparency as further confirmation of some level of political will. The selected cities are thus deviant (frontrunner) cases as they show positive change compared to other cities. It is important to note that civil society assessments of LPAs’ political will were not overwhelmingly strong, however, and that LPA efforts were not made in other cities as well. These six cases show the most positive change, and our analyses seek to map common elements in how it arose. The analyses do not aim to test hypothesized explanations against each other, which would require comparing successful and less successful cases. Our research design instead aims at developing a theory of change by looking at similarities in these cases (Seawright and Gerring 2008, 302ff). 3 The ideal measure would reflect the attitudes of all key actor groups about all other key actor groups (i.e. politicians, civil servants, businesspeople, civil society), but at present, such data is unavailable. 9 FIGURE 2. POSITIVE CHANGE IN POLITICAL WILL BASED ON CSO SURVEYS 2,50 7,00 Selected 2,00 cases 6,00 1,50 5,00 1,00 4,00 0,50 3,00 0,00 2,00 -0,50 -1,00 1,00 -1,50 0,00 PW 2021 PW Change vs 2018 AC Direction 2020-2021 Note: Left y-axis PW2021 (blue bars) show city means of activists' survey responses (n=130) to question To what extent is there political will among the local authorities in your city to fight corruption? Response scale ranged from 1=absence of political will to 11=very high political will, rescaled 1=low (1-3), 2=medium (4-6), 3=high (7-11). PW2018 city mean of activist (n=191) responses to identical question, response scale 1=absence of political will, 3=very high political will. Yellow bars show change between 2018 and 2021 where positive values indicate positive development of political will. AC Direction 2020-2021 (red line) shows values on right y-axis of 2021 city mean of activists' survey responses to question Please recall the anti-corruption actions of the LPA in your city over past 12 months, do you think they were going in the right direction? 1=Not at all, 6=Yes, to some extent, 11=Absolutely. After identifying cases, wev conducted 71 semi-structured online interviews in the six selected cities (be- tween 7 and 10 respondents per city) between May and August 2021. Interviewees included representatives of local authorities (deputy mayors, secretaries and city council members and executive public officers from different departments), members of relevant city council working groups, and representatives of local CSOs and the business sector. Most interviewees were involved in anti-corruption efforts and were well-ac- quainted with the processes. They were identified in official statements, protocols of council sessions, media reporting, and referrals from other deputies and executives. Interview guides were adapted to the specific interlocutors. The guides consisted of 5 themes: definition of corruption and its causes, interlocutors’ and LPAs’ anti-corruption activities (rationale, achievements and pitfalls, conflict lines around their adoption), communication process around anti-corruption initiatives (in- volved actors, organizational forms of communication, content), subjective assessment of the effectiveness of an initiative, resources used (human, material, and international technical assistance), and outlook for the future. Before each interview, we adjusted the guide based on our desk research of a person’s involvement in anti-corruption activities in their city. Most interviews were recorded. We anonymized the responses for interlocutors’ security in the face of the Russian full-scale invasion since reform-minded LPA members and activists became targets of the occupiers. 10 Overview of cases: Local conditions and anti-corruption activities Local conditions and general relevance of findings Before presenting the results, a brief consideration of the scope conditions of our findings is in order. The six cases are typical of regional centers in terms of size with ca. 300,000 residents (except Lviv with ca. 700,000 residents) and are medium-developed cities, with Chernivtsi being the lowest-performing and Pol- tava the highest-performing. The cities shared some conditions considered favorable for anti-corruption political will. Selected cities were all fairly pluralist in the sense of competing economic and political elites, which is a conducive environment for anti-corruption activism (Huss et al. 2020), transparency (Mazepus et al. 2020), and participatory reforms (Keudel 2022). Politically, some level of competition existed in all six cities: no party dominated the councils and mayors face medium to high competition during elections. In Chernivtsi and Poltava, mayors lack party support as opposition parties control the council.4 In terms of economic conditions, all build on trade and services, light industry (textile or automobile assembly), and none are dominated by a single industry; Lviv stands out for its IT-industry. Demand for anti-corruption policies – a possible explanation of governance reforms – differed among the cities. Citizens perceiving corruption as a serious issue varied between 40% in Chernivtsi and 72% in Poltava (Rating Sociological Group 2016, 8); participatory demands also vary between 6% in Chernivtsi and 21% in Vinnytsia respectively, the lowest and the highest values for Ukraine (Rating Sociological Group 2016, 105– 7). Moreover, parties with an anti-corruption agenda are represented in some councils but not others: while in Ivano-Frankivsk and Zhytomyr ca. 2/3 of seats are taken by parties with an anti-corruption agenda, in the rest of the cases this is slightly over 40% (for clustering of Ukrainian parties into anti-corruption parties, we follow: Onopriychuk 2017, 17). Our cases are situated both in regions with Habsburg and Russian im- perial heritage, belonging to the Western and Central macro-regions. Geography and history do not, in other words, systematically explain political will for governance reform in post-Maidan Ukraine.5 In sum, none of the cases have highly unfavorable conditions such as a closed-order oligarchy, and the lessons gleaned from these six cities are unlikely to be relevant for or replicable in such contexts. The aim is to describe how political will can emerge and not how it has emerged elsewhere. The findings are thus not generalizable but are nonetheless potentially relevant for other contexts. Even with auspicious precondi- tions, the emergence of anti-corruption political will is by no means a certainty. The communicative pro- cesses described below may have transformative potential in various types of settings in other words, though perhaps not under the most adverse conditions. 4 Political competition in Chernivtsi and Poltava resulted in the dismissal of mayors by city council votes of no confidence in 2018. According to interlocutors, competition in Chernivtsi was conducive for more transparency (Interviewees 112, 125), while in Pol- tava the mayor’s dismissal undermined transparency initiatives (Interviewee 563). 5 Our sample does not feature cases in South-eastern Ukraine. This is the most industrialized region, with regional centres often dominated by one industry or a sea port, which we rather link to less political will for anti-corruption than geographical location or historical legacies per se (pluralist economic structures are associated with more pluralist governance, and vice versa: Mera 2016). 11 Local government and organization of anti-corruption work Procedurally, either the mayor, the executive departments, or council deputies may initiate policy change in the municipalities.vi Mayors and council deputies are directly elected every five years; the executive commit- tee is proposed by the mayor and confirmed by the council. The local self-government structure follows the “strong mayor” model: as the highest municipal authority and head of both the council and the administra- tion, the mayor sets the council’s agenda and organizes its work, voting, and chairs sessions, and can also issue executive orders (Congress of Local and Regional Authorities 2016, 28, 98). The city council checks the power of the mayor and holds the executive bodies accountable through, inter alia, rules of procedure, deputies’ appeals, and elaborating/reviewing legislative proposals in standing commissions. The council can also cancel the acts of the executive, or initiate a vote of no confidence against the mayor. With respect to anti-corruption work specifically, the organization of work and interactions between the main governing bodies and non-governmental stakeholders varied in the selected cases, suggesting that there is no single model for structuring communication that facilitates anti-corruption political will. For example, there are no patterns in the usage of media or advisory councils or other communication platforms (such as standing commissions of a council) between the cities. Working groups, however, are common in the cases, but they are also common in less successful cases. We did observe two commonalities in the organization of anti-corruption work in the cities with emerging anti-corruption political will. First, CSOs were more likely to use direct forms of communication with LPAs, e.g. roundtables, trainings or providing LPAs directly with constructive policy proposals. This points to the importance of dialogical spaces which allow for direct exchange rather than only mediated communication in the public domain. Second, in many successful anti-corruption initiatives, an individual or unit took on the role of facilitator or standard-bearer and played a key role in sustaining momentum in the communicative processes. Facilitators ensured that ideas generated at workshops were documented and kept on the agenda in subsequent meet- ings. In Lviv and Vinnytsia, LPAs created dedicated administrative units (“integrity sector” and “quality assurance unit” respectively) to coordinate and implement anti-corruption activities such as corruption risk assessments. They also coordinated the efforts of urban development agencies among other things to secure external project funding, which inter alia was used for anti-corruption activities. Chernivtsi, Ivano-Frankivsk and Zhytomyr, in contrast had anti-corruption advisors rather than special units. These advisors were often funded by an external donor and performed similar roles coordinating and sustaining LPA effort (e.g. EU- ACI). While the designated facilitators often assumed coordinating roles, some crucial initiatives were facilitated by CSOs, local deputies, or deputy mayors. For example, the adoption of public auctions was often led by CSOs (e.g. Zhytomyr, Lviv) or, instructively, the very creation of an anti-corruption administrative unit in Lviv, which materialized thanks to coordinating work of a local project manager. This echoes the finding from the citizen participation literature about an often invisible work of facilitators who connect partici- pants, disseminate information, and use specific methods to structure communication (Escobar 2019).6 6 While we do not find any evidence that government units perform better than individual consultants as facilitators, the former may better enable implementation. An external consultant reflected that communication aided in the adoption of an integrity plan, 12 Anti-corruption initiatives Anti-corruption policies introduced in these cities primarily targeted grand administrative corruption though some also sought to address petty corruption. These included for example measures to ensure full public access to drafts, decisions, and proceedings of the council (especially, Chernivtsi, Lviv), public auctions for lease or sale of municipal property (land, real estate) – often via the web-based system Prozorro.Sale (Ivano- Frankivsk, Lviv, Chernivtsi, Vinnytsia),7 geo-information systems (GIS) to visualize open data on municipal property and its lease/sale and on procured services such as street cleaning (Chernivtsi, Lviv, Vinnytsia, Zhytomyr), codes of ethics for executives and for elected deputies (Vinnytsia, Ivano-Frankivsk, Lviv), ded- icated anti-corruption strategies and integrity plans (all). Open data has been introduced in all our cases, albeit with different depth: e.g. while in IT-savvy Lviv open data is regularly updated thanks to the dedicated municipal institution, in, for example, Chernivtsi this has been challenging due to a lack of human resources. Communicative functions: Identification and illustrations In rationalist accounts of collective action, the function of communication is to nurture trust among the actors and to form an implicit or explicit contract about future behavior. In what follows, we identify func- tions both consistent with the rationalist account but also functions that go beyond it, and which resonate with work on norm change, policy formation and collective action. Three overarching communicative functions emerged as crucial to the development of political will (Figure 3). The first relates to the binary component of political will, i.e. developing and sustaining key actors’ commitments, as well as recognition among key actors of others’ commitments. Establishing mutually rec- ognized commitments consisted of three more specific functions: 1) identify and make visible the prospec- tive benefits of collaborative behavior, in this case appropriate and effective rules and regulations to mitigate corruption; 2) make public statements committing to support action to address the issue at hand; and 3) hold each other to account to honor those commitments. The second overarching component relates to the continuous aspect of political will, i.e. making commit- ments operative. Here also three more specific functions were identified. Communication served 1) to map the dimensions and strains of corruption and to prioritize areas most in need of attention; 2) to translate abstract policy goals such as transparency, impartiality and efficiency into operative policy solutions; and 3) to vet policy solutions among relevant stakeholders once they began to take shape. The third overarching function relates to actors’ continuous efforts to persuade relevant stakeholders and especially veto-players of the necessity for policy action. Together, these functions of communication facilitate stakeholder coor- dination and continually reaffirm commitments, helping actors overcome the mutual distrust typical in set- tings with systemic corruption. but that subsequent implementation was weak. Part of his explanation was that an external actor (he) was coordinating the de- velopment of the integrity plan and thus the LPA executives did not have the necessary ownership (715). 7 Use of municipal property is one large corruption issue, which can manifest via grand corruption forms (usually, because influ- ential businesses are represented in the council and can decide on the rental or sale for non-competitive prices by colluding with other deputies) and petty corruption forms (when a public official can be bribed for simply checking the availability of property and/or for a positive decision). 13 FIGURE 3. FUNCTIONS OF COMMUNICATION FOR POLITICAL WILL TO REDUCE CORRUPTION IN LOCAL AUTHORITIES Communication to induce initial commitments Make payoffs visible to stakeholders Work on communicative action points out that communication may shape actors’ preferences (Risse 2000). The first function of communication is thus to define and create a shared sense of value with respect to the benefits of collaboration. Anti-corruption implies an intangible payoff: improved quality of government. While research demonstrates that impartial and efficient government institutions benefit well-being, government effective- ness, and economic growth, the potential payoff may nonetheless be poorly understood. Moreover, as in any collective action dilemma, some actors benefit tremendously from the status quo, and for those actors the benefits of collaboration may be yet more difficult to appreciate. That said, fostering a joint understand- ing of the benefits of quality of government is crucial to inducing such commitments. Establishing a direct and clear causal chain between a policy proposal and expected outcomes has emerged as a possible effect of communicative action and also characteristic of successful norm campaigns (Finne- more and Sikkink 1998). Norm entrepreneurs, actors who aim to raise awareness of a problem and place it on the policy agenda, at times initiate this function (Finnemore and Sikkink 1998). It may also, however, transpire entirely through interactions among actors who share a more diffuse objective to promote positive change. The aim to mitigate corruption may, in other words, not be the primary focus for any single actor initially, but emerge through communication among key actors as a necessary means to achieving other ends. In the six cases studied, eradicating corruption was in fact not the initial goal of stakeholder collaboration. As respondents recounted, this avoided the negative implication of identifying and punishing perpetrators. Instead, corruption was framed as an obstacle to achieving other benefits, i.e. the kind of gains which par- ticipating stakeholders could accrue from introducing measures to increase transparency or accountability, or from rule-bound transactions between authorities and stakeholders. Specifying the payoffs in a positive way “starts” the conversation from the assumption that all want these positive outcomes; political actors, even those engaged in corruption, can hardly argue against such objec- tives. The payoffs varied: some of the payoffs broadly concerned the city’s economic development (Lviv, 433), increased revenue for social services (Lviv, 453) or a vision for a new form of governance that could 14 stimulate socio-economic development (Vinnytsia, 614). Other payoffs mentioned were sector- or agency- specific, such as the city’s attractiveness for international investors (Chernivtsi, 112), a ‘clean’ reputation of a public authority and better standing in transparency rankings (Lviv, 433), protection of architectural her- itage (Ivano-Frankivsk, 233), or better public service for citizens, such as healthcare (Zhytomyr, 732). Blame and post hoc accountability are not the main focus, which allowed bringing a broad range of local officials on board. Interestingly, communicative processes in most cities did not explicitly link underperformance in the past to corruption. The only exception was the city of Chernivtsi, where a change of government had occurred, and respondents could more freely refer to corruption in the past. Instead, most interviewees described that underperformance had resulted from unclear procedures and a lack of transparency, which may also have given rise to suspicions of corruption, which in turn may have hindered the establishment of new businesses or tarnished the reputation of the city. That said, when asked directly what actions they had taken to address corruption, they mentioned the same measures and policy tools mentioned in conjunction with addressing underperformance. Voicing and recognizing others’ commitments Resolving a collective action dilemma and sustaining cooperation requires that actors “gain a reputation for being trustworthy and reciprocate others’ efforts to cooperate” (Ostrom 2010, 551). A first step is that actors need to declare their commitment to pursue a collaborative strategy. While seemingly self-evident, in collaborative efforts under a high degree of uncertainty, as with efforts to reduce corruption, such procla- mations are needed. Developing rules to mitigate corruption requires considerable effort, and tacit support is unlikely to convince others of intended effort. Key actors publicly proclaiming their intention to commit to a collaborative effort may also endogenously induce change in the proclaimer, especially if done repeat- edly. “Over time, people come to believe what they say, particularly if they say it publicly” (Risse, Ropp, and Sikkink 1999, 15). Moreover, when a critical mass of actors publicly commit, others become more suscep- tible to shaming to follow suit. In our six cases, relevant stakeholders verbalized their commitment to a specific measure or to the principles of transparency and accountability in government in front of a relevant audience. Local authorities made such proclamations in a range of settings, such as expert roundtables (e.g. in Chernivtsi and Lviv), assembly meetings at which CSOs were in attendance (Chernivtsi), public hearings, and investor meetings (both Vin- nytsia). Interestingly, our interviewees report one of the functions of the expert roundtables (such as on local democracy or procurement) was to elicit commitment for anti-corruption initiatives or at least for transparency and accountability in general from municipal leaders. When the mayor invited CSOs to internal meetings of public officials, he enabled trilateral political-executive-civil society commitment. An executive officer recalls that these meetings made one feel like “whether you want it or not, [we] had to take measures and act in this [transparency] direction” (Chernivtsi, 112). Communicating with the public directly on con- struction projects, Vinnytsia LPAs committed a broad audience to address issues in the area of construction. Mutual holding to account Perhaps the most important consequence of publicly stated commitments is that they set the stage for another communicative function, which relates to converting cheap talk into costly effort (Risse 2000, 8). 15 Once persons of authority publicly commit to a reform effort, they help create or affirm a normative stand- ard against which they can later be held to account. Hold an authority accountable entails reference to a shared principle or legal standard. Reference to such a standard elevates an expression of criticism into an accountability action, and transforms a chorus of critics into an accountability forum, i.e. the set of actors who might raise uncomfortable questions without having the formal authority of a principal (Bovens 2007; Olsen 2015). Public proclamations of commitments thus make an actor susceptible to accusations of back- peddling, deceit or grandstanding. Publicly stated commitments are in fact indistinguishable from grand- standing absent other communicative actions described here. Invoking others’ commitments to secure follow-through may be more central to consolidating political will in the area of anti-corruption reform than in other policy issues. The tension between individual and com- mon benefits in mitigating corruption increases incentives to defect. If key actors do not monitor one an- other’s follow-through on commitments, costly effort may indeed not follow. In the six cases, this type of communication occurred when activists reminded authorities to adhere to their own procedures, decisions, and public commitments related to corruption prevention. Importantly, politi- cians also exhibited this function in invoking each other’s commitments to pressure each other to stay true to promises made. This took place in public as well as private, professional spaces and was carried out by both governmental and non-governmental actors. In public spaces, citizens would attend council meetings to “remind” the council on their own commitments – often this referred to issuance of construction permits when authorities were making decisions against formal regulations, raising suspicions of corruption (Cher- nivtsi, Ivano-Frankivsk). An employee of the integrity unit in Lviv similarly reminded the mayor in private conversation of his own anti-corruption stance when progress lagged. Local politicians in Chernivtsi estab- lished an informal communication platform between themselves and the mayor to collaboratively identify and remove corruption risks or particularism in each other’s proposals: If such questions even arise in the council of the majority, that this or that proposal should be supported, … it is very strictly checked for compliance with the legisla- tion. No one in this majority wants to support illegal things even if they would come from a deputy of the council majority or mayor himself, [just] because he is a mayor (133). Communication in working groups at times served to hold actors accountable not only for their voiced commitments but also for tacit agreement to an anti-corruption initiative. For example, an executive officer recalls that he would confront a politician-member of the group for not supporting a decision during voting, precisely with an argument that “you have been to the working group, and did not raise any objections” (422, Lviv). Communication to make commitments operative The second overarching function of communication relates to moving from commitments to operative policies and technical solutions. Here as well, three specific functions emerged in the cases. 16 Identifying and prioritizing strains of the problem Some definitions of political will stipulate that a necessary condition is that key actors share a common conceptualization of the problem at hand (Post, Raile, and Raile 2010). As corruption schemes vary and accountability systems have myriad loopholes and shortcomings, thorough mapping of the terrain is needed. Paradoxically, corruption may on the one hand be evident to government officials and business actors, and even civil society and the general public, but only a small set of insiders are likely to know the details of how rents are extracted. Developing effective policy solutions requires thorough analysis and evaluation of the problem. Evaluating corruption and identifying its specific strains is particularly challenging, however, and may require communication carried out under conditions of anonymity. The interviews reveal that actors were able through dialogue to assess the effects of corruption on local business and government operations, and to identify specific corruption schemes prevalent locally. Some of these communicative processes took the form of interviews with local officials conducted by an analyst to assess the quality of the local procedures in which informants described the “bottlenecks” of procurement processes or the “price” (the typical bribe or kickback amount) of a construction permit. In others, analysts mapped corruption risks by identifying government operations with low rates of compliance with transparency requirements, or where officials have a high degree of discretion in the use of resources with insufficient oversight. The goal was not to identify perpetrators but to discover risks in respective processes. On occasion, such interviews were a part of formal Corruption Risks Assessments (CRA) and included end users, mostly businesses, or monitoring organizations from civil society (in Chernivtsi, Lviv and Vinnytsia). Such interviews were not the only form of communication that helped authorities identify local strains of corruption. A municipal unit entrusted with corruption prevention in Lviv, together with a local CSO, kicked off their co-creation of the anti-corruption strategy with a workshop to identify corruption strains. They invited local activists and businesspeople whose activities depend on renting municipal property or receiving permits, and a neutral facilitator, a university professor, helped organize the discussion and distil the primary findings. Developing operative solutions Identifying and reaching agreement on appropriate, effective and viable solutions constitutes another core component of many definitions of political will. While identifying strains of the problem may entail confi- dential dialogues with embedded actors, developing solutions may entail seeking out and harnessing relevant policy and legal expertise. Policy brokers and policy entrepreneurs may, like norm entrepreneurs, be central to developing such solutions (Mintrom and Norman 2009), and may be representatives from higher levels of government or civil society, possibly with backing and training from international donors. Like several of the functions mentioned above, however, developing operative solutions need not result from the efforts of a specific type of actor nor from the transfer and diffusion of policy solutions from elsewhere. Some solutions may genuinely emerge from local deliberations and negotiations regarding the most appropriate solutions to meet local conditions (see Wickberg 2020). 17 In the cases, communication ensured that proposed solutions incorporated expert and experiential knowledge of anti-corruption tools or procedures to close loopholes. In only a few instances, local author- ities developed operative solutions without involving other stakeholders. The forms of communication var- ied between working group discussions, bilateral conversations and municipality-donor dialogue. For exam- ple, in efforts to increase transparency, the coordinating agency or, in some cities, an external consultant, held bilateral conversations with heads of executive departments to “translate” the requirements for trans- parency into specific technical solutions, which it would then refine together with the respective department (444 in Lviv, 125 in Chernivtsi, in Zhytomyr the development of integrity plan followed the same approach, 732). In other cases, the LPA would develop operative solutions together with international experts and local civil society (Chernivtsi, 614). Validating proposed measures Communication to collaboratively develop solutions aids in enhancing the goodness-of-fit of policy solu- tions. In order to further ensure the viability of these solutions, further communicative interactions with the actors directly involved in implementation may be needed. Vetting through formalized consultation with stakeholders often occurs in policy development; policy proposals circulate to agencies, service providers and industry for input and feedback. This consultative work serves both to refine proposals as well as aid in the legitimation of the proposal (Peters 1992). Consultation can allow for a co-constitution of policies rather than a mechanical process of injecting ideas and negotiating interests. Vetting policy solutions may both refine proposals and adapt to local conditions, and preempt opposition and obstruction at a later date. In the cases, once a draft of a policy solution was available, local authorities carried out “reality checks” of the proposed measures with relevant audiences, ranging from experts to potentially affected stakeholders. Authorities sought feedback through informal bilateral consultations, public presentations and round tables, as well as advisory bodies of experts and activists. In Chernivtsi, a personal acquaintance facilitated the process: A councilor, who communicates very well with them – with these civic activists – he threw them our draft regulations, they read and made remarks to us, where they saw that what we were doing was wrong, so to say (133). Elsewhere, drafts were discussed in groups of experts and potentially affected stakeholders. In Lviv, the anti-corruption unit drafted a code of ethics for councilors and convened a meeting with councilors and the council secretariat to “criticize the code and make something better of it.” Some councilors raised objections to the draft, prompting revisions (444). In Vinnytsia, draft proposals of any policy underwent mandatory review by sectoral advisory bodies, to collect “proposals, commentaries and recommendations, on the basis of which executive authorities formulate decisions to be adopted” (624). Politicians of course discussed and modified policy proposals in formal settings, such as through standing commissions, where councilors also signal whether they generally support the proposal (Vinnytsia 673, Lviv 453, 493, Ivano-Frankivsk, 233). Some cities instead used more informal “agreement councils,” ad hoc meeting of faction heads before the plenary, where, once again, participants sense each other’s readiness to support the measures (Chernivtsi 112). At times, consultants funded by international donors would develop proposed measures and then validate them in more informal communication with relevant actors (Lviv on Integrity Sector, Chernivtsi on Integrity Plan). 18 Persuasion to sustain change Persuasion constitutes a core element of both policy change and norm diffusion, and consequently of the work of policy and norm entrepreneurs (e.g. Finnemore and Sikkink 1998; Mintrom and Norman 2009; Risse, Ropp, and Sikkink 1999; Tarrow 2005). Studies of norm change tend to see persuasion as a process of socialization through which entrepreneurs seek to induce internalization of norms, i.e. by changing actors’ understanding of what it means to be a good person, or state. As change almost always implies a cognitive burden for actors involved, and potentially also foregoing rents, continual persuasion is essential. As a communicative function, persuasion was at work in several of the functions mentioned above (as suggested in Figure 3), though it was not necessarily so. For example, key actors sometimes on their own initiative publicly commit to anti-corruption efforts, but at times were persuaded to do so by the force of the better argument, or fear of reputational costs. Heads of agencies participated in the co-constitution of policy solutions, but at times did so only after shared objectives or moral standards were invoked. In our cases, this function was especially prevalent when a draft regulation or initiative needed political backing, or when implementation and follow-through of a policy initiative was weak. In both situations, policy entrepreneurs delivered tailor-made arguments to convince stakeholders to keep on track. Indeed, in the six cases, interviewees recounted the role of persuasion to overcome inaction, disagreement or opposi- tion to proposed initiatives by key municipal managers. Persuasion assumed two forms. First, argument sought to highlight payoffs either for the individual (if an executive agrees to co-create transparent procedures for their department, it would allay suspicions of cor- ruption, 433), or for the agency (e.g. efficiency gains, since clear rules would expedite administrative decision making). Second, in a more deliberative form, actors would address concerns and objections with explana- tions and counterarguments. For example, faced with an idea of an electronic auction, a municipal property department expressed doubts about the economic advantages of the approach and expressed concerns that the tool would be cumbersome for businesses. In response to this argument, actors agreed on a pilot to test the solution small-scale (Lviv). This way, parties allowed themselves to be persuaded of either positive or negative outcome of the experiment instead of simply insisting on own original preferences (453, 493). Discussion and conclusions Political will, often noted as crucial to the success and failure of anti-corruption efforts, rarely takes center stage in corruption research. As such, the question of what processes and factors might induce or foster anti-corruption political will has largely been overlooked. Our study proposes a means to study anti-corrup- tion political will empirically. Based on interviews describing the processes, actors involved and interactions among them that have transpired in six settings in recent years, we inductively pinpointed seven communi- cative functions that ran through all the cases. Institutional settings for communication varied between and within the cities. While working groups to develop measures were used in all six cities, they also were used in the cities where anti-corruption political will was not detected. Similarly, no patterns emerged in the usage of rounds tables, media or advisory coun- cils or other communication platforms, suggesting that effective communication can transpire in a variety of organizational settings. 19 We identified three clusters of communicative functions which together develop a collective commitment to anti-corruption and a mutually shared confidence in others’ commitments. Communication helps key stakeholders define the “stakes” from their own perspective, and then reaffirms commitments throughout the process. Second, communication was crucial in identifying strains of problem, prioritizing the most egregious schemes for rent extraction, developing solutions, and then validating those measures. And third, along the way of negotiating commitments persuasion, often carried out by dedicated standard-bearers, creates coherence in and across the messy communicative events, allowing actors to deepen their collective political will. Importantly, these communicative functions do not represent a linear process: that is, one need not start from making payoffs visible and gradually move towards vetting of the operative solutions through persuasion. In our cases, the development of operative solutions was often the entry point for communication: actors came to realize and articulate the payoffs while exchanging arguments for and against operative solutions. In contrast to some conceptualizations of political will which see political and administrative actors as the source of initiatives for anti-corruption policies (Brinkerhoff 2000, 242–43), the cases suggest that the im- petus at times came from non-governmental actors, and in some cases emerged from interactions among actors rather than a single actor. Political will emerged in cities with both proactive and reactive stances of authorities, provided that they allowed and engaged in communication that then transform their initial open- ness into collaborative action on anti-corruption measures. Reactive openness may be sufficient when local officials have an intent to counter corruption but lack the internal capacity to take needed measures. Inter- national technical assistance in such cases may be highly useful. In the long term, however, developing internal capacity may be needed to sustain political will. From a policy perspective, the findings are instructive and, in several respects, encouraging. If political will is treated as an individual level quality, the best advice for those hoping for reform is simply to wait and hope for the proper leader to emerge. The findings presented here indicate that even where officials are only moderately willing – and do not even create a specific unit to promote anti-corruption reform – it is still possible to initiate and build momentum for anti-corruption initiatives. Given an impetus for reform either on the part of civil society or business, positive change can ensue. Our findings deviate from those in the literature on norm change (Finnemore and Sikkink 1998) on one instructive point. Interviews did not find evidence that communication developed clear cause-and-effect connections between corruption and underperformance (and thus expected payoffs). When discussing op- erative solutions, actors focused on achieving transparency or government efficiency; when discussing pay- offs, corruption was framed in the terms of a mere “suspicion” that could emerge due to flawed procedures and that suspicion (not corruption) could undermine said benefits. Key actors were thus able to discuss the problem while avoiding self-incrimination or the prospect of investigating and punishing perpetrators. From this perspective, focusing the positive outcomes associated with high quality of government and working towards them together, can be productive to start constructive discussions with stakeholders, some of whom may likely benefit from the status quo. Finally, it is worth reiterating the scope conditions. As a country with moderate levels of corruption, cities in Ukraine at the time of the interviews nonetheless had favorable conditions. National protests in 2015 placed corruption at the top of the policy agenda and ignited anti-corruption movements at the local level 20 throughout the country. This, combined with the comprehensive decentralization reforms created oppor- tunities for change and reform in local government. Even with this window of opportunity, we found no evidence of positive change in most cities surveyed. The development of political will to address corruption, even under comparatively favorable conditions, entailed significant and sustained coordination efforts among relevant stakeholders. Epilogue Since the full-scale Russian war of aggression that began on 24 February 2022, some of our interlocutors have enlisted to defend Ukraine as members of the Ukrainian Armed Forces. Others ensure that LPAs function despite the war: they continue providing local public services to residents and internally displaced persons, support (permanently attacked) infrastructure and do so under immense psychological duress of uncertainty and incessant news of fallen soldiers and civilians, friends, and strangers. At the same time, there is a strong belief that Ukraine will prevail, and discussions of post-war reconstruction within Ukraine and with international partners are ongoing. The factors propelling success described here may be informative for designing the recovery interventions, since mutual trust in each other’s commitment for integrity during the recovery process will be crucial to sustain local development. 21 References Ankamah, Samuel Siebie, and S. M. Manzoor E Khoda. 2017. ‘Political Will and Government Anti- Corruption Efforts: What Does the Evidence Say?’ Public Administration and Development 38 (1): 3–14. https://doi.org/10.1002/pad.1815. Bader, Max, Oksana Huss, Andriy Meleshevych, and Oksana Nesterenko. 2019. ‘When Does Anti- Corruption Activism in Ukraine Have Impact?’ Leiden University, Anti-Corruption Research and Education Center. Batory, Agnes. 2012. ‘Political Cycles and Organizational Life Cycles: Delegation to Anticorruption Agencies in Central Europe’. Governance 25 (4): 639–60. https://doi.org/10.1111/j.1468-0491.2012.01599.x. Blickle, G., N. Schütte, and A. Wihler. 2018. ‘Political Will, Work Values, and Objective Career Success: A Novel Approach – The Trait-Reputation-Identity Model’. Journal of Vocational Behavior 107: 42–56. https://doi.org/10.1016/j.jvb.2018.03.002. Bovens, Mark. 2007. ‘Analysing and Assessing Accountability: A Conceptual Framework1’. European Law Journal 13 (4): 447–68. https://doi.org/10.1111/j.1468-0386.2007.00378.x. Brinkerhoff, D.W. 2000. ‘Assessing Political Will for Anti-Corruption Efforts: An Analytic Framework’. Public Administration and Development 20 (3): 239–52. https://doi.org/10.1002/1099- 162X(200008)20:3<239::AID-PAD138>3.0.CO;2-3. Cabinet of Ministers of Ukraine. 2014. ‘Concept of local self-governance and territorial power reforming’. Order 333-2014-р. https://despro.org.ua/en/news/partners-news/detail.php?ELEMENT_ID=1381. Chêne, Marie. 2010. ‘Overview of Corruption and Anti-Corruption in Sierra Leone’. 256. U4 Expert Answer. U4 Anti-Corruption Resource Center CMI Chr. Michelsen Institute. https://www.u4.no/publications/overview-of-corruption-and-anti-corruption-in-sierra-leone. Dudley, William. 2019. Ukraine’s Decentralization Reform. Research Division Eastern Europe and Eurasia. Research Division Eastern Europe and Eurasia. Berlin: SWP. Escobar, Oliver. 2019. ‘Facilitators: The Micropolitics of Public Participation and Deliberation’. In Handbook of Democratic Innovation and Governance, 178–95. Elgar Online. https://www.elgaronline.com/view/edcoll/9781786433855/9781786433855.00022.xml Fedynchuk, Ivanna, Mykola Yatskov, Taras Sluchyk, Maksym Latsyba, and Maksym Lukiniuk. 2018. Indeks Demokratychnosti Mist - 2018 [City Democracy Index]. Kyiv: Ukrainian Centre for Independent Political Research. https://drive.google.com/file/d/1XjK6eC7NHBdVpPn2z2M-IH2o5l7OiJjy/view. Finnemore, Martha, and Kathryn Sikkink. 1998. ‘International Norm Dynamics and Political Change’. International Organization 52 (4): 887–917. 22 Huss, Oksana. 2020. How Corruption and Anti-Corruption Policies Sustain Hybrid Regimes. Strategies of Political Domination under Ukraine’s Presidents in 1994-2014. Stuttgart: ibidem-Verlag. Huss, Oksana, Max Bader, Andriy Meleshevych, and Oksana Nesterenko. 2020. ‘Explaining Variation in the Effectiveness of Anti-Corruption Activism in Ukraine’s Regions: The Role of Local Context, Political Will, Institutional Factors, and Structural Factors’. Demokratizatsiya: The Journal of Post-Soviet Democratization 28 (2): 201–28. https://doi.org/10.1093/ajae/aau104. Jones, David Seth. 2017. ‘Combating Corruption in Botswana: Lessons for Policy Makers’. Asian Education and Development Studies 6 (3): 213–26. https://doi.org/10.1108/AEDS-03-2017-0029. Jongen, Hortense. 2021. ‘Peer Review and Compliance with International Anti-Corruption Norms: Insights from the OECD Working Group on Bribery’. Review of International Studies 47 (3): 331–52. https://doi.org/10.1017/S0260210521000097. Karklins, Rasma. 2005. The System Made Me Do It: Corruption in Post-Communist Societies. Routledge. https://www.routledge.com/The-System-Made-Me-Do-it-Corruption-in-Post-communist- Societies/Karklins/p/book/9780765616340. Keudel, Oleksandra. 2022. How Patronal Networks Shape Opportunities for Local Citizen Participation in a Hybrid Regime. Stuttgart: ibidem Press. https://cup.columbia.edu/book/how-patronal-networks-shape- opportunities-for-local-citizen-participation-in-a-hybrid-regime/9783838216713. Khutkyy, Dmytro, and Kristina Avramchenko. 2019. ‘Impact Evaluation of Participatory Budgeting in Ukraine’. Kyiv. https://doi.org/0.13140/RG.2.2.11468.36485. Lukerya, Ivan, Taras Sluchyk, Ivanna Fedynchuk, Mykhailo Shelep, and Mykola Yatskov. 2016. Indeks Demokratychnosti Mist [City Democracy Index]. Kyiv: Ukrainian Cntre for Independent Policy Research. https://drive.google.com/file/d/0B8u_pYIkoApyV1NfMnZuaXpUak0/view. Mintrom, Michael, and Phillipa Norman. 2009. ‘Policy Entrepreneurship and Policy Change’. Policy Studies Journal 37 (4): 649–67. https://doi.org/10.1111/j.1541-0072.2009.00329.x. Mungiu-Pippidi, Alina. 2015. The Quest for Good Governance: How Societies Develop Control of Corruption. Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9781316286937. Nitsova, Silviya, Grigore Pop-Eleches, and Graeme Robertson. 2018. ‘Revolution and Reform in Ukraine. Evaluating Four Years of Reform’. Washington: PONARS Eurasia. www.ponarseurasia.org. OECD. 2017. ‘Anti-Corruption Reforms in Ukraine: Pilot 5th Round of Monitoring Under the Istanbul Anti-Corruption Action Plan’. https://doi.org/10.1787/b1901b8c-en. Olsen, Johan P. 2015. ‘Democratic Order, Autonomy, and Accountability’. Governance 28 (4): 425–40. https://doi.org/10.1111/gove.12158. 23 Ostrom, Elinor. 1998. ‘A Behavioral Approach to the Rational Choice Theory of Collective Action: Presidential Address, American Political Science Association, 1997’. The American Political Science Review 92 (1): 1–22. https://doi.org/10.2307/2585925. Ostrom, Elinor. 2010. ‘Polycentric Systems for Coping with Collective Action and Global Environmental Change’. https://doi.org/10.1016/j.gloenvcha.2010.07.004. Persson, Anna, and Martin Sjöstedt. 2012. ‘Responsive and Responsible Leaders: A Matter of Political Will?’ Perspectives on Politics 10 (03): 617–32. https://doi.org/10.1017/S1537592712001648. Peters, B. Guy. 1992. ‘The Policy Process: An Institutionalist Perspective’. Canadian Public Administration 35 (2): 160–80. https://doi.org/10.1111/j.1754-7121.1992.tb00686.x. Post, Lori Ann, Amber N. W. Raile, and Eric D. Raile. 2010. ‘Defining Political Will’. Politics & Policy 38 (4): 653–76. https://doi.org/10.1111/j.1747-1346.2010.00253.x. Quah, Jon S. T. 2015. ‘Evaluating the Effectiveness of Anti-Corruption Agencies in Five Asian Countries: A Comparative Analysis’. Edited by Dr Jon S.T. Quah and Professor Chilik Yu. Asian Education and Development Studies 4 (1): 143–59. https://doi.org/10.1108/AEDS-10-2014-0050. Quah, Jon S.T. 2017. ‘Five Success Stories in Combating Corruption: Lessons for Policy Makers’. Asian Education and Development Studies 6 (3): 275–89. https://doi.org/10.1108/AEDS-03-2017-0031. Risse, Thomas. 2000. ‘“Let’s Argue!”: Communicative Action in World Politics’. International Organization 54 (1): 1–39. https://doi.org/10.1162/002081800551109. Risse, Thomas, Stephen C. Ropp, and Kathryn Sikkink, eds. 1999. The Power of Human Rights: International Norms and Domestic Change. Cambridge Studies in International Relations. Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9780511598777. Roberts, Sean, and Oleksandr Fisun. 2014. Local Governance and Decentralization Assessment: Implications of Proposed Reforms in Ukraine. The Ukraine Mission of the United States Agency for International Development (USAID). https://doi.org/10.13140/RG.2.1.5093.0721. Romanova, Valentyna, and Andreas Umland. 2019. Ukraine’s Decentralization Reforms Since 2014. Initial Achievements and Future Challenges. London: Chatham House. https://doi.org/10.1787/9789264204836-en. Sally, David. 1995. ‘Conversation and Cooperation in Social Dilemmas: A Meta-Analysis of Experiments from 1958 to 1992’. Rationality and Society 7 (1): 58–92. https://doi.org/10.1177/1043463195007001004. Šmigoc, Elena B. 2015. ‘Political Will for Political Behaviour’. In Political Will: A Short Introduction, edited by Dino Abazović and Asim Mujkić, 33–61. Sarajevo: Friedrich Ebert Stiftung. Tarrow, Sidney. 2005. The New Transnational Activism. Cambridge Studies in Contentious Politics. Cambridge: Cambridge University Press. https://doi.org/10.1017/CBO9780511791055. 24 TI Ukraine. 2018. ‘Transparent Cities 2018’. 2018. https://transparentcities.in.ua/en/rating/?rating- year=current. Tilly, Charles. 1985. “War Making and State Making as Organized Crime” in Eds. Peter Evans, Dietrich Rueschemeyer, and Theda Skocpol. Bringing the State Back. Cambridge: Cambridge University Press, 1985. Tsebelis, George. 2002. Veto Players: How Political Institutions Work. Princeton University Press. https://www.jstor.org/stable/j.ctt7rvv7. USAID Office for Democracy and Governance. 2006. ‘Anticorruption Program Brief: Anticorruption Agencies (ACAs)’. https://pdf.usaid.gov/pdf_docs/Pnadm208.pdf. Wickberg, Sofia. 2020. ‘Global Instruments, Local Practices: Understanding the “Divergent Convergence” of Anti-Corruption Policy in Europe’. These de doctorat, Paris, Institut d’études politiques. https://www.theses.fr/2020IEPP0030. i Veto players (Tsebelis 2002) are those with the power and position to block a reform and whose positive assent is necessary to change the status quo (Post et al. 2010, p. 661). ii We excluded Kyiv, because as a capital city it has a different governance model. We exclude Uzhhorod, because despite multiple efforts, we could not find civil society respondents for the survey. We also excluded Symferopil in the Autonomous Republic of Crimea and Donetsk and Luhansk due to their temporary illegal occupation by Russia. We included cities that per- formed the functions of the regional centres in the two latter regions: Mariupol and Sievierodonetsk. iii https://transparentcities.in.ua/en, last accessed 17 July 2022 iv https://publicityindex.org/, last accessed 1 September 2021 v The interviews were conducted in the Ukrainian language by the research team: Oksana Huss, Dmytro Iarovyi, Khris- tina Khambekova, Oleksandra Keudel, and Alina Los. At each interview, one team member led the conversation, and another made notes. Interviews were transcribed and coded (inductive, focused coding) with the help of MAXQDA. vi The 1997 Law of Ukraine On Local Self-Government stipulates the structure of local government and the main provi- sions of legislative-executive relations at the local level. 25