Beleidsvorming en beleidsadvies in zwakke democratische omgevingen: inzichten uit Bulgarije, Moldavië en Oekraïne

Untangling the ways in which policy makers make decisions and consider different inputs in the policy process has been a central endeavour of much of public policy scholarship. In this context, the notion of a policy advisory system (PAS) has emerged as a helpful conceptual heuristic. The concept has given rise to a prolific research agenda seeking to better understand the complex processes underpinning the development of public policies. The departure point of this study is the acknowledgment that much of the thinking about policy advice provision and the policy advisory system, however, is derived almost exclusively from the experience of consolidated and affluent Western and Northern European and Anglo-Saxon democracies. Consequently, such scholarship is based, at least implicitly, on a series of assumptions about the policy and administrative environment, within which advice provision and utilisation takes place, which simply do not hold in many countries. Such assumptions include the presence of a capable and merit-based bureaucracy, the relative strength of formal institutions and established rule of law as well as the presence of coherent party structures, among others. Existing literature therefore tells us little about what to expect of advisory processes taking place in settings not sharing those characteristics. The dissertation addresses this gap by exploring how key governance features of what I call weak democratic settings, or political systems characterised by poor democratic quality and enduring authoritarian practices, impact the nature and dynamics of policy advisory systems. To do that, the dissertation studies advisory practices in three post-communist countries in Eastern Europe, namely Bulgaria, Ukraine and Moldova. Through a mixed-method research design relying on semi-structured interviews, data on public expenditure and contracting and social network data, the study addresses the overarching research question: How do policy advisory systems operate in weak democratic settings and how does this compare to what we know about advisory practices elsewhere? The dissertation's individual chapters address two distinct but related subthemes within this larger question. The first subtheme is concerned with the place of individual (types of) advisors within PAS in weak democratic settings. Chapters 2, 3 and 6 speak to this first component of the dissertation by zooming in on the advisory role of three distinct groups of actors. Chapter 2 examines the role of public sector consultants in policy making in Bulgaria, a country with historically limited in-house analytical and policy capacities within the bureaucracy. The chapter relies on a mixed-method research design drawing on an original dataset of government tenders for the procurement of policy-related services in the period 2010-2023 and 14 interviews with consultants and civil servants. The chapter demonstrates the significant propensity to rely on external consultants for advice provision and other core policy work and identifies a number of important trends in the use of such external sources of advice. It also highlights the role of the EU in both driving and facilitating the outsourcing of policy work. Chapter 3 represents a qualitative case study focusing on the process of formulation of Bulgaria's National Climate and Energy Plan for 2021-2030. The aim of the analysis is to examine how expert advice and different types of evidence are utilised in contexts characterised by historically weak culture of technocratic rationality and strong politicisation. The chapter finds that the instrumental use of evidence-based advice provided by experts remains exceedingly rare. Instead, such advice typically serves merely a perfunctory and symbolic function, while advice rooted in more populist and political logics proves much more successful. Chapter 6 zooms in the advisory role and activities of non-governmental organisations (NGOs) by examining a "deviant" case study, namely the process of the formulation of Bulgaria's Law on Social Service. The Law was a cornerstone of the country's service provision reform and featured near-unprecedented inclusion of the NGO community in the substantive policy making. The second subtheme within the dissertation deals with the question of influence within the policy advisory process and, more specifically, with the factors and conditions, which facilitate influence. Chapters 4, 5 and 6 fit within this subtheme of the dissertation. Chapter 4 investigates how current theoretical thinking about the origins of influence measures up to the realities of policy processes within hybrid regimes. Through a qualitative analysis of 45 semi-structured interviews with policy advisors from Ukraine and Moldova, the chapter explores the empirical utility of the theoretical model put forward by Manwaring (2018), which explains policy influence as a result of the interaction of content, context, demand and supply factors. Although the chapter confirms the overall applicability of the framework in the hybrid regime context, it also highlights important differences in the empirical manifestation of its key elements in such settings. Chapter 5 then goes a step further and examines the extent to which different logics of influence apply to different policy domains in order to explore the comparative importance of regime-level vs. domain-level influence determinants. The chapter utilises a mixed-method research design combining social network analysis and qualitative analysis of 50 semi-structured interviews in the comparative analysis of two PAS in Bulgaria, one in the area of sustainability policy and another in the field of social policy. It does so in order to examine the extent of externalisation of both advice provision and advisory influence in typically state-centric political systems. It finds meaningful differences in the dynamics of externalisation across the two domains, suggesting that a domain-level as opposed to a country-level approach to studying PAS dynamics may be optimal. Finally, Chapter 6 builds on those insights and opens the "black box" of the policy making process. By focusing on a single policy process as a case study, it unpacks the complex combination of conditions which allow otherwise disadvantaged advisory actors, in this case non-governmental organisations, to achieve influence. The chapter utilises an analytical approach based on analytic eclecticism. It not only recognises the role of agency, structure and institutions in bringing about policy outcomes (in this case advisory influence), but highlights the various complementarities between those three groups of explanatory factors in producing the observed outcome. Four key findings emerge from the analysis of PAS dynamics, namely the tendency of advisory systems and processes in weak democratic settings towards personalisation, politicisation, informality and internationalisation. In terms of the origins of advisory influence, a nuanced and complex picture emerges from the dissertation. The study underscores the fact that influence is an outcome characterised by equifinality. Far from being subject to a single explanatory logic, advisory influence is a product of diverse range of favourable conditions, often acting complementarily. Still, the analysis indicates that in weak democratic settings like those under examination power and influence are intensely individualised. As such, they often intersect with individual actors' embeddedness in and access to informal (and even interpersonal) networks of economic and political power. Given the high level of political control over all aspects of the policy process, influence was further found to be strongly dependent on demand-side explanations. Those typically related to the attitudes of key decision makers and informal veto players (e.g. oligarchs and state captors) vis-à-vis certain advisors or types of advice or the perceived need (or lack thereof) for specific advice.