https://orcid.org/0000-0001-5674-2343
Figures
Abstract
In democracies, public support for policies is crucial to their legitimacy, and the absence can impede necessary reforms, which are needed to keep most disastrous effects of climate change at bay. This study emphasizes the role of policy goal communication as an important but often overlooked dimension of climate policy discourse, arguing that how policy proposals are linked to their intended goals during political debates directly influences public support. We present findings from a novel large-scale (n = 5,655) survey with an embedded randomized experiment that systematically manipulates the type of goal communication and the level of policy goal ambition. Unlike previous studies, the expected policy effectiveness was generated and agreed upon through an iterative process of expert elicitation to provide respondents with a most accurate statement. The results highlight that the presentation of information on policy effectiveness as an inherent element of policy design and not as the larger context in which the policy is proposed significantly increases support for highly ambitious policies renewable energy policies. This study implies that policymakers seeking to promote ambitious climate policies should focus on directly linking proposed policies with the goals these policies should reach.
Citation: Brückmann G, Stadelmann-Steffen I (2026) Policy goal communication increases support for ambitious renewable energy policies. PLOS Clim 5(2): e0000823. https://doi.org/10.1371/journal.pclm.0000823
Editor: Valerio Lembo, CNR: Consiglio Nazionale delle Ricerche, ITALY
Received: July 11, 2025; Accepted: January 7, 2026; Published: February 25, 2026
Copyright: © 2026 Brückmann, Stadelmann-Steffen. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The replication data and code for this study are deposited on the Harvard Dataverse https://doi.org/10.7910/DVN/QPP2BU.
Funding: This work was supported by the Swiss Federal Office of Energy (https://www.bfe.admin.ch/bfe/en/home.html) as part of the SWEET (https://www.bfe.admin.ch/bfe/en/home/research-and-cleantech/funding-program-sweet.html) consortium EDGE (https://www.sweet-edge.ch) (to ISS). The authors bear the sole responsibility for the conclusions and the results presented in this publication. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
As the current barriers to limiting climate change are not geophysical but political [1], and so far, climate policies resulting in substantial emissions reductions often include renewable energy policies [2], the transition to renewable energy systems is a climate policy priority. Many countries around the world have committed to limit climate-relevant emissions through international agreements [3]. They are now setting renewable energy goals [4] and need to implement policies to achieve fully renewable energy systems [5]. It is widely acknowledged that, without at least a certain degree of public support, policies are difficult to implement successfully in democracies, either because their introduction poses electoral risks [6] or because unpopular measures are likely to be rather ineffective [7]. In this context, recent research dealing with policy support has emphasized the multidimensional nature of political measures and individual preferences, and has focused on how specific policies and their design affect policy support [8–10]. The main argument of this literature is that policy measures will typically include elements that individuals like and others they dislike [11], sometimes generating trade-offs. Hence, if policy measures are designed in the right way, popular elements of that policy could compensate for less popular ones and eventually facilitate popular support [12].
While recognizing the significant role of policy design in garnering policy support, this study aims to underscore another essential yet often overlooked element that potentially influences individuals’ opinion formation and, consequently, their eventual support for policies – namely, the role of policy goals. We argue that in scientific and public debates, policies are surprisingly often discussed quite independently of their ability to reach specific goals. This is particularly the case in the field of energy policy, where public discourse typically revolves around the costs of these measures [13,14]. An illustrative example from Switzerland was the campaign in the context of the rejected CO2 act in June 2021, where opponents used claims such as “car driving only for the rich” [15], while the government advertised the act based on financial calculations and arguments showing that the measures would not cost much for households and the economy [16].
The disconnect between policies and policy goals is remarkable, since the essence of policymaking is policy effectiveness [17], that is, designing and implementing political measures to achieve government goals. At the same time, we observe, for example, concerning climate change policy, that policies, once enacted, often do not align with policy goals [18,19], which is one reason why nations lag behind their climate targets. Moreover, theoretically, there are good reasons to believe that information about policy goals affects opinion formation. On the one hand, similar to the policy design perspective, it can be argued that if policy goals are not explicitly communicated or debated, people suffer from an information deficit [20]. On the other hand, a policy goal can define the framing of a policy [21] and serve as a mental anchor in the opinion formation process [22]. An ambitious goal could create an ambitious reference point and increase support for policies that achieve this goal.
In this context, the central argument of this study posits that the way policy proposals are connected to their intended and achievable goals during political debates or campaigns substantially influences their level of backing (for a related argument regarding nationally set emission reduction targets within the framework of the Paris Agreement, see [23]). The primary objective is to examine, theoretically and empirically whether and how the communication of the policy goals shapes individual policy support.
This study presents findings from a novel survey experiment, in which we systematically manipulate the way policy goal information is communicated, as well as the level of goal ambition. We hypothesize that individual support for ambitious policy proposals will be greater if ambitious policy goal information is provided (Hypothesis 1). We further expect that individuals who learn that their preferred policy proposal is not effective in reaching the goal will update their preference and accept more ambitious policies (Hypothesis 2). Instead of focusing on individual policy instruments, we focus on policy mixes, as usually a combination of several policy instruments is needed to solve complex problems such as climate change [24]. Henceforth, we use the term “policy proposals” for policy proposals consisting of a mix of multiple policy instruments.
We propose Switzerland as a suitable study country because policy making is highly dependent on direct-democratic decisions, which makes public support crucial. Furthermore, Swiss citizens regularly express their opinions on policy proposals, which is likely to increase the validity of the data in the survey experiment [13,25].
Our study contributes to the literature in at least two respects. First and foremost, to the best of our knowledge, no study has assessed how different types of policy goal communication affect individuals’ support for corresponding policies. Although it is widely acknowledged that the lack of public support poses a significant obstacle to the adoption of effective climate policies [26–29], scientists disagree on whether the communication of specific targets or deadlines for climate policies [30] is beneficial for adopting these policies. We theoretically develop and empirically test a differentiated conceptualization of how policy goals could be communicated and integrated into a public debate. We distinguish a) between a situation where goal information is considered an inherent element of a policy (henceforth, the Policy+Goal treatment) and a situation where policy goals form the context in which policy preferences are built (other treatment groups) and b) between different levels of goal ambition. Moreover, we analyze whether respondents, when receiving feedback that their preferred policy does not reach the goal, update their policy preference or their goal. Incorporating the role of policy goals into opinion formation on policies may not only be relevant for future referendums and popular initiatives in Switzerland, but also extend its significance beyond the Swiss context. This is due to the increasing importance of energy policies in tackling climate change [31], and the global surge in direct democratic practices [32].
Second, we present a novel interdisciplinary procedure to develop a realistic information treatment in survey research. Energy modelers and energy economists actively supported us in a multi-step reiterative elicitation process to formulate policy mixes in line with current climate-related pledges and differently ambitious policy goals to not deceive respondents. Whereas estimating or even predicting policy effects evidently includes high levels of uncertainty, we argue that this procedure allowed us to develop information that is as realistic as possible for integration into our experiments. Although correct information is important from an ethical perspective, it also contributes to greater external validity in the experimental study.
Theoretical background
State of research - The role of policy goals in public opinion formation
Contemporary societal challenges in various policy domains, such as climate change mitigation and health and welfare state reforms, necessitate state intervention. However, a crucial obstacle to successful policy implementation and change lies in the lack of popular support [33]. Not only is a certain degree of policy support needed in democratic policy-making, but it also facilitates the subsequent viability and effectiveness of the measure [7]. In this context, it has been suggested that conveying how effectively a measure can reach its intended objectives may serve as a way to increase support for a policy prior to its implementation [33]. When the public perceives that policy goals lead to favorable outcomes and are feasible through specific measures, they are more inclined to support such policies. However, while Reynolds et al.[33], in their comprehensive literature review, mainly focused on studies on health and environmental policies, and identified a small but overall positive effect on policy support, theoretical arguments and empirical evidence regarding the impact and mechanisms of communicating policy goals on policy support remain inconclusive. Furthermore, it is unclear what constitutes the most effective approach to conveying a policy’s ability or inability to achieve specific goals [34].
The subsequent subsections examine three different strands of literature that shed light on the significance of policy goals in garnering support for policies. Subsequently, we apply these approaches to the case of renewable energy goals and policies to formulate our theoretical expectations.
Goal information to reduce an information deficit.
A prominent assumption in research on policy support is that a “knowledge deficit” can hinder public support for policy measures [20,35]. Especially in the field of environmental and climate change policies, which typically involve short-term and visible costs but mostly long-term and uncertain benefits, a lack of understanding of how these measures work and, therefore, generate future benefits may be problematic for policy support [11,36].
More specifically, it can be argued that individuals who lack awareness of the goal a policy is aimed at or the goals a policy measure can achieve face a knowledge deficit. They lack relevant pieces of information about the policy. Therefore, providing information about a policy’s intended and achievable goals can reduce the knowledge deficit, and consequently increases policy support.
Policy goals in the politics of trade-offs.
Another strand of research emphasizes that opinion formation on policies involves multidimensional decisions [37] and trade-offs [8,11,12,38,39]: Individuals will probably like some elements of a policy, while disliking others. Eventually, when forming an opinion about a policy, they need to weigh the perceived costs and benefits.
These contributions have in common the expectation that policy design, along with other factors such as socio-demographics, ideology, or framing, affects how individuals perceive and think about a policy. In this context, it is generally assumed that the more direct and visible the benefits, for example, as a result of the type of revenue recycling, the higher public support [40–45]. These perceived benefits are important to compensate for unpopular elements of a policy, such as costs or restrictions [46].
According to this perspective, policy goals and information about policies’ ability to reach goals can be considered a relevant dimension of a policy. If discussed and communicated to citizens, the goal dimension of policies is likely integrated into the multidimensional opinion-forming process. In weighing advantages and disadvantages, policies that explicitly address the concerns and objectives of individuals, especially if they are expected to be successful in achieving those targets, can reinforce the affirmative view.
Policy goals as framing and mental anchor.
Drawing on issue framing theory [21,47], issue frames emphasize a policy’s distinct yet potentially relevant aspects, guiding citizens to think about specific issues from particular perspectives. By highlighting the specific features of a policy, such as its positive or negative effects, frames can mobilize citizens to consider and engage with the issue in a particular manner. While research has identified some effective framing strategies [48–50], it has also been emphasized that “simple re-framing” is not a panacea to increase climate policy support [51].
We argue that approaching the process of opinion formation on a policy from the perspective of the goals that the policy should or can achieve can serve as an influential—yet largely overlooked—frame. When a specific policy proposal is presented as a solution to achieve a particular goal, it can generate a positive perception of that policy. Generally, policy goals as a framing mechanism have the potential to stimulate a goal-oriented opinion-formation process, which differs from the more cost-focused approach typically associated with policy design. This argument aligns with the literature on policy credibility, which holds that actors are more likely to change their behavior in accordance with a goal when they believe policymakers’ pledges will translate into concrete action [52,53]. From this perspective, it can be expected that communicating information about policies, including the goals that the measures are expected to achieve, will contribute to policy credibility and thus increase the likelihood of individuals supporting them.
Furthermore, in line with the concept of mental anchoring [22,54], ambitious policy goals can create a more ambitious reference point and help individuals to solve commitment problems [23]. In this context, Tingley and Tomz [23] found that international pledges, namely the Paris Agreement, increased support for emission reduction policies among US citizens.
Theoretical expectations
Based on these insights from existing literature, we formulate two hypotheses on how the policy goals dimension is expected to affect policy support. We focus on goals with respect to renewable energy production at different levels of ambition and different policy mixes that may or may not achieve these different policy goals. Hence, increased policy support in that context means that individuals support policy mixes that are able to reach (more) ambitious renewable energy targets.
Based on the previous discussion, we distinguish between two facets of the policy goal dimension: policy goals and policy effectiveness. The term policy goals encompasses both the purpose and direction of policy change. In contrast, policy effectiveness pertains to the extent to which a specific policy measure is able to achieve its intended goal successfully.
The first hypothesis relates to a policy goal effect, i.e., the fact that opinion formation takes place in a situation where the policy goal information is available compared to a situation without such information:
Hypothesis 1. Individuals are more likely to support ambitious policies if ambitious policy goal information is provided.
Generally, we expect that support for ambitious policy mixes is stronger if the opinion-formation process about these policy proposals involves information about policy goals, that is, the ambition of renewable energy production. This expectation is backed by several of the aforementioned theoretical approaches. Initially, the formation of opinions without information about the policy objective can be considered a situation with information deficit [20,35]. Therefore, individuals may underestimate the benefit side of a policy proposal [11]. Hence, information linking policy proposals to policy goals and policy effectiveness closes the information deficit, adds an additional “benefit” to the multidimensional decision-making, and leads to a more goal-oriented framing of opinion formation process [21]. Moreover, showing that a policy relates to previously formulated goals can contribute to policy credibility and thus facilitate actors engagement with the policy [52,53].
However, it is likely that the degree to which information about policy goals enhances support for ambitious policies depends on how ambitious those goals are. Two rationales underpin this assertion. Firstly, a framing that builds on ambitious policy goals is likely to produce a more pronounced positive influence on ambitious policy support than framing that emphasizes less ambitious objectives. Secondly, to strengthen support for ambitious policies, the mental anchor [22,54] must be ambitious.
The second hypothesis revolves around information concerning policy effectiveness, which relates to the ability of policies to accomplish their intended goals:
Hypothesis 2. The provision of information that indicates a lack of policy effectiveness increases the support for more ambitious policies.
Drawing on the knowledge deficit model [20,35], we posit that information on a potential mismatch between a selected policy and a goal, indicating a lack of policy effectiveness, could amplify support for more ambitious policy proposals. According to this perspective, mismatch information reveals to individuals that the current policy does not meet the goal and underscores the need to implement more ambitious measures. Naturally, a mismatch between an ambitious policy goal and a preferred policy mix could also be addressed by reducing the level of goal ambition. However, given the general recognition of the need to increase renewable energy generation [55], we expect people to be more likely to endorse more ambitious policies when their preferred policies fall short of the policy objectives, rather than reducing the ambition of those goals.
Methods and data
Ethics statement
The institutional ethics committee of the ETH Zurich has approved this research without reservation in their decision EK 2022-N-109. This research was designed and implemented considering the Principles and Guidance for Human Subjects Research to ensure (potential) participants’ well-being.
In the postal invitation letter, we explicitly stated that potential participants have to consent to take part in this survey. We also mentioned that survey participation was voluntary. At the beginning of the online survey, see S1 Table, we repeated that any answers would be treated strictly confidentially and anonymously. We provided this information in four languages so all invitees understood it, and they could give informed consent if they wished to do participate in the survey experiment.
We randomly selected from the Swiss population and abstained from inviting minors or people aged 75 or above to exclude vulnerable groups. We did not target any other vulnerable group. As S2 Table shows, we also did not expose a larger survey burden on other vulnerable groups, for example, those from less affluent households.
We ensured no deception by co-creating the energy policies and the goals they can reach with multiple experts. Also, at no other point of the survey have participants been deceived.
Respondents were compensated by the possibility to participate in a random lottery for cinema vouchers and a tablet. Through survey participation, no harm or impact was inflicted on participants.
Data
To examine our hypotheses, we developed a novel vignette experiment incorporated in a comprehensive survey conducted in Switzerland from August 26 to October 31, 2022.
A stratified random sample based on location quota was obtained from the Federal Statistical Office population register, comprising Swiss residents aged 18 to 75 years (we set the lower age cut-off at voting-age and the upper limit was set not to burden older residents with an online survey). A total of N = 13,500 invitations were sent out via postal mail, inviting recipients to participate in an online survey on the future of Swiss energy. We hosted the online survey on qualtrics.com, and the survey started with a question to obtain informed consent. Detailed information on the survey can be found in S1 Table and in the pre-analysis plan (which can be found on OSF registries https://doi.org/10.17605/OSF.IO/JNS4H, but as only some sections of it are relevant to this manuscript, an excerpt can be found with the replication files on https://doi.org/10.7910/DVN/QPP2BU).
The study’s findings are derived from the 5,655 respondents who participated in the experiment. Comparison with official population statistics reveals that the sample aligns closely with the Swiss population in terms of age, gender, and education, but exhibits a skew toward high-monthly household income groups (see S2 Table). Therefore, we conclude that our sample should have high external validity, especially X-validity [56], for the Swiss population.
The policy goals experiment
The overview of the experiment is depicted in Fig 1. First, we randomly assigned respondents to the No Goal condition (as a control group) or one of the three treatment groups (see Fig 1):
- No Goal: No goal information is ever displayed. Respondents can select their preferred policy proposal without any goal-related information. Serves as a control group.
- Policy+Goal: Respondents can select their preferred policy proposal which is presented alongside the respective policy goal the different policy mixes are each likely to reach.
- Goal Assigned: Respondents are randomly assigned to and informed about one of three policy goals with different levels of ambition (low, medium, or high) before they can select their preferred policy proposal. They receive feedback in case of a mismatch between the goal and the selected policy proposal.
- Goal Selected: Respondents have to select one of three policy goals with different levels of ambition (low, medium, or high) before they can select their preferred policy proposal. They receive feedback in case of a mismatch between the goal and the selected policy proposal.
Diamonds indicate randomization into treatment groups, light gray fields are displayed to respondents, and dark gray fields prompt respondents to answers.
The likelihood of assignment to the Goal Assigned or the Goal Selected treatment groups is p=1/3, and p=1/6 for assignment to either the Policy+Goal treatment group or the No Goal (control) group. The higher assignment probability for the Goal Assigned and Goal Selected groups was deliberate to enhance the number of observations in groups that could receive feedback on goal misalignment, allowing us to test Hypothesis 2.
In the following, the steps of the experimental design are described for each of the different experimental groups.
Step 1 Initial information on policy goal
For participants, the beginning of the experiment was the presentation of a common introductory text, as shown in Figs 2–4. Subsequently, the different experimental groups received varying information on the policy goals that Switzerland should achieve.
Grey boxes on the right were not displayed to respondents but added for clarity through alignment with the overview of the experimental design.
Grey boxes on the right were not displayed to respondents but added for clarity through alignment with the overview of the experimental design.
Grey boxes on the right were not displayed to respondents but added for clarity through alignment with the overview of the experimental design.
For members of the two groups Policy+Goal and No Goal, no further information on policy goals was provided at this stage (Fig 2). However, for comparability between experimental groups, a goal ambition level was assigned for these groups as well, but never displayed to the participants.
The experimental group Goal Assigned was assigned and informed about one of the three goal statements in Table 1. These goals describe different levels of goal ambition, as indicated in the right column of the table. In the example provided in Fig 3, the medium ambition goal is displayed to respondents.
In the treatment group Goal Selected, after the introductory text, participants were asked to select one of the three policy goals used in this survey experiment (see Table 1) which is exemplified in Fig 4: “Several concrete goals are discussed for this. Thinking of the electricity sector, which goal do you think Switzerland should pursue?”
Hence, while assignment to this group was random, their level of goal ambition is self-selected and, therefore, not random. Consequently, for the analyses related to Hypothesis 1, where we experimentally test the role of goal communication and goal ambition, we exclude this group.(Nonetheless, the results on policy support for this group are presented in S3 Fig, showing that individuals who selected a highly ambitious goal indeed also supported more ambitious policies.)
Step 2 Baseline policy decision - selecting the preferred policy proposal
For the baseline policy decision, respondents were randomly shown three out of the six policy mixes, and had to indicate which one they would support most (see Fig 1): “Various measures can be taken to achieve this goal. We present three packages of measures below. Please indicate which one you would support most. The packages of measures should be financially balanced; any deficits in the national budget will be financed through taxation of energy consumption. Unless explicitly stated otherwise, the measures are designed to last until 2050.”
This was identical for respondents from all experimental groups (Fig 5) but those from the Policy+Goal treatment group. The latter received the additional information about what energy goal the respective policy proposal would likely achieve (Fig 6).
Grey boxes on the right were not displayed to respondents but added for clarity through alignment with the overview of the experimental design.
Grey boxes on the right were not displayed to respondents but added for clarity through alignment with the overview of the experimental design.
The six policy mixes used in this study (given in full in the boxes below) have been carefully selected and defined through expert elicitation. The expert assessment was conducted through an iterative process involving a diverse group of Swiss energy researchers with multidisciplinary backgrounds, including energy economics and modeling. The interdisciplinary approach aligns with the growing recognition of social science-led “all hands on deck” research on energy and climate topics [57, p. 3].
The aim was to ensure that each of the three policy goals was met with two policy mixes each. Within each goal ambition level, one policy mix emphasized market-based incentives such as purchase price guarantees, contributions to investment costs, or electricity surcharge, while the other focused on regulation, including requirements for solar PV on buildings, obligations for electricity suppliers, and designated areas for open-space solar PV. With the two different policy mixes per level of ambition, we wanted to provide respondents with clearly described policy mixes and ensure that our treatment effects were not driven by the instrument type. For each policy mix, expert elicitation indicated an overarching consensus that the policy mix was likely to achieve the intended goal but likely not the more ambitious goals(s). Thus, through this extensive co-creation procedure, we ensured to investigate realistic policy mixes and goals, maximize external validity and minimize deception of respondents.
Below, we present the policy options in descending order of ambitiousness. In the survey experiment, the three (out of the six) randomly displayed policy mixes were displayed in a random order to the respondents, while ensuring at least one policy mix reached the assigned or selected policy goal. For respondents with low or medium policy goals, it always suffices to select at random (with equal probability) three out of six policy proposals to ensure that at least one achieves the policy goal. However, this is not the case for participants with a high goal ambition. Thus, in the randomization process we ensured that there was always at least one policy option displayed that, according to expert elicitation, was likely to achieve that high goal (technical details on randomization with restrictions can be found in the S1 Text). Moreover, in the survey, no labels (such as, “High ambition – Emphasis on incentives”) were displayed. Parts depicted in light gray were only displayed to the treatment group Policy+Goal.
Two policy mixes likely achieve the high-ambition policy goal
High ambition – Emphasis on incentives: To promote the production of domestic renewable electricity, there is a purchase guarantee for renewable electricity. This means that anyone who builds a plant for the production of domestic renewable electricity today is guaranteed at least CHF 0.12 per kWh for the electricity produced from it during the first 10 years. In addition, the construction of such a plant is supported with an investment contribution amounting to 50 % of the investment costs. This support is financed through the taxation of electricity consumption (via a surcharge on grid usage). Finally, there are requirements in the area of renewable electricity: from 2030 there will be a solar obligation on new buildings and renovations, and from 2050 on all buildings. Electricity suppliers will also be obliged to have at least 80 % domestic renewable electricity in their electricity mix from 2030. Solar installations will be permitted on open areas and in the Alpine region. Based on the assessment of various researchers in this area, this combination of policy measures is sufficient to achieve the goal of “100 % of one year’s electricity consumption is covered by domestic renewable production from 2040 onwards”.
High ambition – Emphasis on regulation: The construction of a plant for the production of domestic renewable electricity is supported with an investment contribution amounting to 40 % of the investment costs. This support is financed through the taxation of electricity consumption (via a surcharge on grid usage). Finally, there are requirements in the area of renewable electricity: From 2030, there will be a solar obligation on new buildings and renovations, and from 2040 on all buildings. Electricity suppliers will also be required to have at least 85 % domestic renewable electricity in their electricity mix from 2030. Solar installations will be allowed on open areas and in the Alpine region. Based on the assessment of various researchers in this area, this combination of policy measures is sufficient to achieve the goal of “100 % of one year’s electricity consumption is covered by domestic renewable production from 2040 onwards”.
Two policy mixes likely achieve the medium ambition policy goal
Medium ambition – Emphasis on incentives: To promote the production of domestic renewable electricity, there is a purchase guarantee for renewable electricity. This means that anyone who builds a plant for the production of domestic renewable electricity today is guaranteed at least CHF 0.08 per kWh for the electricity produced from it during the first 10 years. In addition, the construction of such a plant is supported with an investment contribution amounting to 40 % of the investment costs. This support is financed through the taxation of electricity consumption (via a surcharge on grid usage). Finally, there are requirements in the area of renewable electricity: from 2030 there will be a solar obligation on new buildings and renovations. Based on the assessment of various researchers in this area, this combination of policy measures is sufficient to achieve the goal of “100 % of one year’s electricity consumption is covered by domestic renewable production from 2050 onwards”.
Medium ambition – Emphasis on regulation: In order to promote the production of domestic renewable electricity, the construction of such plants is supported with an investment contribution amounting to 40 % of the investment costs. This support is financed through the taxation of electricity consumption (via a surcharge on grid usage). Finally, there are requirements in the area of renewable electricity: From 2030, there will be a solar obligation on new buildings and renovations, and from 2050 on all buildings. Electricity suppliers will also be required to have at least 80 % domestic renewable electricity in their electricity mix from 2030. Based on the assessment of various researchers in this area, this combination of policy measures is sufficient to achieve the goal of “100 % of one year’s electricity consumption is covered by domestic renewable production from 2050 onwards”.
Two policy mixes likely achieving the low ambition policy goal
Low ambition – Emphasis on incentives: In order to promote the production of domestic renewable electricity, the construction of such plants is supported with an investment contribution amounting to 40 % of the investment costs. This support is financed through the taxation of electricity consumption (via a surcharge on grid usage). In addition, there are requirements in the area of renewable electricity: From 2040, there will be a solar obligation on new buildings and renovations. Based on the assessment of various researchers in this area, this combination of policy measures is sufficient to achieve the goal of “90 % of electricity consumption in one year is covered by domestic renewable production from 2050 onwards”.
Low ambition – Emphasis on regulation: There are requirements in the area of renewable electricity: From 2040, there will be a solar obligation on new buildings and renovations. Electricity suppliers will also be required to have at least 75 % domestic renewable electricity in their electricity mix from 2030. Based on the assessment of various researchers in this area, this combination of policy measures is sufficient to achieve the goal of “90 % of electricity consumption in one year is covered by domestic renewable production from 2050 onwards”.
Step 3 Evaluation of feedback on goal misalignment
In the final phase, if participants in the Goal Assigned or Goal Selected groups had chosen a policy mix in Step 2 that did not align with their assigned or chosen goal, they received feedback regarding this mismatch. Therefore, for those respondents, Fig 7 zooms into Step 3 of the experiment for them and presents the questions and respondents’ response options.
Light gray fields are displayed to respondents, dark gray fields prompt respondents to answers, and arrows with unfilled tip indicate conditional flows.
Operationalization of the hypotheses
To evaluate Hypothesis 1, we assess the probability of supporting ambitious policy mixes among the No Goal, Policy+Goal, and Policy Assigned groups. We do not use respondents who did self-select into their goal (Goal Selected), to keep the randomized nature of this experiment. For the examination of Hypothesis 2, only respondents from the Goal Assigned and Goal Selected group are used who received feedback about a discrepancy between their preferred policy and the assigned or selected goal. Table 2 summarizes the utilization of treatment groups and outlines the analytical procedures applied to test our hypotheses.
Analysis
Our dependent variable is whether or not respondents selected a policy proposal in alignment with the most ambitious goal, as it provides the best proxy for highly ambitious climate policies. Therefore, we perform our analysis using logistic regression models, due to the binary nature of the dependent variables.
We use R for computation and cite relevant packages [58–71], following recent recommendations [72].
Due to the randomized nature of our experimental data and the balancing between experimental groups (as demonstrated in the Tables included in S2 Text), we follow our pre-analysis plan, and do not use any covariates, as we are conceptually interested in Average Treatment Effects (ATE). (However, we test for heterogeneous treatment effects in S5 Text.)
Results
Among all respondents from the No Goal, Policy+Goal and Goal Assigned groups, we first observe their policy choices, based on which level of goal ambition the selected policy proposals are likely to achieve. As depicted in S5 Fig, which presents on the left the support for policies aligned with the high goal for each treatment group (regardless and independent of their goal), there is no general and statistically significant increase in the likelihood of opting for a high-ambition policy mix in the two treatment groups Policy+Goal and Goal Assigned compared to the No Goal (control) group. Only on the 10 % level is the random assignment to the Policy+Goal treatment leading to more high-ambition policy choices than No Goal.
As theorized in Hypothesis 1, one reason might be that different goals could communicate different messages to respondents. We proceed to evaluate our hypothesis, which considers the degree of goal ambition.
Hypothesis 1: Policy support conditional on the level of goal ambition
Fig 8 provides evidence that a high goal effectively increases the likelihood of selecting an ambitious policy among all groups, significantly in comparison to a low goal, p = .0000. Most clearly, informing about the ambitious goal a policy can achieve along with the description of the policies (Policy+Goal) leads to a significantly higher likelihood that ambitious policy proposals are selected, as the results of a linear hypothesis test with = 6.855 and p = .0088 demonstrates. These results confirm our theoretical expectation that it is not only relevant that goal information is goal information is provided but also the message it contains. More specifically, in order to generate stronger support for ambitious policy proposals, an equally ambitious goal information message is required whereas information about a less ambitious goal does not make a difference.
The associated regression is column 1 of S13 Table.
It is important to note that strictly speaking, we cannot directly compare the predicted probabilities between the groups with the high goal and the others. This follows from the slightly higher probability of seeing high-ambition policies for respondents with the high-ambition policy goal (as detailed in the Methods and data Section under Step 2), which ensures that all respondents in this group encounter at least one policy aligned with their respective goal. In contrast, randomization did not differ between the low and medium goal groups. We therefore present the predicted probabilities grouped by goal ambition. Hence, the observable differences between the experimental groups for the High Goal condition, as seen in Fig 8, are not the result of varying randomization but actual treatment effects. The High Goal among the Policy+Goal group is significantly higher than the High Goal among the Goal Assigned group, as a linear hypothesis test with = 7.0827 and p = .0078 indicates, and significantly higher than the No Goal group with the same high goal (
= 51.547 and p = .0000). We further tested the robustness of the treatment effect when accounting for the fact that some respondents refrained from choosing any policy proposal (i.e., selecting ‘none’). As Fig 9 shows the probability to chose any policy does not vary between treatment groups, while our main findings also hold when conditioning on choosing any policy (not ‘none’) (see S4 Fig). Moreover, the experimental treatment does not significantly alter the choices for policy mixes with more emphasis on market-based incentives or more regulation (as we detail in S4 Text).
Finally, in the S5 Text, we present exploratory analyses in which we tested heterogeneous treatment effects. These results confirm our main conclusions and include political self-placement, educational attainment, climate change priorities and trust in science as well as gender, age and income (see S5 Text). For trust in science, S9 Fig demonstrates that individuals with low trust in science are generally less likely to endorse ambitious policy mixes but information on policy effectiveness increases support for ambitious policies even within this sub-group. Moreover, we consider the finding that treatment effects do not significantly vary between educational groups as an indication that our experiment was understandable also for lower skilled respondents.
Hypothesis 2: Feedback about mismatch between preferred policy and policy goal
We examine how individuals responded to information that their preferred policies did not align with their (assigned or selected) policy goals. This analysis focuses exclusively on respondents in the Goal Assigned or Goal Selected treatment groups who were assigned or selected a high or medium goal, as those groups could potentially receive feedback on a mismatch of goal and policy. In total, 668 respondents in the Goal Assigned or Goal Selected treatment groups received feedback about such a mismatch (descriptive information on feedback can be found in S12 Table).
Initially, it can be seen from Fig 10 that the reaction to the feedback does not differ between the two treatment groups, that is, whether respondents were initially assigned or selected their goal (statistically the differences between the groups are not significant with N = 668, = 1.88, df = 2 and p = .390). Overall, almost 40 percent (38.9 %) of respondents, upon learning of a mismatch between their preferred policy and the goal, chose to adjust their policy preference (Fig 10). Furthermore, 32.9 % decided to change the policy goal, while 28.1 % opted not to modify either the selected policy or the policy goal. This distribution of reactions differs from chance (N = 668,
= 11.689, df = 2 and p = .0028).
327 respondents stem from the Goal Assigned and 341 respondents from the Goal Selected treatment group.
Consequently, a relative majority preferred to revise the policy rather than the goal (N = 480, = 3.3333, df = 1 and p = .0679). The probability of accepting the discrepancy without modifying either the policy or the goal is significantly different from altering the policy (N = 448,
= 11.571, df = 1 and p = 0.007).
Moreover, among the 257 respondents who decided to modify their policy choice after feedback, the majority (158 or 61.2 %) selected policy mixes in the second round that indeed reduced the discrepancy between the ambition of their goal and the selected policy proposal (depicted in S8 Fig). In general, these findings support Hypothesis 2.
However, almost one-third of respondents opted to adjust their policy goals following feedback on mismatches (see Fig 10). Given they only received feedback if goals where more ambitious than selected policies, this implies goals might have been decreased. Indeed, Fig 11 shows the likelihood of these respondents to lower their goal ambitions after feedback. We find that respondents in the Goal Assigned group, who were given a high target, were significantly more likely to reduce their goals. This suggests that respondents tended to decrease their goals if the assigned targets seemed too difficult to meet with their preferred policy selections. This observation underscores that setting goals reliant on unpopular policies might lead individuals to revise or abandon these goals entirely.
There are 46 respondents in the Goal Assigned treatment group that decreased their policy goal ambition and 56 in the Goal Selected treatment group. The other individuals kept (40 or 39 % in Goal Assigned and 51 or 44 %in Goal Selected) or increased (17 or 17 % in Goal Assigned and 9 or 8 % in Goal Selected) their goal ambition.
Overall, when comparing both, goals achieved and ambition levels from selected policies before and after reactions to feedback, we observe an increase in goals achieved by 1.7 percentage points, from 57.8 % to 59.4 % (when not considering goal changes, and 2.9 percentage points when also considering altered goals, see S14 Table). Moreover, the share of policies aligned with the most ambitious goals increases by 1.3 percentage points from 35 % to 36.3 % through the feedback reactions (see S15 Table). The (combined) share of policies selected in alignment with at least the medium goal increases by 1 percentage point from 58.1 % to 59.1 % (as depicted in S16 Table, see also S2 Fig (before feedback) and S7 Fig (after feedback)).
Discussion and conclusion
This study aimed to investigate the influence of policy goal information on individual policy support, highlighting that the dimension of policy goal has been overlooked both in public debates and in opinion formation research. Empirically, we presented results from a novel survey experiment that systematically manipulated the type of goal communication and the level of goal ambition within the context of renewable energy policies in Switzerland. Our analyses were guided by two theoretically derived hypotheses. The first hypothesis stated that support for ambitious policy proposals would be higher if information on ambitious policy goals was available during opinion formation. Moreover, the second hypothesis suggested that providing individuals with the information that their preferred policy proposal does not reach the policy goal would increase their support for more ambitious policy mixes.
The analyses conducted in this study reveal that the communication of policy goals can affect the support for ambitious policies, lending empirical support to Hypothesis 1. More specifically, including information about the objectives the policy can likely achieve, i.e., highlighting their effectiveness, increases the likelihood that citizens choose highly ambitious policy packages. The finding implies that the use of policy effectiveness as an inherent element of policy design [73] makes the purpose of a policy more visible and accessible, contributes to higher policy credibility [52,53], and therefore has the potential to increase support for ambitious measures. In contrast, using ambitious policy goals as a frame [21] or mental anchor [22] to define the decision context did not significantly affect the policy preferences of the respondents. At the same time, unlike Tingley and Tomz [23], we also did not find a backfire effect when the goals were very ambitious. We attribute this to the broad Swiss consensus on the overarching goal of renewable energy expansion. Against this background, even the most ambitious goal is unlikely to provoke fundamental opposition, i.e., the kind of reaction that would likely be needed to trigger backfiring effects.
In addition, mixed results were obtained with respect to Hypothesis 2. If informed of a mismatch between policy goals and the preferred policy, respondents express a notable intention to align their policy goals with their policy choices, as indicated by their readiness to modify these choices. Nevertheless, it is important to recognize that this intention does not always materialize. This study does not definitively resolve whether this is due to challenges in evaluating policy ambition or for other reasons.
Our results also bring attention to a significant caveat: a notable portion of individuals (roughly 26%) do not support any policy and do not strive to achieve any renewable energy goal. We observe that a considerable share of respondents did not endorse any policy mix. Furthermore, when respondents became aware of the mismatch between their policy goal and their preferred policies, particularly in the Goal Assigned group, they willingly accepted a reduction in the goal ambition. This illustrates the limits of goal selection as a means of increasing policy support.
The present study is not without limitations. That is, the mismatch information to the respondents was not fully exploited, as not enough respondents in the respective treatment groups missed their policy goal in the first policy choice. Moreover, while we are confident that our study has strong internal validity, the external validity of survey experiments on political attitudes can always be questioned [74]. Future research could investigate whether these patterns also hold in countries without direct-democratic elements, where citizens are not equally used to engaging with and deciding between policy proposals. Moreover, we acknowledge that the goal of increasing domestic renewable energy generation may be relatively uncontested compared to other policy goals. For example, other climate policies such as carbon taxes [27,40], fossil fuel bans [75], or even geoengineering [76] seem to be much more controversial and we cannot be sure that our findings travel to such policy debates. For example, when the goal as such is more conflictual, it may be more strongly integrated in policy communication already, compared to the policy field under investigation in this study. Consequently, the role of goals might be more ambiguous in policy areas with conflicting overarching objectives. Given that such conflictual policy debates might gain in importance in the future, more research is needed to explore whether the relevance of policy goals is more important in some policy domains or discourses than in others.
Nevertheless, our study has broader conceptual and practical implications beyond the specific case we investigated. First, the theoretical framework developed in this study, the survey experiment based on expert elicitation, and the empirical results emphasize the importance of policy goals and their communication in policy making and research. They highlight that the way policy goals and, in particular, the ability of policy measures to effectively achieve these goals are conveyed and discussed can influence the formation of opinions about policies and, consequently, policy support. This opens an avenue for future research to explore whether and how the relevance of policy goals varies between institutional and political contexts. The question of how information about effectiveness affects policy support is clearly relevant beyond energy and climate policy. For example, our approach could also be valuable in other policy areas such as social policy where there is a broad consensus on the need for reform in general, but a lack of majority support for specific proposals. Similarly, further insights would be valuable to understand whether the visibility or tangibility of policy goals for the general public moderates the role of goals in the opinion formation process (see also [77]). Finally, this study’s main implication for policymakers aiming to increase support for ambitious renewable energy policies is to make policies’ capacity to achieve their goals—or, their effectiveness—an inherent element of their communication about these policies.
Supporting information
S1 Table. Survey structure.
While the full survey instrument can be found with the replication files on https://doi.org/10.7910/DVN/QPP2BU, this Table gives an overview of the survey structure.
https://doi.org/10.1371/journal.pclm.0000823.s001
(PDF)
S1 Text. Technical information on the randomization.
Includes additional technical information on randomization to ensure policy proposals presented to respondents reach policy goals.
https://doi.org/10.1371/journal.pclm.0000823.s002
(PDF)
S2 Table. Sample descriptives.
Descriptive comparison between the sample and the Swiss population in terms of socio-demographic characteristics (official Swiss statistics are for the survey year (2021) and for all ages, if not indicated otherwise).
https://doi.org/10.1371/journal.pclm.0000823.s003
(PDF)
S2 Text. Balancing tables.
Includes all balancing tables.
https://doi.org/10.1371/journal.pclm.0000823.s004
(PDF)
S3 Table. Balancing table.
No Goal assignment vs. remaining experimental participants.
https://doi.org/10.1371/journal.pclm.0000823.s005
(PDF)
S4 Table. Balancing table.
Policy+Goal assignment vs. remaining experimental participants.
https://doi.org/10.1371/journal.pclm.0000823.s006
(PDF)
S5 Table. Balancing table.
Goal Assigned assignment vs. remaining experimental participants.
https://doi.org/10.1371/journal.pclm.0000823.s007
(PDF)
S6 Table. Balancing table.
Goal Selected assignment vs. remaining experimental participants.
https://doi.org/10.1371/journal.pclm.0000823.s008
(PDF)
S7 Table. Balancing table.
Low Goal assignment vs. remaining experimental participants.
https://doi.org/10.1371/journal.pclm.0000823.s009
(PDF)
S8 Table. Balancing table.
Medium Goal assignment vs. remaining experimental participants.
https://doi.org/10.1371/journal.pclm.0000823.s010
(PDF)
S9 Table. Balancing table.
High Goal assignment vs. remaining experimental participants.
https://doi.org/10.1371/journal.pclm.0000823.s011
(PDF)
S1 Fig. Initial policy goals and policy choices.
Initially assigned policy goal (not shown to the No Goal and Policy+Goal groups) and initially selected policy goal (for Goal Selected treatment, only) (N = 5,655). The same graph after feedback reactions can be found in S6 Fig.
https://doi.org/10.1371/journal.pclm.0000823.s012
(PDF)
S3 Text. Additional results.
Information on initial goal randomization vs. initial goal selection. treatment effects on the likelihood to choose an ambitious policy package, and first policy goal selection.
https://doi.org/10.1371/journal.pclm.0000823.s013
(PDF)
S4 Text. Additional results.
Information on differences in emphasis within policy mixes.
https://doi.org/10.1371/journal.pclm.0000823.s014
(PDF)
S2 Fig. Descriptive results.
First chosen policy, given treatment status, and assigned or chosen policy goal (N = 5,655). For the No Goaltreatment group, no goal was displayed, ever, while for the Policy+Goal treatment group, the goals that each offered policies could reach were displayed next to the policy options, but no overall goal to be achieved.
https://doi.org/10.1371/journal.pclm.0000823.s015
(PDF)
S3 Fig. Predictions plot.
Predicted probabilities for ambitious policy choice among Goal Selected group. Predicted probabilities for choosing ambitious policy proposals among the respondent who previously selected their desired policy goal. The dependent variable of the logistic regression is a binary indicator, which assumes the value of one, if the first chosen policy achieves the most ambitious policy goal (on the left) and medium or maximum (right panel) ambition. Associated regression tables can be found in S18 Table.
https://doi.org/10.1371/journal.pclm.0000823.s016
(PDF)
S4 Fig. Predictions plot.
Conditional logit version of Fig 8. This logistic regression is conditional on choosing any policy, not none. Again, the dependent variable is a binary indicator, which assumes the value of one, if the first chosen policy mix is of maximum ambition, therefore, achieving the highest policy goal. The associated regression is column 1 of S10 Table.
https://doi.org/10.1371/journal.pclm.0000823.s017
(PDF)
S10 Table. Regression table.
Predicted probabilities for initial policy choices, given any policy is chosen.
https://doi.org/10.1371/journal.pclm.0000823.s018
(PDF)
S5 Fig. Predictions plot.
Predicted probabilities for choosing ambitious policy proposals. The dependent variable of the logistic regression is a binary indicator, which assumes the value of one, if the first chosen policy achieves the most ambitious policy goals (left panel) or either the most or medium ambitious policy goals (right panel). The associated regression models can be found in S19 Table.
https://doi.org/10.1371/journal.pclm.0000823.s019
(PDF)
S11 Table. Regression table.
Predicting initial policy goal selection.
https://doi.org/10.1371/journal.pclm.0000823.s020
(PDF)
S12 Table. Descriptive information on feedback potential.
Distribution of respondents who could receive feedback. Percentages are shares from those respondents who were randomly assigned into the Goal Assigned or Goal Selected treatment group and were assigned to or selected a medium and high goal, and therefore became eligible for feedback, given policy choices not aligning with the goals and not selecting none of the proposed policies.
https://doi.org/10.1371/journal.pclm.0000823.s021
(PDF)
S6 Fig. Policy goals and policy choices after (potential) alterations.
Last policy goal after alterations and initial policy goals (for those who did not receive feedback and for those in the No Goal and Policy+Goal groups).
https://doi.org/10.1371/journal.pclm.0000823.s022
(PDF)
S7 Fig. Descriptive results after (potential) alterations.
Policy choice, after potential alterations in response to feedback about mismatches, given treatment status and originally assigned or selected policy goal.
https://doi.org/10.1371/journal.pclm.0000823.s023
(PDF)
S13 Table. Regression table.
Treatment Effects for Policy Choice as displayed in Fig 8.
https://doi.org/10.1371/journal.pclm.0000823.s024
(PDF)
S14 Table. Descriptive information.
Goal achievement (before and after feedback).
https://doi.org/10.1371/journal.pclm.0000823.s025
(PDF)
S15 Table. Descriptive information.
Policy choice (before and after feedback).
https://doi.org/10.1371/journal.pclm.0000823.s026
(PDF)
S16 Table. Descriptive information.
Aggregated policy choice (before and after feedback).
https://doi.org/10.1371/journal.pclm.0000823.s027
(PDF)
S17 Table. Regression table.
Treatment effects for the second decision displayed in Fig 11.
https://doi.org/10.1371/journal.pclm.0000823.s028
(PDF)
S18 Table. Regression table.
Initial policy choices among Goal Selected treatment group (Predictions based on the first two columns are displayed in S3 Fig.)
https://doi.org/10.1371/journal.pclm.0000823.s029
(PDF)
S8 Fig. Predictions plot.
Logit regression results, see left column of S17 Table, for a binary indicator if after feedback the policy choice was more ambitious, for n = 257 respondents who wanted to adapt their policy selection and answered the second policy preference question, there are 121 from Goal Assigned treatment and 136 from Goal Selected.
https://doi.org/10.1371/journal.pclm.0000823.s030
(PDF)
S19 Table. Regression table.
Treatment effects (Predictions based on the first two columns are displayed in S5 Fig.)
https://doi.org/10.1371/journal.pclm.0000823.s031
(PDF)
S5 Text. Heterogenous treatment effects.
Information on heterogenous treatment effects with respect to trust in science, political self-placement, educational attainment, climate change priorities, gender, age, and income.
https://doi.org/10.1371/journal.pclm.0000823.s032
(PDF)
S9 Fig. Predictions plot.
Predicted effects of initially choosing maximum policies by stated trust in science. The associated regression table can be found in S9 Fig.
https://doi.org/10.1371/journal.pclm.0000823.s033
(PDF)
S20 Table. Regression table.
Heterogeneous Treatment Effects with respect to Trust in Science.
https://doi.org/10.1371/journal.pclm.0000823.s034
(PDF)
Acknowledgments
This project is enabled through intensive collaboration with the SWEET EDGE consortium members and collaborators, thank you. A special thanks to the “energy experts”. We are grateful for inputs on our pre-analysis plan from PolMeth Europe 2022 participants. We thank participants at the EPG Online 2022, Swiss Political Science Annual Conference in Basel, the 2023 Meeting of the European Public Choice Society in Hannover, seminar participants in Zurich (ETH and UZH CIS Colloquium), Bern and London (LSE, Social Policy department), and participants at the European Political Science Association Annual Conference in Glasgow for comments and questions on previous versions.
Rahel Freiburghaus and Karin Ingold provided valuable feedback on earlier drafts of this work. We thank Giada Gianola, and Karel Ziehli for help with the translations of the survey instruments, Moira Ettlin for assistance with survey programming and proof reading, and Céline Imobersteg with formatting assistance. This project was carried out with the support of the Swiss Federal Office of Energy (SFOE) as part of the SWEET EDGE (https://www.sweet-edge.ch) project. The authors bear sole responsibility for the results and conclusions.
References
- 1. Matthews HD, Wynes S. Current global efforts are insufficient to limit warming to 1.5°C. Science. 2022;376(6600):1404–9. pmid:35737785
- 2. Stechemesser A, Koch N, Mark E, Dilger E, Klösel P, Menicacci L, et al. Climate policies that achieved major emission reductions: Global evidence from two decades. Science. 2024;385(6711):884–92. pmid:39172830
- 3. Stern PC, Dietz T, Vandenbergh MP. The science of mitigation: closing the gap between potential and actual reduction of environmental threats. Energy Research & Social Science. 2022;91:102735.
- 4. Osorio-Aravena JC, Reyes-Chamorro L, Haas J, Aguilera J, Palma-Behnke R, Breyer C. Insights for informing energy transition policies – are decision makers listening to science? The case of Chile. Energy Strategy Reviews. 2025;58:101644.
- 5. Beiser-McGrath LF, Bernauer T. Current surveys may underestimate climate change skepticism evidence from list experiments in Germany and the USA. PLoS One. 2021;16(7):e0251034. pmid:34232962
- 6. Burstein P. The impact of public opinion on public policy: a review and an Agenda. Political Research Quarterly. 2003;56(1):29.
- 7. Vanhuysse P, Jankowski M, Tepe M. Vaccine alliance building blocks: a conjoint experiment on popular support for international COVID-19 cooperation formats. Policy Sci. 2021;54(3):493–506. pmid:34393278
- 8. Barnes L, Blumenau J, Lauderdale BE. Measuring attitudes toward public spending using a multivariate tax summary experiment. American J Political Sci. 2021;66(1):205–21.
- 9. Hellwig T, Kweon Y. Taking cues on multidimensional issues: the case of attitudes toward immigration. West European Politics. 2016;39(4):710–30.
- 10. Hainmueller J, Hopkins DJ. The hidden american immigration consensus: a conjoint analysis of attitudes toward immigrants. American J Political Sci. 2014;59(3):529–48.
- 11. Stadelmann-Steffen I, Dermont C. The unpopularity of incentive-based instruments: what improves the cost–benefit ratio?. Public Choice. 2018;175(1–2):37–62.
- 12. Häusermann S, Kurer T, Traber D. The politics of trade-offs: studying the dynamics of welfare state reform with conjoint experiments. Comparative Political Studies. 2018;52(7):1059–95.
- 13. Dermont C. Environmental decision-making: the influence of policy information. Environmental Politics. 2018;28(3):544–67.
- 14. Stadelmann-Steffen I. Citizens as veto players: climate change policy and the constraints of direct democracy. Environmental Politics. 2011;20(4):485–507.
- 15.
Bolzli M. CO2-Gesetz: So viel müssten Autofahrer mehr zahlen — nau.ch; 2021. https://www.nau.ch/news/wirtschaft/co2-gesetz-so-viel-mussten-autofahrer-mehr-zahlen-65925102
- 16.
Swiss Federal Office of the Environment. Kosten des revidierten CO2-Gesetzes für eine Durchschnittsfamilie. 2021. https://www.bafu.admin.ch/dam/bafu/de/dokumente/klima/dossiers/faktenblatt-07-kosten-fuer-eine-durchschnittsfamilie
- 17.
Peters BG, Capano G, Howlett M, Mukherjee I, Chou MH, Ravinet P. Designing for policy effectiveness. Cambridge ed. Cambridge/New York/Melbourne/New Delhi/Singapore: Cambridge University Press; 2018.
- 18.
Climate Action Tracker. Glasgow’s 2030 credibility gap: net zero’s lip service to climate action. 2021. https://climateactiontracker.org/documents/997/CAT_2021-11-09_Briefing_Global-Upyear_Glasgow2030CredibilityGap.pdf
- 19.
UNFCCC. Nationally determined contributions under the paris agreement synthesis report by the secretariat, summary. 2021.
- 20. Stoutenborough JW, Vedlitz A. The effect of perceived and assessed knowledge of climate change on public policy concerns: An empirical comparison. Environmental Science & Policy. 2014;37:23–33.
- 21. Chong D, Druckman JN. Framing theory. Annu Rev Polit Sci. 2007;10(1):103–26.
- 22. Furnham A, Boo HC. A literature review of the anchoring effect. The Journal of Socio-Economics. 2011;40(1):35–42.
- 23. Tingley D, Tomz M. International commitments and domestic opinion: the effect of the Paris agreement on public support for policies to address climate change. Environmental Politics. 2019;29(7):1135–56.
- 24. Schmidt TS, Sewerin S. Measuring the temporal dynamics of policy mixes – an empirical analysis of renewable energy policy mixes’ balance and design features in nine countries. Research Policy. 2019;48(10):103557.
- 25. Dermont C, Stadelmann-Steffen I. The role of policy and party information in direct-democratic campaigns. International Journal of Public Opinion Research. 2019;32(3):442–66.
- 26. Drews S, van den Bergh JCJM. What explains public support for climate policies? A review of empirical and experimental studies. Climate Policy. 2015;16(7):855–76.
- 27. Mildenberger M, Lachapelle E, Harrison K, Stadelmann-Steffen I. Limited impacts of carbon tax rebate programmes on public support for carbon pricing. Nat Clim Chang. 2022;12(2):141–7.
- 28. Schmid N, Guinaudeau B. Mapping public support for climate solutions in France. Environ Res Lett. 2022;17(4):044035.
- 29. Wicki M, Huber RA, Bernauer T. Can policy-packaging increase public support for costly policies? Insights from a choice experiment on policies against vehicle emissions. J Pub Pol. 2019;40(4):599–625.
- 30. Mallett R, Maslin M. Setting climate deadlines could be counterproductive. Nature. 2022;612(7940):404. pmid:36513833
- 31.
Clarke L, Wei YM, Navarro ADLV, Garg A, Hahmann AN, Khennas S, et al. Contribution of Working Group III to the sixth assessment report of the intergovernmental panel on climate change. In: Shukla PR, Skea J, Slade R, Khourdajie AA, van Diemen R, McCollum D, et al., editors. Climate Change 2022 : Mitigation of Climate Change. Cambridge, UK and New York, NY, USA: Cambridge University Press; 2022. p. 613–743.
- 32. Altman D. The potential of direct democracy: a global measure 1900 –2014). Soc Indic Res. 2016;133(3):1207–27.
- 33. Reynolds JP, Stautz K, Pilling M, van der Linden S, Marteau TM. Communicating the effectiveness and ineffectiveness of government policies and their impact on public support: a systematic review with meta-analysis. R Soc Open Sci. 2020;7(1):190522. pmid:32218927
- 34. Reynolds JP, Pilling M, Marteau TM. Communicating quantitative evidence of policy effectiveness and support for the policy: Three experimental studies. Soc Sci Med. 2018;218:1–12. pmid:30312911
- 35. Rhodes E, Axsen J, Jaccard M. Does effective climate policy require well-informed citizen support?. Global Environmental Change. 2014;29:92–104.
- 36. Ruprecht S. Bridging the gap: the influence of information and education on acceptance of environmental taxes in Switzerland. Environ Res Commun. 2023;5(7):075010.
- 37. Hainmueller J, Hopkins DJ, Yamamoto T. Causal inference in conjoint analysis: understanding multidimensional choices via stated preference experiments. Polit anal. 2014;22(1):1–30.
- 38. Armingeon K, Bürgisser R. Trade-offs between redistribution and environmental protection: the role of information, ideology, and self-interest. Journal of European Public Policy. 2020;28(4):489–509.
- 39. Stadelmann-Steffen I, Eder C, Liem M. Trade-off or co-evolution? Support for climate change policy in times of crisis. Renewable and Sustainable Energy Reviews. 2026;226:116186.
- 40. Beiser-McGrath LF, Bernauer T. Could revenue recycling make effective carbon taxation politically feasible?. Sci Adv. 2019;5(9):eaax3323. pmid:31555740
- 41. Bürgisser R, Stadelmann-Steffen I, Armingeon K. Can information, compensation and party cues increase mass support for green taxes?. Journal of European Public Policy. 2024:1–28.
- 42. Carattini S, Kallbekken S, Orlov A. How to win public support for a global carbon tax. Nature. 2019;565(7739):289–91. pmid:30651626
- 43. Jagers SC, Martinsson J, Matti S. The impact of compensatory measures on public support for carbon taxation: an experimental study in Sweden. Climate Policy. 2018;19(2):147–60.
- 44. Klenert D, Mattauch L, Combet E, Edenhofer O, Hepburn C, Rafaty R, et al. Making carbon pricing work for citizens. Nature Clim Change. 2018;8(8):669–77.
- 45. Kotchen MJ, Turk ZM, Leiserowitz AA. Public willingness to pay for a US carbon tax and preferences for spending the revenue. Environ Res Lett. 2017;12(9):094012.
- 46. Fesenfeld LP. The effects of policy design complexity on public support for climate policy. Behav Public Policy. 2022;9(1):106–31.
- 47. Druckman JN, Nelson KR. Framing and deliberation: how citizens’ conversations limit elite influence. American J Political Sci. 2003;47(4):729–45.
- 48. Stokes LC, Warshaw C. Renewable energy policy design and framing influence public support in the United States. Nat Energy. 2017;2(8).
- 49. Feldman L, Hart PS. Climate change as a polarizing cue: framing effects on public support for low-carbon energy policies. Global Environmental Change. 2018;51:54–66.
- 50. Gustafson A, Goldberg MH, Bergquist P, Lacroix K, Rosenthal SA, Leiserowitz A. The durable, bipartisan effects of emphasizing the cost savings of renewable energy. Nat Energy. 2022;7(11):1023–30.
- 51. Bernauer T, McGrath LF. Simple reframing unlikely to boost public support for climate policy. Nature Clim Change. 2016;6(7):680–3.
- 52. Campiglio E, Lamperti F, Terranova R. Believe me when I say green! Heterogeneous expectations and climate policy uncertainty. Journal of Economic Dynamics and Control. 2024;165:104900.
- 53. Sitarz J, Pahle M, Osorio S, Luderer G, Pietzcker R. EU carbon prices signal high policy credibility and farsighted actors. Nat Energy. 2024;9(6):691–702.
- 54. Tversky A, Kahneman D. Judgment under uncertainty: heuristics and biases: biases in judgments reveal some heuristics of thinking under uncertainty. Science. 1974;185(4157):1124–31.
- 55. Plum C, Olschewski R, Jobin M, van Vliet O. Public preferences for the Swiss electricity system after the nuclear phase-out: a choice experiment. Energy Policy. 2019;130:181–96.
- 56. EGAMI N, HARTMAN E. Elements of external validity: framework, design, and analysis. Am Polit Sci Rev. 2022;117(3):1070–88.
- 57. Brückmann G, Berger S, Caviola H, Hahnel UJJ, Piana V, Sahakian M, et al. Towards more impactful energy research: the salient role of social sciences and humanities. PLOS Clim. 2023;2(2):e0000132.
- 58.
R Core Team. R: A Language and Environment for Statistical Computing; 2024. https://www.R-project.org/
- 59. Wickham H, Averick M, Bryan J, Chang W, McGowan L, François R, et al. Welcome to the tidyverse. JOSS. 2019;4(43):1686.
- 60. Lüdecke D. ggeffects: Tidy data frames of marginal effects from regression models. JOSS. 2018;3(26):772.
- 61.
Hlavac M. stargazer: well-formatted regression and summary statistics tables; 2022. https://CRAN.R-project.org/package=stargazer
- 62.
Lüdecke D. sjPlot: data visualization for statistics in social science; 2024. https://CRAN.R-project.org/package=sjPlot
- 63.
Pedersen TL. patchwork: the composer of plots; 2024. https://CRAN.R-project.org/package=patchwork
- 64.
Kassambara A. Ggpubr: ‘ggplot2’ based publication ready plots. 2023. https://CRAN.R-project.org/package=ggpubr
- 65.
Huntington-Klein N. vtable: variable table for variable documentation; 2024. https://CRAN.R-project.org/package=vtable
- 66.
Lüdecke D. sjlabelled: labelled data utility functions (version 1.2.0); 2022. https://CRAN.R-project.org/package=sjlabelled
- 67.
Sharpsteen C, Bracken C. tikzDevice: R Graphics Output in LaTeX Format; 2023. https://CRAN.R-project.org/package=tikzDevice
- 68.
Xie Y. tinytex: helper functions to install and maintain TeX Live, and Compile LaTeX Documents; 2025. https://github.com/rstudio/tinytex.
- 69. Xie Y. TinyTeX: a lightweight, cross-platform, and easy-to-maintain LaTeX distribution based on TeX Live. TUGboat. 2019;40(1):30–2.
- 70.
Wickham H, Pedersen TL, Seidel D. scales: scale functions for visualization; 2023. https://CRAN.R-project.org/package=scales
- 71.
Rinker TW, Kurkiewicz D. Pacman: package management for R. 2018. http://github.com/trinker/pacman
- 72. Arel-Bundock V, McCrain J. Software citations in political science. APSC. 2023;56(3):398–401.
- 73.
Howlett M. Designing public policies: principles and instruments. Routledge Taylor & Francis Group; 2011.
- 74. Barabas J, Jerit J. Are survey experiments externally valid?. Am Polit Sci Rev. 2010;104(2):226–42.
- 75. Tröndle T, Annaheim J, Hoppe J, Hanger-Kopp S, Patt A. Public preferences for phasing-out fossil fuels in the german building and transport sectors. Environ Res Commun. 2023;5(8):081001.
- 76.
Raimi D, Zhu Y, Newell R, Prest B, Bergman A. Global energy outlook 2023 : sowing the seeds of an energy transition. 2023. https://media.rff.org/documents/Report_23-02_NPKKYws.pdf
- 77. Soss J, Schram SF. A public transformed? Welfare reform as policy feedback. Am Polit Sci Rev. 2007;101(1):111–27.