This article empirically investigates how different institutional designs of implementation arrangements are associated with specific configurations of policy capacity and, in turn, with variation in implementation outputs. Building on the policy capacity framework developed by Wu et al. (2015) and its extension to collaborative settings, the article analyses three irrigation public–private partnership (PPP) projects in Kenya—Galana Kulalu, Perkerra, and Chakama—through a small-N, case-oriented comparative design. The study suggests that variation in implementation outputs is best understood through the interaction between institutional design and the configuration of analytical, managerial, and political capacities activated across individual, organizational, and systemic levels. The findings indicate that political capacity serves as an important enabling condition for more effective collaborative arrangements, while the relative weight of analytical and managerial capacities varies with the coordination logic embedded in each arrangement. More hierarchical and rule-based arrangements are associated with stronger implementation outputs where analytical and managerial capacities are institutionalized, whereas more networked and adaptive arrangements rely more heavily on managerial-political capacities rooted in trust, partner mobilization, and negotiated coordination. By connecting policy capacity research with implementation research on collaborative arrangements, the article contributes to a more relational, institutionally grounded understanding of why similar policy interventions yield different implementation outputs under broadly comparable conditions.
Artificial intelligence is transforming public governance, yet existing policy design frameworks remain theoretically inadequate to this transformation. This article develops the Algorithmic Policy Evaluation Framework (APEF), a diagnostic-normative framework grounded in critical theoretical resources, Foucauldian subjectivation, Fricker’s epistemic injustice, Habermasian accountability theory, and Fraser’s parity of participation, for evaluating the democratic legitimacy of algorithmic governance once adopted. The APEF identifies four structurally interrelated dimensions, epistemic opacity, distributional injustice, responsibility displacement, and contestability foreclosure, theorised as a mutually reinforcing configuration of algorithmic power, distinct from and complementary to existing explanatory accounts of why such systems are adopted. Using a critical interpretive design with a most-different-cases logic across three policy domains and five countries, the article illustrates the framework’s cross-domain reach within liberal-democratic administrative states, an explicit scope condition. It situates the APEF against the wider literatures on algorithmic power, algorithmic governmentality, and algorithms in the policy process, and engages debates on algorithmic impact assessment, participatory AI, and the EU AI Act. Five normatively grounded design principles are proposed, their operationalisation and jurisdictional variation discussed explicitly, with the tensions between them treated as sites of democratic negotiation rather than technical resolution.
Why do decentralized governance systems underperform in coordination-dependent policy domains despite substantial policy activity, institutional mandates, and technical knowledge? This study addresses this question through non-structural flood governance in Southeast Sulawesi, Indonesia, as a case of fragmented cross-sector policy integration. Drawing on a qualitative multi-level case study, the research combines documentary analysis, 42 semi-structured interviews, and light process tracing to examine how intelligence, prescription, implementation, and appraisal are connected across decentralized bureaucratic systems. Across contrasting local settings, the findings identify recurrent brokerage gaps—weakly authorized, insufficiently routinized, or episodic connective arrangements linking distributed knowledge, authority, and decision arenas. Process evidence shows that risk intelligence could generate cross-sectoral prescriptions without integrated organizational commitments, substantial implementation could coexist with weak coordination, and formal appraisal could remain disconnected from collective learning. These dynamics produced programmatic density without governance integration and may reinforce structural bias toward visible interventions alongside broader fiscal, political, institutional, and environmental constraints. The study conceptualizes brokerage gaps as a meso-level processual mechanism through which bureaucratic fragmentation can generate decision-process disconnection and, under conditions of high institutional interdependence, contribute to governance underperformance.
To what degree do mass publics learn, such that they can engage in the substantive debates of policy design? And how does the nature of that learning relate to public trust? These questions are both relevant to policy makers and central to the literature on learning, but scholars have struggled to draw a clear link between public sentiment, modes of elite and mass public learning, and the maintenance (or loss) of trust in government decisions. We argue that social media offers a particularly useful lens through which to examine the relationship between public sentiment and opinions regarding political events, how those lead to public engagement around policies, and how political elites’ responses to that engagement contribute to the maintenance of public trust. Drawing on the case of Canada, we use Twitter to construct a database of over 450,000 distinct tweets between March 10th and September 10th, 2020 on the topic of child care and school closures and re-openings during the first wave of the COVID-19 pandemic. Using automated coding, topic modeling, and sentiment analysis expressed across government actors and the Canadian public, we find evidence that, while the Canadian public demonstrated increasingly negative sentiment and high levels of anxiety, public trust—including support for government institutions in Canada—remained high. We posit that this finding is linked to the fact that the public was heavily engaged in conversations about the specific instruments and settings needed to ensure a safe return to school in the fall of 2020. That ability to engage in public puzzling alongside governments served to maintain trust, even under conditions of uncertainty and stress.
Our study uses a context-mechanism-outcome approach to highlight how trust diversity emerges in wildfire collaboratives and how it impacts collaborative outcomes. Available literature identifies four distinct trust traits in natural resource collaborations including affinitive trust, dispositional trust, rational trust and procedural trust. This study is based on a survey disseminated from an inventory of 133 collaboratives yielding 38 responses. In an earlier study, we identified three distinct wildfire collaborative archetypes and survey responses reveal that each archetype is equivalently represented in the survey response sample. We conduct a qualitative comparative analysis (QCA) of the relative importance of the presence/absence of trust traits for environmental, social-political-economic and collaborative process outcomes. We find that environmental outcomes require the presence of complete procedural trust while rational and affinitive trust are key factors impacting collaborative process outcomes. Additional qualitative analysis of survey responses reveals the different trust traits whose presence or absence are significant for different collaborative archetypes. Social outcomes show more reported success than environmental outcomes, though collaborative process gains display the most reported success. This study highlights the importance of future investigations of the influence of the geography of landscape and property rights in trust development processes and subsequently collaborative outcomes.
This article examines cross-national diffusion in the UNESCO Creative Cities Network (UCCN) by focusing on a gateway outcome: whether and when a country records its first entry into the network. Existing diffusion research generally expects cultural similarity to facilitate adoption. We argue that this expectation does not necessarily hold in an international organization-led network that explicitly values cultural diversity and broad geographical representation. Using a directed dyad-year panel of 176 countries from 2004 to 2019, we estimate discrete-time dyadic event history models of the annual probability that a country records its first UCCN designation. The results show that greater cultural distance from prior member countries is positively associated with successful first entry, while institutional distance remains a significant constraint and domestic readiness continues to matter. These findings suggest that cultural distance can operate as a conditional advantage in diversity-valuing global networks and that diffusion theory should distinguish first entry from later expansion.
The literature has gone to great lengths to describe how administrative coordination can support climate policy implementation. Yet few studies have adopted both a configurational and an actor-oriented perspective to understand which combination of policy capacities and policy design features bureaucrats perceive as necessary or sufficient to effectively engage in administrative coordination for climate policy implementation. Analyzing eighteen local departments implementing climate plans in four cities (Bologna, Lausanne, Padua, Zurich) and across various policy sectors, a configurational approach based on mixed methods allows testing two hypotheses: the first (H1) is that a combination of multiple, high-level organizational capabilities is sufficient for effective coordination. The second one (H2) is that both integrated planning and high organizational policy capacities together are sufficient for effective coordination. Finding evidence for H2, this paper contributes to the ongoing debate regarding the integrative policy capacities needed to foster the ecological transition at the local level. It makes a distinctive contribution by systematically clarifying how the interplay between policy capacities and policy design features shapes administrative coordination outputs.
Why do some policy agendas disappear not merely as a matter of declining attention, but as a consequence of becoming politically unsayable? Scholarship on policy processes has devoted extensive attention to agenda‑setting and non‑decision, but far less to the mechanisms through which once‑contested policy options are actively withdrawn from the space of legitimate political action. This article develops the concept of agenda exit to explain how policy contestation is stabilised when proposals become non‑viable to articulate and are subsequently constrained against return. Drawing on a process‑tracing analysis of housing tax reform in Australia from 2014 to 2021, the article shows how a coordinated field of property‑sector actors transformed electoral defeat into a durable agenda constraint through practices of reputational risk construction, substitution, and boundary‑policing. The core contribution is to theorise agenda exit as a condition of structured absence: an interregnum in which a policy remains technically conceivable but is rendered politically unsponsor‑able within ordinary arenas of contestation. This clarifies a distinct mode of policy stability, in which governability is secured not through consensus or resolution, but through the defensive narrowing of political possibility. By specifying how policy options become absent without formal decision, the analysis extends existing work on agenda control, blame avoidance, and policy continuity, and highlights the importance of studying the conditions under which policies can no longer be credibly proposed.
Does the “nationality” of artificial intelligence (AI) shape public acceptance of its integration into public policymaking? We contend that, amid the rise of techno-nationalism, citizens are more supportive of embedding domestically developed AI systems in government policymaking than foreign alternatives. Drawing on a survey experiment conducted in China, we find that having domestic AI rather than foreign AI increases individuals’ support for the use of AI in the formulation of both major and minor policies, with this effect particularly pronounced among those with strong techno-nationalistic orientations. People favor domestic AI because they believe it can improve the quality of policymaking while mitigating potential risks. Moreover, individuals are more likely to endorse the use of domestic AI in minor policies than in major ones. This study contributes to our understanding of the intersection of techno-nationalismand public policy.
Fragility is a term widely used in research and policy, yet its conceptualization remains contested even within fields. We develop a general framework for understanding and measuring fragility. It applies across fields and is oriented toward policy use. The framework provides practitioners with a common structure for thinking about fragility, a set of candidate measures, and guidance for constructing them. Fragility is defined as the ease with which an object fails. We distinguish it from risk, resilience, and vulnerability and identify five types of measures: critical stress, failure condition, conditional probability, unconditional probability, and composite indices. We examine the use of these measures and identify the contexts in which they are most appropriate. Together, they imply a trade-off between causal interpretability—the ability to explain how failure occurs—and operational feasibility—the ease with which a measure can be constructed with reasonable validity. We then discuss the use of fragility measures in policy and develop a protocol for constructing fragility measures, which we apply to a case study of urban areas exposed to wildfire. Even small differences in failure definitions lead to different fragility measures and policy implications.
Electronic voting (e-voting) is pivotal for modern democratic institutions in the digital age. Despite nearly six decades of e-voting technology development, significant disparities exist in its global introduction. Existing studies primarily focus on country-specific experiences, leading to limited generalizable insights into what defines successful e-voting introduction and its enablers. This study bridges this gap by proposing a comprehensive definition of successful e-voting introduction. We then explore on what grounds e-voting introduction succeeds or fails by applying fuzzy-set qualitative comparative analysis (fsQCA) to data from 35 OECD countries. Results reveal that the broader policy environments, digital readiness and citizens’ trust and ability to use technology significantly shape e-voting policies. Government commitment and policy alignment are identified as plausible complementary explanations for both success and failure. The study underscores the interplay of institutional and societal factors, rather than isolated influences, is the key driver of e-voting success.
To date, scholars have not devoted sufficient attention within the Multiple Streams Framework (MSF) to the emotional strategies employed by policy entrepreneurs—a gap this study addresses through an analysis of Israel’s judicial reform (January 2023‒December 2024). Using qualitative content analysis, AI-assisted thematic analysis, and a public survey, we examine statements made by two figures who spearheaded the reform: the minister of justice and the chair of the Knesset’s Constitution, Law and Justice Committee. Findings show that, contrary to our hypothesis, nearly all emotional appeals made by these figures triggered negative emotions among respondents, with no variation across political affiliations and no difference between reactions to the emotional content and to the political figures themselves. However, as expected, right-leaning individuals reported lower levels of negativity than those who identify with the center or the left. These findings suggest that fueling negative emotions served as the entrepreneurs’ primary mobilization strategy. In theoretical terms, the study contributes to the MSF by illustrating how emotional appeals may function as an endogenous mechanism within the problem stream. The findings are consistent with an interpretation that associates emotional framing with early stabilization of public perceptions of reform, potentially limiting subsequent opportunities for problem reframing as the policy process unfolds, and with observable agenda effects, including mass mobilization and a temporary government suspension of the reform. Moreover, initial emotional responses to the political figures are also consistent with an anchoring dynamic that may limit subsequent opportunities for reframing. By linking entrepreneurial emotional strategies to patterns of public emotional responses, the article highlights how emotions can accelerate problem recognition and reinforce contested interpretations, thereby influencing agenda dynamics at the stage of problem definition, prior to the coupling of problems, policies, and politics within the MSF conceptualization.
Actors involved in the governance of a given issue often attribute blame and responsibility to other actors. While such dynamics are usually studied for highly politicized and conflictive issues involving acute crises, pressure for actors to act, and related reputational implications, we study blame and responsibility attribution for a situation of little public attention, a so-called creeping crisis. In this context, we ask how patterns of blame and responsibility attribution develop across actors from different governance levels, actor types, and organizational roles. We do so by descriptively mapping the public media discourse on antimicrobial resistances (AMR) in Switzerland and Germany, based on 633 statements coded in media articles. Results reveal generally more responsibility than blame attribution to others. Whether blame and responsibility are rather attributed to similar rather than dissimilar others depends on actors’ governance level, actor type or organizational role. Overall, understanding the dynamics of blame and responsibility attribution is important, as they influence policy design, and might increase the risk of political deadlock.
Does celebrity endorsement encourage individuals to incur private costs to provide public goods? We examine this issue in the context of climate policy, specifically aviation emissions. We ask: Does the information that a celebrity purchases carbon credits to offset their air travel emissions encourage individuals to do the same? We focus on endorsement from Taylor Swift, one of the most famous pop stars. To differentiate a “Swift Effect” from a generic celebrity effect, we examine the hypothetical purchase of offsets by two additional celebrities who share similar demographic characteristics (White women), Caitlin Clark and Angelina Jolie. We work with a pre-registered vignette-based online survey experiment (N = 1309; N = 609 attentive respondents) conducted on a representative panel of U.S. respondents. We find that, compared to the reference group (no celebrity endorsement), endorsement of any of the three celebrities does not change respondents’ willingness to purchase carbon offsets. Overall, respondents inclined toward environmental action (e.g., concerned about climate change, not frequently flying for work/leisure), and non-White respondents, were more willing to pledge to purchase offsets. Blue-collar respondents were less willing, whereas other socio-demographic attributes, such as gender, did not have a significant effect on this willingness to purchase offsets.
Why do countries facing similar pressures for electric vehicle (EV) transitions adopt different policy coordination arrangements? Existing scholarship pays less attention to how historically rooted institutional configurations shape the emergence and persistence of coordination modes. This article addresses this gap through a comparative study of Indonesia and Thailand, two Southeast Asian middle-income economies that adopted national EV strategies in the late 2010s but developed divergent coordination trajectories. Drawing on historical institutionalism, it develops a framework of compound path dependence, arguing that coordination arrangements are shaped by the interaction of technical lock-in, relational networks, and organizational learning. Using comparative process tracing from the 1960s–2024, the analysis identifies two distinct governance modes: in Thailand, consensus-based incrementalism built on strong industrial capacity, institutionalized consultation, and committee-based routines; while in Indonesia, directive-driven ad-hocism rooted in a resource-centered development path and weaker industrial coordination legacies. Through this, this study advances the literature on public policy and sustainability transitions. It reveals a critical trade-off between institutional coherence and policy agility in sustainability transition, and argues that policy coordination choice is a product of structure-induced agency constrained by historical legacies.
In the European Union (EU), substantial policy efforts aim to conserve and enhance forest biodiversity. Under the European Green Deal (EGD), flagship policies, including the EU forest and biodiversity strategies for 2030, envision the ecological transformation of nature, society, and economy. Central to these efforts are norms—soft institutions and shared standards that guide actors’ behaviour. Drawing on diffusion approaches, we identify key norms promoted by the EU and examine their normative impact on forest biodiversity politics in member states. Using qualitative content analysis and expert interviews, we analyse current EU strategies and confront them with past and current national forest and biodiversity strategies in Germany and Spain to investigate whether and to what extent they reflect European norms. Our results indicate that the EU articulates norms through the EGD that contribute to shaping national discourses on forest biodiversity policies. However, contested interpretations among the analysed member states reveal a dynamic and context-dependent norm diffusion process. By shedding light on such complex interactions, this study advances our understanding of the EU’s influence on member states and contributes to the European integration literature and norms research by offering empirical insights into diffusion and contestation processes.
Policy feedback is often viewed as either reinforcing or undermining policy stability. However, it can arise in multiple ways simultaneously, creating mixed policy feedback with complex effects on policy stability. We examine how mixed policy feedback affects policy stability by developing a framework to identify within- and across-level feedbacks, and applying it to two cases of contested climate policy involving domestic carbon markets in Australia (2014–2025) and Canada (2015–2025). In Australia, a nascent baseline-and-credit scheme eventually became entrenched despite initial policy instability, while in Canada, a seemingly stable carbon pricing policy survived early conflict but was later partially reversed. By examining how multiple feedbacks arose and interacted within changing political contexts, we show how mixed policy feedback can affect policy stability, leading to potentially surprising outcomes. This implies a need to go beyond assessing the net effects of mixed policy feedback to consider complex interactions alongside exogenous developments.
Public governance has become increasingly saturated with plans, targets, indicators, delivery units, and performance routines, yet persistent policy underperformance continues to be diagnosed mainly as an implementation problem. This comment argues that part of that pattern reflects not just empirical difficulty, but an analytical habit built into how the policy cycle is commonly used. While the cycle remains valuable as a heuristic, it under-specifies a distinction between two different but interdependent governing processes: the political authorization of collective purpose and the administrative-operational organization of delivery. When that distinction is blurred, unresolved ambiguity in strategy is absorbed into plans, targets, and performance frameworks, which are then treated as if they expressed already-settled commitments. Tensions originating upstream are thus displaced downstream and recoded as problems of coordination, implementation, and managerial weakness. The result is a recurring misdiagnosis: what appears as delivery failure may begin in weakly stabilized political direction. Distinguishing between purpose design and purpose delivery helps clarify why accountability so often settles on delivery systems, why performance regimes may optimize action without reopening purpose, and why tools such as PDCA are more appropriate for delivery than for the political authorization of direction. The policy cycle need not be abandoned, but used with greater analytical precision.
How evidence is used to frame policy issues plays a crucial role in shaping which knowledge is deemed relevant to policymaking. Drawing on argumentation theory and framing literature, we argue that evidence functions as a tool that speakers use strategically and systematically to support specific elements of frames (regarding causes, consequences, or solutions), thereby determining which bodies of knowledge are mobilised in political discourse. Using quantitative content analysis of media, trade, and parliamentary discourse in Switzerland from 2013 to 2022, we identified four frames relating to the issue of pesticide risk reduction: health, environment, economy, and practice. We found that across all frames, the most emphasised frame element is also the one most likely to be supported by evidence. Overall, consequence claims were more likely to be backed by evidence than claims about causes or policy solutions. These patterns reveal which bodies of knowledge are mobilised in political discourse in support of certain policy stances and where the presented frames lack supporting evidence. Recognising the role of framing in evidence use can thus support more transparent and reflexive policymaking by helping researchers and policymakers to identify unavailable or overlooked knowledge and addressing gaps in both research and policymaking.
This article explains how Ecuador decriminalized euthanasia—an unlikely outcome in a conservative setting—through an INUS configuration that coupled: (1) an institutional design enabling citizens to bypass legislative deadlock; (2) a problem broker whose personal struggle reframed a moral conflict as a constitutional claim; and (3) a policy community able to translate that claim into constitutionally legible safeguards. This triad, when combined with the strategic coupling efforts of policy entrepreneurs in a given political context, can create a configuration of conditions that enables binding normative change despite adverse partisan dynamics. Using theory-building process tracing, we reconstruct the causal sequence from court filings to the ruling, drawing on official decisions, interviews, and public statements, complemented by World Values Survey evidence on attitudinal shifts and digital traces from X that capture coalition building and framing. Conceptually, we put institutional design in the foreground of the Multiple Streams Framework as a structural determinant of both policy and decision windows. Where constitutional activation is available, an apex constitutional review body can convert agenda salience into binding normative change, even under adverse political conditions, provided that proposals reach policy readiness and are institutionally legible.
