Policy Design, Diffusion, and Policy Instrument Attitudes

Policymakers are in the constant struggle to design policies that appropriately address societal, economic, and ecological needs. Literature on policy design has contributed to the field of policy studies with a systematization of possible policy instruments (Howlett 2004, 2023; Acciai and Capano 2021; Capano and Howlett 2020). But how do policymakers decide whether and which policy to implement and how does the public respond to different policy instruments? These questions will be treated in the fourth issue of this year's Review of Policy Research volume from different perspectives. The first part of this issue focuses on governments and what influences their decision to implement or not implement certain policies. In line with research on policy diffusion (Marsh and Sharman 2009; Mallinson 2021; Mistur et al. 2023), the first three papers illustrate how the decisions of neighboring regions can influence (local) governments to follow the same approach. The subsequent papers address how compatibility with already established policies through path dependence (Beyer 2010; Pierson 2014), institutional capacity (Toshkov et al. 2021; Khan and Hussain 2024), and policy learning (Dunlop et al. 2018; Crow et al. 2023; Zaki et al. 2023) shape policymakers' decisions, and what constitutes effective policymaking. The second part of this issue turns toward citizens as the recipients of political decisions and looks at how policies' consequences differ throughout societal groups looking at environmental justice (Sovacool et al. 2017; Hanke and Guyet 2023; Ambrose et al. 2025) and how the public responds to different policy instruments (Veselý 2021; Nekola et al. 2022). This covers analyses on public support and compliance with governmental policies as well as the impact of public information on citizen behaviors. In the first contribution, Gregoire-Zawilski and Siddiki (2025) investigate how state-level net metering policy designs differ and diffuse across the United States. Bridging literature on policy diffusion (Graham et al. 2013; Porto de Oliveira et al. 2023) and policy design (Schneider and Ingram 1993; Howlett and Mukherjee 2018), the authors take on a more nuanced approach to understand how specific design elements such as goals, incentives, and eligibility diffuse at the sub-national level instead of the policy as a whole. Based on text data from legislations between 1981 and 2016, the authors use dyadic event history analysis to model policy diffusion across 45 US states that have enacted net metering policy, that is, a regulation which allows electricity customers to sell excess renewable electricity back to the grid. The analysis reveals that internal factors such as citizen ideological alignment as well as bipartisan legislatures served as key drivers of diffusion, while the democratic control of the legislature or the presence of a democratic governor did not have a significant effect. As such, the study presents net metering as a citizen-oriented, politically flexible policy instrument whose diffusion is mainly influenced by the local political and economic context. Moreover, the study confirms that strategic learning from neighboring states has had a significant effect on net metering policy design, thus stressing the role of policy imitation. Contributing to the understanding of how policies can diffuse transnationally and thus be transferred across institutional context, the paper by Li et al. (2025) explores China's approach to policy transfer by zooming in on the development of its science funding system. To date, China's policy transfer has been dominated by so-called “pilot cities” in which local governments implement relatively low-cost trials of foreign innovative policies (see e.g., Qian 2017; He et al. 2024). In this experimental governance approach, if the pilot does not reach envisioned results, underlying knowledge is usually discarded. Calling for a more sustainable strategy towards “learning from abroad” (Dolowitz and Marsh 2000) and drawing on strategic niche management research (Schot and Geels 2008), Li and colleagues present “adaptive niche cultivation” that allows to appropriately modify foreign expertise to national political conditions through interaction between policymakers and experts. The authors illustrate the effectiveness of this approach with an in-depth case study of China's science funding system based on document analyses and interviews. Here, scientists and policymakers collaboratively adjusted foreign ideas through incremental adaptation and learning which enabled China to overcome institutional barriers and resource limitations in the development of its scientific infrastructure. Moreover, the authors shed light on interaction dynamics and power relations in the domestication of foreign policy knowledge which offers valuable insights into how policy transfer can succeed across contexts. Zhang et al. (2025) continue this line of argument by investigating the conditions under which local governments “reinvent”, that is, alter policy innovations transferred from abroad, using the empirical example of Municipal Solid Waste Classification Policy (MSWCP) in China. Drawing on Everett Rogers' seminal diffusion of innovations theory (Rogers 1983) and related scholarship on policy innovation and diffusion (Glick and Hays 1991; Nicholson-Crotty and Carley 2016; Dolowitz et al. 2020), Zhang et al. conduct a crisp-set Qualitative Comparative Analysis (Ragin 1987; Schneider and Wagemann 2012) of 31 Chinese cities to identify condition configurations that lead to policy reinvention. The analysis focuses on the presence or absence of six key conditions which cover internal factors (financial resources, government attention, policy experience), intergovernmental relations (peer competition, organizational learning), and external influences (social demand). The results reveal nine distinct pathways to policy reinvention, illustrating how different causal configurations can lead to the same outcome. Among the most influential configurations are those combining intergovernmental learning and peer competition, which can drive policy reinvention even in cases with limited financial capacity and low social demand. However, the study showed that all conditions can favor reinvention, while their absence can be compensated for by other factors. Not only neighboring regions but also previous policies of the same government can shape the formulation of future policies, for instance through path dependency or policy feedback (Wilsford 1994; Peters et al. 2005; Beyer 2010; Berry and Berry 2017). By examining how established cigarette taxes influenced the adoption of new vaping regulations across US states, Kim (2025) contributes to this scholarship with an operationalization of policy compatibility that facilitates the alignment of new policies with those in place. In line with the previous paper of this issue, Kim also draws on the diffusion of innovations theory (1983) and argues that the normative compatibility of policies (i.e., similarity of goals and underlying values) and operational compatibility (i.e., suitability to the state's regulatory landscape) shape policymakers' decision to either align new policies with old ones or to innovate. The article empirically illustrates this concept with statistical analyses of an original panel dataset on state-level vaping regulations across the United States from 2010 to 2020. The findings confirm that rather than treating vaping as a novel policy issue, states often incorporate it into their existing tobacco control systems. As such, states with more comprehensive anti-smoking measures and higher cigarette taxes are significantly more likely to adopt accordingly strict vaping regulations, thus staying in line with the institutional structures, normative commitments, and regulatory approaches of existing policies. McAlexander et al. (2025) extend the diffusion approach by incorporating institutional capacity and regime type into the explanation of inefficient policy instruments. They investigate why governments maintain inefficient and costly subsidies, particularly in the realm of fossil fuels such as gasoline. The study draws on a dataset of monthly gasoline price fluctuations across 157 countries and distinguishes democracies from autocracies using the Polity IV score. Institutional capacity is operationalized with two indicators from the World Bank's Worldwide Governance Indicators (WGI). The authors explore how regime and institutional capacity interact to shape a government's willingness to phase out inefficient gasoline subsidies. Their key finding is that meaningful subsidy reform occurs only when democratic governance is coupled with strong institutional capacity. The study offers valuable connections to broader policy process perspectives. From an empirical standpoint, it would be worthwhile to replicate the analysis using alternative datasets. The Sustainable Governance Indicators (SGI), for instance, provide comparable measures of democracy, governance capacity, and policy outcomes (Croissant and Pelke 2022), along with additional control variables, though only for a subset of the countries examined here. The article adds important dimensions by linking policy capacity to democracy measurement and empirical policy performance. While McAlexander et al. (2025) emphasize that effective policy reform depends on both democratic legitimacy and institutional capacity, the sixth contribution in this issue shifts the focus to crisis governance and examines how specific configurations of policy capacity explain government effectiveness. Tan et al. (2025) investigate the conditions under which Chinese provincial governments were able to respond effectively to the initial wave of the COVID-19 pandemic. Rather than assuming that early action alone determines policy success, the study draws on the differentiated concept of policy capacity to identify institutional combinations that support high-performing crisis responses. Building on an analytical framework of coordination (Koop and Lodge 2014) and especially the multi-level policy capacity framework developed by Wu et al. (2015), the authors distinguish four core types of policy capacity: analytical, regulatory, delivery, and mobilization (a crisis-specific adaptation of coordination capacity). Using crisp-set Qualitative Comparative Analysis (csQCA, Rihoux and Ragin 2009) across 27 Chinese provinces, the study identifies multiple causal configurations associated with successful outcomes. The outcome is the effectiveness of crisis management measured by the case fatality rate. Conditions include early regulatory response, high analytical capacity, high mobilization capacity, and sufficient medical resource delivery. The findings show that no condition is sufficient on its own. Early regulatory response, for instance, contributes to success only when embedded in a broader configuration that includes strong analytical or mobilization capacity. Provinces with limited healthcare infrastructure were still able to respond effectively when they could mobilize collective action and coordinate inter-agency efforts. Likewise, robust delivery systems were not enough if risk anticipation and political coordination were weak. Tan et al. (2025) reinforce the broader argument of this issue that policy effectiveness is shaped by institutional configurations rather than isolated drivers. By adapting the concept of policy capacity to the context of crisis response, it demonstrates the importance of complementarities among distinct functional capacities. In doing so, it contributes to a growing literature on how institutional design shapes the responsiveness and resilience of governments under pressure (Kuhlmann et al. 2024; Friede 2025). The seventh article in this issue deepens our understanding of policy change during a crisis by focusing on the role of policy learning. Wang (2023) investigates how instrumental and political learning interact to shape state-level decisions to tighten or loosen COVID-19 public gathering restrictions in the United States between May and December 2020. Using a Qualitative Comparative Analysis (QCA) of 27 cases across six states (Delaware, Minnesota, North Carolina, New Jersey, Virginia, and Louisiana), the study identifies five distinct learning configurations that explain policy change in both directions. Wang finds that decisions to strengthen restrictions were primarily driven by instrumental learning, such as worsening health indicators or updated evaluations of infection risk. In contrast, decisions to relax restrictions were more complex and often motivated by political learning, particularly declining gubernatorial approval ratings and public protests. These findings suggest that instrumental learning dominates expansion decisions, while political learning plays a larger role in relaxation decisions. The article contributes to policy process research by demonstrating how policy learning can operate within a crisis (Kamkhaji and Radaelli 2017; Hornung and Bandelow 2022; Crow et al. 2023). It demonstrates that instrumental and political learning are not mutually exclusive but can interact in dynamic configurations depending on the political and epidemiological context. Methodologically, the use of QCA to trace learning patterns across time and cases (Verweij and Vis 2021) offers an alternative to conventional statistical models. Theoretically, the study extends ongoing debates about learning under uncertainty and confirms the interdependence of different learning modes (Dunlop and Radaelli 2013; Lesch and Millar 2022). While Wang (2023) examines how policy learning shaped governments' decisions during the COVID-19 crisis, the eight contributions shift attention to the societal consequences of these decisions, particularly their unequal environmental impacts. Campbell et al. (2025) investigate whether the pandemic-induced reductions in economic activity and mobility led to more equitable exposure to air pollution across racial and ethnic groups in the United States. The study addresses the implications of policies targeting reduced use of gasoline-powered cars, as seen, for example, in California, and their interaction with patterns of environmental justice. It uses GIS data, federal datasets, and sensor measurements for the development of air pollution between 2019 and 2020, taking into account the respective regional proportions of ethnic groups, people in poverty, average income, population density, and other variables. While some average pollution levels declined, the reductions were not evenly distributed. In some cases, disparities persisted or widened. Black communities remained disproportionately affected by some pollution levels even in the context of overall improvement. The study shows that structural disparities in exposure to environmental harm are not automatically resolved through macro-level behavioral changes such as reduced traffic volume. This article complements the institutional and behavioral perspectives presented earlier in this issue by focusing on distributive outcomes. It highlights the importance of embedding equity considerations into policy evaluation. The results make it clear how important it is to distinguish between policy output and outcome. The study invites further research into whose interests are served by specific policy changes, and thereby may contribute to scholarship on policy windows (Bodenheimer and Leidenberger 2020; Amri and Logan 2021; Shirazi et al. 2024). When studying policy design and instruments, three mechanisms are generally of interest: those that link actors' decisions in favor of or against policy instruments to the adoption of policy instruments; those that connect policy instruments to the desired goal and evaluate whether they attain it; and those that assess the extent to which policy instruments influence public opinion. Starting with the first mechanism, public policy scholars have recently proposed the concept of an Autonomy of Policy Instrument Attitudes (APIA) (Nekola et al. 2022; Veselý 2021) which suggests that individuals have specific preferences for policy instrument types that are independent of the policy problem or other contextual factors. To provide further empirical evidence on this matter, Veselý et al. (2025) investigated whether a general preference for a particular instrument type could also account for specific preference for policy measures, using the example of the response to COVID-19. The results confirm this relationship but also indicate that the attitudes towards policy instruments are related to political views, and that APIA's predictive power is relatively low. They call for further research, particularly concerning the measurement of APIA. Another major area of interest when adopting policy measures is their effect, particularly on the behavioral change of target groups and potentially their preferences. The effectiveness of policy instruments is a well-established area of research in public policy (Capano and Lippi 2017; Eicke 2025; He et al. 2025; Mukherjee et al. 2021). While existing studies have acknowledged the effectiveness of public information campaigns, there is a research gap regarding the impact of the framing and presentation of information on their effectiveness. In their study, Kim et al. (2025) conducted a vignette experiment to test the effectiveness of different message types in altering citizens' attitudes, intentions and behavior in recycling programs. They employed a 2 × 2 design, varying the dimensions of abstract versus concrete messages (“why to recycle” vs. “how to recycle”) and global versus national distance. Regardless of the socially distant or close context, citizens' attitudes and behavior were more pro-recycling when confronted with concrete information on what to sort and recycle, and how to do so, than when confronted with abstract information on the reasons why people should recycle household waste. Policy instruments usually do not come alone but in mixes (Howlett and Ramesh 2023; Wu et al. 2024). Therefore, change at the level of individual attitudes, intentions, and behavior is often not the result of one policy instrument but of several. Ideally, policy instruments should complement each other in order to achieve the set objective. But which instruments do this? Taking the example of waste sorting, Li, Jiang, et al. (2025) performed a conjoint experiment to analyze the effects of different combinations of policy tools. In three rounds, participants were presented with two policy scenarios reflecting different combinations of nudges, boosts, social norms, and monetary incentives. Participants were asked to choose one and also to indicate their willingness to comply. The authors found that nudges and economic incentives were the most effective, and that combining nudges and boosts with economic incentives could enhance effectiveness, meaning that policy instruments work together. Willingness to comply is strongly associated with the acceptability of policy measures, which in turn depends on how the measures are conceived. In environmental policy, one of the most prominent instruments is carbon taxation (Andersen 2019; Beiser-McGrath and Busemeyer 2023; Knaggård and Hildingsson 2025). However, increasing taxes also increases revenue, and how the government utilizes revenues from carbon taxes, a process known as revenue recycling (Hedegaard and Kongshøj 2024), might also influence the acceptability of the policy measure. Drawing from the example of the Irish carbon tax reform, Muth (2025) investigates the relationship between revenue recycling and acceptability. Based on data from 16 interviews with political elites, he concludes that direct compensatory measures and green investments are more politically acceptable than the “fee and dividend” model, whereby revenues are redistributed equally across society, which only received strong support from the Green Party. Furthermore, in addition to revenue recycling, more an important of the of carbon taxes as rather than