Policy calibration and policy acceptability: assumptions, evidence, and practical implications

Abstract

Policy calibration is often described as a technical phase of policymaking after the most important aspects of the proposed policy have already been decided. In this article, we show that policy calibration also has an important political dimension because the acceptability of policies depends often more on concrete aspects of policies than on general policy views. We argue that understanding policy acceptability, i.e. attitudes to proposed policies, requires understanding the link between general and specific policy attitudes. The article starts with an illustrative example, showing how acceptability might change when policy becomes calibrated. It is followed by a summary of five theoretical approaches to the link between general and specific policy attitudes. These theories are confronted with empirical evidence. This article demonstrates that although attitudes to general principles of policy design are often hard to change simply by providing individuals with new information, it is easier to increase (or decrease) the acceptability of specific policies in that manner. This paradox might be explained by the relatively loose link between individuals’ attitudes to general policy principles and their attitudes to the specific use of policy tools.

Keywords:

• Policy calibration
• policy acceptability
• policy attitudes
• survey experiments
• principle-policy gap

1. Introduction

Politicians are usually reluctant to design and implement policies that are likely to face strong public opposition or not be endorsed by the “critical mass” of the public (Grelle and Hofmann 2024). Acceptability of policies thus plays an increasingly important role in policy-makers’ decisions on policy design and is also of growing interest to policy scholars (De Groot and Schuitema 2012). The concept of policy acceptability is usually defined as the extent to which people accept the introduction of newly proposed public policy. In other words, this concept denotes peoples’ attitudes to policy proposals (Kojo et al. 2022).¹

Many factors influencing policy acceptability have been identified, including the level of policy coerciveness, perceived benefits and effectiveness of policy, or socio-psychological factors such as political orientation or personal values (Drews and Van den Bergh 2016). However, only tiny attention has been devoted to the “when” question of policy acceptability, i.e. to the question of how acceptability changes during the policymaking cycle. It is well known that the content of policy proposals changes over time. The policymaking process usually starts with rather broad (macro) policy proposals that are gradually developed into more concrete and specific policies. Interestingly, people with different values and perspectives may want to participate at different times of decision-making. Some are more interested in the macro (and more abstract) phase, while others step in only when the policy becomes more concrete (Perlaviciute and Squintani 2023).

However, the different representation of policy views during the policymaking process is not the only reason why the acceptability of policy might change over time. People often change their individual attitudes toward proposed policies when the abstract ideas are translated into specific policies. Acceptability of policies at abstract and specific level often differ (Sütterlin and Siegrist 2017). This means that the process of “policy calibration,” during which concrete decisions about “the specific ways in which the instrument is used” are made (Gofen, Wellstead, and Tal 2023), has an important impact on policy acceptability. While policy calibration is often depicted as a technical process after the “big political battles are over,” as we try to show in this paper, the specifics of proposed policies might play a crucial role in policy acceptability. Policy calibration thus has not only an important substantial aspect (in terms of policy effectiveness) but also a central political dimension (in terms of policy feasibility).

As all practitioners know, reaching an agreement at the general level is often easier than agreeing on the concrete parameters of a new policy. For instance, it is unlikely a proposal to increase teachers’ salaries will be contested until specific dollar amounts begin to be discussed (e.g. how much salaries will increase, what will be in the pay package, what bonuses will be given, and what tradeoffs with other measures must be made). Sütterlin and Siegrist (2017) have documented that renewable energy technologies enjoy high levels of acceptability in the abstract, but their popularity decreases when respondents are questioned on a more specific level. According to them, that is attributable to the fact that when asked about specifics, people consider more of the negative aspects of renewables and adjust their acceptability accordingly.

However, the opposite might also be true. Consider the example of a government plan to phase out fossil fuel heating. Because individual fossil fuel boilers currently provide almost 60% of the energy consumed for heating living space and water in the EU (Braungardt et al. 2023), announcing such a plan would likely provoke much initial disagreement. Opposition, however, might be softened by providing more information about the plan. For instance, the policy could be accompanied by compensation paid to households to help them replace their source of heating, or it might exempt particular groups of citizens from compliance. Also, as the policy receives greater attention on the public agenda, more information about the rationale for it might be provided, such as the fact that carbon emissions from heating buildings make up 30% of global final energy consumption. This might make people more aware of the issue and consequently more willing to accept the policy.

It is thus vital for practitioners to know how policy acceptability develops over time as a policy becomes more specific and calibrated. Unfortunately, whereas there have been attempts to discuss “how to make policy more acceptable” (De Groot and Schuitema 2012), to the best of our knowledge, no one has systematically explored the “when” question of policy acceptability. Because time and resources are always limited, for practitioners, it is important to know when it is most effective to invest into efforts to increase the policy acceptability. While the related literature on consensus building might provide some guidance, it remains contradictory. On the one hand, it has often been claimed that up-front investment in consensus-building saves considerable time and energy in downstream implementation (Ansell and Gash 2008). Consequently, more effort should be invested in the initial phases of policymaking rather than in later policy calibration. On the other hand, it has been observed that premature stress on consensus-building might suppress political conflict and narrow probing of viable solutions (Hoppe 2022). It could also be time-consuming and unnecessarily slow down the whole process.

This paper aims to fill that gap and connect policy calibration with policy acceptability. We proceed in four steps. First, we provide an example from practice of how policy acceptability has changed when a general idea becomes a specific and calibrated policy. Second, we organize the theories of the link between general and concrete policy attitudes. Third, we confront expectations derived from those theories with the available empirical evidence. In so doing, we focus on how providing information can change the acceptability of calibrated policy. We conclude with implications for practice.

2. Illustrative example

Having once been responsible for creating a national educational strategy, I still remember my surprise at how easy it was to reach an agreement among very diverse stakeholders on complicated topics such as school curricula. However, that agreement was only on a general (one might even say vague) level. When the discussion started to address more specific proposals, the heated debates began.

A particular example in which agreement on basic principles does not ensure agreement on calibrated policy is the introduction of middle-tier in Czech educational governance. The Czech Republic has one of the most decentralized and fragmented systems of educational governance in the world. Basic schools are operated by municipalities that are often very small and have responsibility for just one school. Not surprisingly, these small municipalities do not have much capacity to support their schools. School principals must spend much of their time on administrative and technical duties rather than on leading and supporting teachers. In contrast to most other countries, there is no government institution between the school and the central state (Ministry of Education, Youth, and Sport) that would have any capacity to support and coordinate highly autonomous but often isolated schools.

The introduction of a so-called “middle tier” of elementary school governance is the cornerstone of the Czech Republic’s Strategy for the Education Policy to 2030 and beyond. This document was approved by the Czech government in 2020. It was prepared following wide discussion and consensus-seeking. As one of the members of the core team preparing the strategy, I was struck by the ease with which the idea of introducing a middle tier of governance for elementary schools was accepted by a wide range of heterogeneous actors. There was a near-consensus on the need to establish a new institution that would shoulder some of the principals’ burdens and provide concentrated support for information and communications technology, legal advice, project management services, and other tasks. The idea was elaborated in a pilot project in which middle-tier government institutions were organized in two of the country’s regions by the Ministry. After three years, an external evaluation found the pilot project to be very successful.

And yet, as the idea of adding a middle tier of governance came closer to fully fledged implementation, a number of questions arose: What legal form should the middle-tier institution have? What types of employees would fit the institution best? How many schools should it be responsible for? What powers should it have to enforce its decisions? These ostensibly technical questions proved to be quite contentious. At least two main alternative policies were proposed (one by the Ministry, another by a group of powerful NGOs). These two policy alternatives were not mutually exclusive. They shared the same view of the problem and the design of the proposed solutions, but they differed in many specific ways that caused tension between their supporters.

A very recent evaluation of the stakeholders’ attitudes (Srb 2024) showed that the general idea of introducing a middle tier of school governance is still well-accepted.² However, the “technical” aspects of both proposed implementation alternatives were challenged. Some actors argued that the other proposal was overly ambitious, while others believed that the problem should be solved in “a more complex way.” Still, others asserted that the proposal should, first of all, consider adapting current institutions and should not aspire to create new ones. According to another stakeholder, any profound change should only occur after no less than a decade of thorough discussion. The overall tendency to look at the “bright side” of the general proposal is nicely put by another actor: “We welcome the idea of a middle tier. But only if it does not affect the competencies of principals.”

The main takeaway from this example is that the acceptability of basic principles does not automatically lead to acceptability about how they should be precisely designed and then implemented. In general, the more respondents think about a proposal (and the more they are affected by it), the more “yes, buts” they are likely to see in the proposal. The “buts” do not necessarily change their opinion of the overall policy principle, but they might substantially reduce acceptance of a calibrated policy based on that principle. Let’s look at how this puzzle has been explained in theory.

3. The abstract-concrete link in policy attitudes

As explained above, policy acceptability refers to attitudes to policy proposals. In general, attitudes are assumed to have three components: cognitive, affective, and behavioral (Breckler 1984). The cognitive component concerns one’s beliefs; the affective component involves feelings, and the behavioral component consists of ways or intentions of acting. People’s feelings, beliefs, and behaviors with regard to a policy are related but distinct (Maio et al. 2003, 290). In recent years, academics’ conceptualization of policy attitudes has started to emphasize the affective component over the other two (and especially the behavioral component) because some studies have empirically demonstrated the primacy of the affective component (Banaji and Heiphetz 2010, 358). More often, however, attitudes are conceptualized as a dual construct, i.e. attitudes are assumed to be based on a cognitive component (beliefs and knowledge) and an affective component (feelings and emotion) (Wilson, Lindsey, and Schooler 2000).

Heated debate exists about how policy attitudes can best be conceptualized and measured (Bachner and Hill 2014). There is, however, agreement that policy attitudes exist at different levels of abstraction. The most common conceptualization identifies three levels (Hall 1993; Howlett 2009; Jenkins‐Smith et al. 2014). In any event, it is clear that policy attitudes might differ according to the level of policy concreteness and specificity. This is why acceptability might change in policymaking processes. As noted above, most policies are initially formulated at a relatively general level and later are calibrated in the policy process (Howlett, Ramesh, and Capano 2024).

The process of policy calibration is both political and cognitive. Contrary to the dominant portrayal of policy calibration as a merely technical process, individuals change their attitudes toward the proposed policy throughout the process. Actors involved in the calibration process concretize their expectations and negotiate about specific parameters. Not only the policy itself but also individual attitudes change as the policy becomes calibrated. The actors’ initially rather general views of the policy become more nuanced and sophisticated. Furthermore, with a deeper understanding of what the proposal actually entails, individuals’ acceptability of the policy can increase or decrease. Policy actors gradually accumulate evidence about policy problems and solutions through mutual interaction and, therefore, update their beliefs and preferences with regard to the appropriateness of a policy over time (Moyson 2017).

Understanding the linkage of attitudes to abstract (general) and specific (calibrated) policy attitudes is crucial for understanding policy acceptability. If policy attitudes are hierarchically structured, then policy acceptability requires agreement on general policy principles. On the other hand, if attitudes to concrete policies might deviate substantially from the general principles, acceptability can be achieved later in the policy calibration process. Unfortunately, the links between macro and micro levels of policy attitudes have remained a black box (Kammermann and Angst 2021, 2).

However, based on the academic literature, we can distinguish at least five main theories about how the macro and micro levels of policy attitudes relate to each other (see Table 1).³ Because different theories are based on different assumptions, they lead to different hypotheses on the link between macro and micro policy attitudes. They also have different expectations about the effect providing more information might have on specific policy attitudes.

Table 1. Theories of the formation of attitudes toward policies.

	Assumptions	Subtype	Authors	Empirical expectations on specific policy attitudes
1. Non-attitudes or weak attitudes	Attitudes on specific policy issues are weak; no internal structure can be found.		Converse (2006) and Zaller (1992)	No structure and consistency in specific policy attitudes
2. Top-down models^a	Attitudes on specific policy issues change only when higher-order policy attitudes change	2A. Hierarchical model	Hall (1993), Hurwitz and Peffley (1987), and Peffley and Hurwitz (1985)	Almost impossible to manipulate with new information
		2B. Constrain models	Jenkins‐Smith et al. 2014)	Only secondary aspects open to change (that do not compromise values)
3. Pluralism models	Attitudes on specific policy issues change depending on the structure of higher-order policy attitudes	3A. Complexity model	Sniderman, Brody, and Tetlock (1991) and Tetlock (1986)	It depends on the structure of higher-order aspects.
		3B. Anarchic models	Carstensen (2011)	Specific policy attitudes are easy to manipulate with new information

Source: Author.

^a Also labeled as nested model, hierarchical model, and paradigm model.

First, attitudes toward policies have been conceptualized as “constructed preferences” or “non-attitudes.” This view was first established by Converse (2006) and echoed by many others (Bishop 2008; Zaller 1992). It is based on early public opinion research in which much inconsistency in attitudes was observed. According to this view, individuals often lack political sophistication, and therefore, they form their opinions about public policies on the spot. Consequently, policy acceptability is based on whatever clues they have at their disposal, including various heuristics. If this conceptualization is correct, then policy attitudes should be completely unstable, inconsistent, and highly manipulable by providing new information. There would be no link between attitudes to policies in general and attitudes to concrete policy tools.

The second conceptualization of policy attitudes is more optimistic about individuals’ ability to process relevant information. It can be labeled as a “top-down approach.” This is probably the most common perspective in public policy. It assumes that individuals derive their attitudes toward specific policies from their existing general orientations: “[G]eneral orientations are assumed to ‘constrain’ or determine more specific policy preferences” (Hurwitz, Peffley, and Seligson 1993, 64). This view of top-down formation of policy attitudes exists in at least two versions. The “strongly” deterministic version was formulated by Hurwitz, Peffley, and Seligson (1993) and Hurwitz and Peffley (1987). They created a “hierarchical” model of belief systems about international affairs, where broad attitudes to foreign policy postures (i.e. abstract beliefs about the general direction a government should take in international affairs) were assumed to structure and constrain attitudes to more specific policies (Hurwitz and Peffley 1987). In their later work, they indeed found that general orientations (e.g. preferences for military solutions or containment of adversaries) are more stable than many specific policy preferences (e.g. attitudes toward defense spending or U.S. interventions in Central America) and that much of the over-time consistency in policy attitudes is generated by these more general orientations (Hurwitz, Peffley, and Seligson 1993).

Although somewhat less deterministic, a similar conceptualization of top-down policy beliefs can be found in public policy. It is most explicitly pronounced in the theory of an advocacy coalition framework (ACF), where specific, calibrated policy proposals are constrained by higher-order policy attitudes (Jenkins‐Smith et al. 2014). ACF posits that an individual actor’s belief system is hierarchically structured in that general beliefs (“deep core” and “policy core” beliefs) constrain more specific beliefs. Specific “secondary beliefs” are preferences for different tools for implementation of a core policy: “These may be about the appropriateness and efficacy of specific tactics for realizing the strategies derived from the policy core and, more generally, the deep core. … According to the ACF, these are the beliefs most susceptible to change. Indeed, adjustment of these beliefs may be necessary to protect deep core and policy core beliefs” (Jenkins‐Smith et al. 2014, 486). In contrast to Hurwitz and Peffley (1987), ACF is not entirely deterministic, but it is constraining. Higher-level beliefs define what possible solutions to a problem can be considered and which one from this pre-selected set of solutions is then implemented (Sabatier and Mazmanian 1980). The link between abstract and specific is not as straightforwardly hierarchical as Hurwitz and Peffley’s.

Pluralism models do not assume that specific policy attitudes are derived from general and abstract policy attitudes in that way. Advocates of pluralism models observe and acknowledge a disconnect between abstract ideas and concrete policies. Pluralism models explicitly address the so-called principle-policy puzzle (or principle-policy gap), i.e. the observation that people often fail to reason from the general to the specific and support for a general principle (e.g. racial equality) to concrete policies designed to realize it (Sniderman, Brody, and Tetlock 1991, chapter 4). Again, we can distinguish between “strong” and “weak” versions of pluralism models.

According to a weak model (Sniderman, Brody, and Tetlock 1991; Tetlock 1986), the disconnect between abstract and concrete results from the complexity of policy reasoning. People often have not one but many fundamental values or beliefs. They have “pluralistic ideologies” in which different values are in frequent and intense conflict. Because realizing important values often point to contradictory policies, individuals must make painful tradeoffs (Tetlock 1986). Given that all human beings have limited cognitive capacity for processing information, they rely on knowledge structures derived from past experience or on low-effort heuristics (Tetlock and McGuire 1986).

The strong version of the pluralism model was advanced by Carstensen (2011), who conceives of human beings as “bricoleurs.” According to Carstensen, bricoleurs are non-dogmatic problem-solvers who take stock of existing ideas and pragmatically combine bits and pieces from several paradigms to achieve a solution to a problem. Importantly, the ideas used by bricoleurs do not necessarily fit together coherently. Carstensen likens his concept to Kingdon’s (1984) “anarchic ‘multiple streams’ model” (Carstensen 2011, 157). According to Carstensen, ideas often contain elements of meaning that conflict with each other. “Actors [choose] the different elements an idea should contain, but a more proper picture is probably one of kneading or molding an idea to try to get it to hang together and gain the acceptance of other actors – often while at the same time other actors knead and mold their related ideas” (Carstensen 2011, 157).

4. Some empirical evidence

The theories described above have been tested by empirical research. Survey experiments are a particularly well-suited method of testing attitude formation (Haselswerdt and Bartels 2015). They enable the researcher to control the information that individual research subjects acquire. In that way, they simulate a real-life policy calibration process in which individuals are exposed to many factors (including new information) that might play a role in their attitudes toward policies.

First, the literature does not entirely support the view of Converse, Zaller, and others (Converse 2006; Zaller 1992) about the instability and inconsistency of policy attitudes. Even though most of the public has very unstable views on public policy, 20–40% of the population actually does hold stable and consistent views (Freeder Lenz, and Turney 2019). The most decisive factor for consistency is political sophistication (Sniderman, Brody, and Tetlock 1991). Consequently, elite respondents exhibit more stability in their opinions (Converse and Pierce 1986). Because actors involved in the policymaking process (public officials, politicians, and on-the-ground implementers of policies) are likely to have a high degree of political sophistication and more structured and consistent policy views, it might be deduced that they do not arrive at their views either randomly or entirely pragmatically. Evidence also suggests that individuals’ policy attitudes are at least sometimes internally consistent across levels. To take just one recent example, Gerace, Rigney, and Anderson (2022) found a high correlation between their subjects’ general attitudes toward easing COVID-19 restrictions and their acceptability of specific measures.

Second, robust empirical findings come from comparing the effects of information provision on problem and solution perceptions. Interestingly, it has been repeatedly reported that it is much easier to change the perception of a problem than the perception of a policy solution. Lergetporer, Werner, and Woessmann (2020) found that providing information about the extent of educational inequality strongly increases concerns about educational inequality. However, providing more information to individuals affects their acceptability of equity-oriented education policies only marginally. Similar findings were reported by Kuziemko et al. (2015) on income inequality. Veselý (2023) reports that providing information about the general effectiveness of different types of policy instruments does not affect individuals’ attitudes toward concrete policy measures (in Veselý’s study, support for lowering or raising unemployment benefits). This study suggests that during the policy calibration phase, radical changes in the direction of policy design are not likely.

On the other hand, there is evidence that policy attitudes toward well-calibrated policies can be manipulated by providing selected new information and induce individuals to change their views of a policy in the intended direction (Haaland, Roth, and Wohlfart 2023). Lergetporer et al. (2018) showed that providing information about current levels of education spending and teacher salaries reduced support for more spending and salary increases. Telling the public that increased education spending would be financed through higher taxes further reduced the acceptability of such spending. Finally, informing the public about tradeoffs between different spending options reduced support for smaller class sizes and increased support for teacher salary increases and the purchase of new teaching materials. Other surveys (Cattaneo et al. 2020) confirmed that it is relatively easy to change preferences for calibrated policies such as specific levels of educational spending.

Because attitudes to calibrated policies are relatively easy to manipulate by new information, whereas attitudes to more general or abstract policies are less manipulable, the disconnect between different measurements is likely to be observed (Sütterlin and Siegrist 2017). This is not necessarily due to the complete lack of cognitive connection between the abstract and the concrete. Rather, people might reason differently about abstract and concrete policies. Sometimes, the observed disconnect might be even due to too much reasoning. Heberlein et al. (2005) found in their study on public support for wolf restoration that people who liked wolves most would pay less money to restore 500 wolves and more to restore 200 wolves. That seems paradoxical as so for wolf lovers $500 should be better than $200. In-depth interviews with respondents showed that wolf lovers worried that too many wolves could create problems for wolf restoration, which could lead to wolf poaching and harassment, so they supported fewer, not more, wolves. Consequently, being a strong supporter of any generally formulated policy does not have to translate into being an advocate of the strongest measures at the calibrated level.

To complicate things even further, it has been documented that in certain cases, policy attitudes can change more substantially by providing new information. This change, however, is conditioned by other factors. Diamond, Bernauer, and Mayer (2020), for instance, conducted a survey about climate change and genetically modified food (GMO) policy preferences in Germany and the United States. They found that providing information about the prevailing scientific opinion significantly shifted their subjects’ policy views in favor of scientific opinion, primarily among individuals whose prior attitudes conflicted with the scientific message. Haselswerdt and Bartels (2015) found that citizens prefer tax breaks over equivalent direct government payments, particularly when tax breaks are the government’s usual means of intervention. When direct payments are the status quo, or when any government involvement on the issue is unfamiliar to the subjects, their preference for tax breaks is reduced. They also found that their subjects’ political orientation (i.e. their general policy attitude) has an effect, with conservatives strongly preferring tax breaks over direct payments. The authors suggested that the policy status quo influences citizens’ perceptions of policy proposals.

5. Conclusions and Implications for practice

In democratic societies, public acceptability of the critical mass is believed to be a prerequisite for policy success. Acceptability of policies evolves throughout the policy process. It might decrease or increase substantially during the policy making process during which originally rather abstract policy proposals are developed into more specific calibrated policies. For practitioners it is important to know how the attitudes to general and concrete policies relates, because with this knowledge they can more easily predict how the acceptability will evolve during the policy process.

Unfortunately, the empirical evidence suggests that the link between macro and micro policy attitudes is very complex and depends on a number of factors. More research is thus needed to uncover the link between the acceptability of the general direction of policies and the acceptability of calibrated policies. Currently, it is not possible to formulate simple advice on how to increase public acceptability during policy calibration. However, the available evidence leads to the following tentative practical implications.

First, available evidence suggests that although attitudes to general principles of policy design are often hard to change simply by providing individuals with new information, it is easier to increase (or decrease) the acceptability of specific policies in that manner. More specifically, providing new information to individuals has a more substantial effect on how they view a problem than on how they view proposed general solutions to that problem. Individuals’ opinions about proposed solutions are more difficult to change than the way they view the problem. That is especially true when the proposed general policy design differs from the policy status quo. Nevertheless, the provision of new information can change attitudes to calibrated policies relatively easily, such as an agreement for a specific dollar amount of educational spending. In short, although new information can have a substantial influence on how a final (concrete) policy proposal is perceived, it may not create substantial changes in “deep core” beliefs.

This paradox might be explained by a cognitive disconnect between support for a general principle and support for a specific policy derived from that principle. This principle-policy attitude gap is an everyday reality. Consequently, the acceptability of general and calibrated policies might be – and often are – two different things. The link between attitudes toward general and specific policies is usually rather loose. From a practical point of view, it means that agreement on general policy principles does not ensure their smooth translation into calibrated policy. Individuals’ acceptability of policies can often be substantially increased or decreased by adding (or changing) concrete policy parameters.

As we have seen in our example of introducing a middle tier of school governance in the Czech Republic, even in “clear-cut cases” where agreement on the principle is almost unanimous, technical aspects can be crucial for policy acceptance. Simply put, the devil is in the details, and “secondary aspects” sometimes matter more to the acceptability of a policy than basic principles.

Consequently, if policymakers want to know the real level of acceptability of their proposals, they should focus on specifics and provide people with all the necessary details about the proposed policy (Sütterlin and Siegrist 2017). After all, what is finally approved and implemented is not a general but very concrete and calibrated policy. In practice, the acceptability of calibrated and actually implemented policies matters more than the acceptability of initial general proposals. In other words, general opinions tell us little about whether a policy will be accepted as it is actually implemented. Even seemingly concretely defined policies, such as “to increase public transport subsidy” or “increase fuel taxes,” are still too unspecified to predict their acceptability.

Empirical evidence thus does not fully support the prevailing belief that it is always impossible to achieve policy acceptability without reaching a broad agreement on general principles. Nevertheless, that does not mean that the effort to achieve “sufficient consensus” (Bardach 1998) at the general level of policy design is unnecessary. Also, that is not to say that general policy attitudes do not matter in policy calibration. They do. Nevertheless, their effect is not deterministic but rather probabilistic and might be very selective.

The key practical question is “which policy core beliefs affect which secondary aspects” (Kammermann and Angst 2021, 763). Research shows that the interaction of various factors usually makes a difference conjointly. For example, Liu, Qin, and Zhang (2023) showed that providing information about the risks of a proposed policy has different effects depending on the type of policy instrument (direct versus indirect), how the information is framed (positively or negatively), and the motivations for the introduction of the proposed policy (personal versus public). Particular combinations of these parameters have different effects.

Consequently, the validity of the five different theories described above is likely to differ according to context. For instance, in cases when the individual is very familiar with the situation, the link between attitudes to abstract and concrete policy is likely to be more direct. This may explain, for instance, the relatively high correlation between support for general and calibrated measures taken against COVID-19. Here, the theories assuming hierarchical structure are likely to be more plausible than in cases where the policy issue is more distant to people’s knowledge or in issues upon which people have only vague knowledge. It can also be assumed that for policy issues with a moral aspect (Mooney and Schuldt 2008), the link between attitudes to general policies and calibrated policies is tighter than it is for less value-laden policy problems. Instead of assuming that one particular theory fits to all circumstance, practitioners should consider different theories and their assumptions.

From a practical point of view, while disagreement about “core beliefs” might be unimportant in some cases, it matters in others. Disagreements about general principles of policy design (e.g. how much regulation a policy should include) are not likely to dissipate later in the policy design process. If a “sufficient consensus” on general policy design (e.g. the primary mechanism for delivery of services) is needed for policy success, efforts to achieve consensus should be made early on. Simply providing information is not likely to do that. More “collaborative and consensus-building methods aimed at challenging stakeholders’ mental models” (Bianchi 2022, 408) should be employed in this situation.

Little is known about the opposite situation, i.e. whether (and how) a policy based on unpopular general principles can become more acceptable when it becomes calibrated. As noted in the introduction to this article, it is not uncommon for people to accept calibrated policies that do not align – at least at first sight – with their views of the general policy design. Policy scholars usually suggest that too much disagreement on basic principles will eventually lead to policy failure. It is still unclear how much initial disagreement can be tolerated and overcome by subsequent discussion and information sharing about specific policy parameters. In their study, Jagers, Matti, and Nilsson (2017) showed that stakeholders’ acceptability of environmental policy instruments tends to increase proportionally to their experience with them. In some cases, public acceptability of general projects was relatively low before implementation, but then acceptance increased as the experience of the project grew. However, there can be no guarantee that a policy will be accepted. Acceptability is conditional on factors such as personal experience with or with the view that the decision-making process was fair.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the European Regional Development Fund project “Beyond Security: Role of Conflict in Resilience-Building” (Reg. No.: Z.02.01.01/00/22_008/0004595).

Notes

1 Policy acceptability thus might be contrasted with policy acceptance, i.e., attitudes to policies that are already in place. While both acceptability and acceptance refer to attitudes, the term policy support is sometimes used to refer to the behavioral dimension of public policies (Kojo et al. 2022). In other words, policy support denotes how people actually behave (e.g., active protests against proposed policies). Concepts of acceptance, acceptability, and support are thus closely related but not synonymous.

2 Ninety percent of respondents agree that there is a need for an institution that would take administrative burdens off school principals and enable them to become pedagogical leaders.

3 It should be noted that the theories summarized below sometimes employ concepts other than policy attitudes. While Zaller, Hurwitz, and many others focus on the concept of attitudes, ACF centers on beliefs, Sniderman and Tetlock focus on preferences, and Carstensen simply talks about ideas. Distinguishing these related terms is not easy because they overlap and differ mostly in their level of abstraction (Maio et al. 2003). It is beyond the purpose of this article to untangle the perplexity of these concepts, but we acknowledge that different usage of these concepts might have consequences for the assumptions of theory.