BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Capturing household beliefs, preferences and attitudes in ECA Acknowledgments This report was prepared by Ailin Tomio (Consultant, DIME5), Juni Singh (Extended Term Consul- tant, EECPV), Alessandro Silvestri (Consultant, DIME5) and Jonathan Karver (Economist, DIME5), under the guidance of Anna Fruttero (Senior Economist, EECPV), Alexandru Cojocaru (Senior Economist, EECPV), Obert Pimhidzai (Lead Economist, EECPV), Ambar Narayan (Practice Manager, EECPV) , and Asad Alam (Regional Practice Director, EECDR). The team is grateful to UDA Consulting and LLC Zerkalo Central Asia, who collected survey data for this report, and to colleagues at the World Bank that provided advice and inputs during the preparation of diagnostic activities and the report, especially Renos Vakis (Lead Economist, DIME5), Patrick Behrer (Research Economist, DECSI), Abigail Dalton (Senior Operations Officer, DIME5), Essienawan Essien (Program Assis- tant, EECPV) and Armanda Carcani (Program Assistant, ETIRI). The team would like to thank the peer reviewers - Monica Robayo (Senior Economist, EECPV), Esma Kreso (Senior Environmental Specialist, SCAE2), and Jorge Luis Castañeda (Economist, DIME5) -for their valuable feedback. This work was jointly supported by the World Bank’s Climate Support Facility (CSF) and the Foreign, Commonwealth & Development Office (FCDO)’s Effective Governance for Economic De- velopment in Central Asia (EGED) Trust Fund Climate Window. The mission of the CSF is to support developing countries in accelerating their transition to low-carbon and climateresilient develop- ment and elevate the national decarbonization agenda. The Whole-of-Economy (WoE) Program in EFI supported by the Climate Support Facility (CSF) was launched in December 2021. Its goal is to strengthen the analytical basis and diagnostics to support effective policy advice on issues relating to a WoE approach to climate change and to improve the capacity of client countries in this approach. The WoE under the Climate Support Facility (CSF) Trust Fund will be deployed over four years (fiscal 2022−25) through global and regional programmatic block grants. Table of Contents Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1. INTRODUCTION AND CONTEXT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2. A BEHAVIORAL APPROACH TO CLIMATE POLICY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 3. CLIMATE CHANGE PERCEPTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 4. SUPPORT FOR GOVERNMENT POLICIES AND AWARENESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 5. ENERGY EFFICIENCY AND HEATING AND COOLING PRACTICES . . . . . . . . . . . . . . . . . . . . . . . . . . 31 6. SOLID WASTE MANAGEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 7. BARRIERS TO THE ADOPTION OF CLIMATE-FRIENDLY BEHAVIORS AND TECHNOLOGIES . . . . . 43 8. INFORMATION TREATMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 9. POLICY RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 9.1) Communication-related solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51 9.2) System and program-level solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 10. CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 APPENDIX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Appendix A The survey and sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Executive summary The Europe and Central Asia region faces significant climate-related risks and promoting sustainable practices is crucial for long-term growth.1 Climate change is having a profound impact on the countries of the Europe and Central Asia (ECA) region. ECA has some of the highest per capita carbon emissions in the world due to fossil-fuel-dependent, energy-intensive economies. The consequences of this dependency on the environment, coupled with rising energy prices in the region, signal a strong need for sustainable clean energy solutions (particularly in re- newable energy) and transitions to less-energy-intensive technologies 1 Sustainable practices are actions that aim to reduce environmental impact, preserve resources, and maintain ethical operations. They support ecological, human, and economic health and vitality. Some examples of sustainable practices are reducing, reusing, and recycling waste; avoiding single-use plastics; disposing of waste in the proper bins; using energy efficiently, for example by upgrading heating and cooling devices; reducing fossil fuel use to generate electricity and heating; and using alternative transportation modes such as walking, biking, or using public transit instead of driving. 4 Capturing household beliefs, preferences and attitudes in ECA and practices. The World Bank’s ECA Climate Roadmap identifies “tran- sition risk” as the biggest challenge facing the region, requiring critical action on areas related to energy, industry, transport, urban develop- ment, food, landscapes, and water. Addressing these areas will require a ‘whole of economy’ approach that considers the unique contributions of policy, financing, and people. This report provides valuable insights into the behavioral dimensions of climate change in the region, high- lighting the need for tailored policies that address both individual and systemic barriers. Behavioral insights can meaningfully help address climate change challenges in the region. Insights from behavioral science support a holistic approach to addressing climate issues by eval- uating the systemic, physical, social, and individual (including psychological) factors that affect the decision-making process concerning environmentally harmful (or positive) behaviors. Beyond a deeper understanding of the human component of climate change challenges, behavioral in- sights can support the design of effective policy interventions that promote sustainable actions. This report examines the behavioral dimensions of climate change and the green transition in the ECA region, focusing on how individual beliefs, attitudes and behaviors influence the adoption of sustainable practices. The study makes use of a survey of 16,500 households across 16 countries that aimed to collect information from participants about climate change percep- tions, support for government policies, and willingness to adopt energy-efficient technologies, clean heating and cooling devices, and sustainability-oriented waste management practices. The survey was implemented in person or via telephone or the web, depending on the country, in order to have a nationally representative sample. The analysis follows the levels of behavioral determinants of Nielsen et al. (2024), namely system, physical, social, and individual factors that influence decision-making. We also measure individuals’ willingness to upgrade existing technol- ogies and propose an information intervention to increase willingness to upgrade. The results suggest a complex interplay of system, physical, social, and individual factors as influencing ECA households’ practices. While a majority of ECA residents acknowledge the reality of climate change, there is a significant underestimation of the level of concern among their peers. This “pluralistic ignorance” suggests a need for communication strategies that highlight the widespread support for climate action. ECA residents generally favor incentives and access improvements over taxes and bans to promote sustainable practices. However, support for these policies is significantly influenced by individual attitudes, with those willing to make lifestyle changes for the environment being more likely to support government action. A significant portion 5 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION of the population relies on traditional heating systems, highlighting the need for promoting ener- gy-efficient technologies. While willingness to upgrade heating systems and insulation is high, particularly in Central Asia, financial constraints and lack of trust in sustainable technologies remain significant barriers. Waste separation practices are more prevalent in EU countries, Türkiye, Tajikistan, and Uzbekistan and less so in other regions. Most households are willing to pay more for improved waste collection services, particularly in Central Asia, South Caucasus, and Moldova. Financial constraints and lack of trust in sustainable technologies are the most cited barriers to the adoption of climate-friendly behaviors. The report proposes a range of communication and program-level solutions to address these barriers. Lack of information and trust in technology were identified as barriers to adopting cli- mate-friendly technologies, hence the focus on communication. In terms of communications, policy makers should leverage trusted messengers such as academics and NGOs, utilize preferred communication channels, highlight the widespread support for climate action, emphasize future gains over present costs, clarify policy mechanisms, and tailor messages to specific population segments. In tandem with these efforts, targeted support programs should increase access to sustainable practices, use choice architecture to constrain behaviors that are not sustainable, consider distributional concerns in their design, and reduce the “sludge” in them. 6 Capturing household beliefs, preferences and attitudes in ECA 7 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Introduction and context The impacts of climate change are profoundly affecting countries of the Europe and Central Asia (ECA) region. The heavy reliance on fossil fuels and energy-intensive economies in Eastern Europe and Central Asia have resulted in that region’s having one of the highest per capi- ta carbon emission rates in the world—and its total emissions exceed those of Northern, Southern, and Western Europe combined. Moreover, since February 2022 geopolitical events have made it evident that ECA needs to diversify its energy sources and avoid excessive dependence on fossil fuel imports. The drop in Russian natural gas flows to ECA in 2022 marked the single largest supply shock in the history of global gas markets (World Bank 2024a). The clean energy transition will require extraordinary private and public investment and that both the supply and demand side move to clean technologies. These include on the 8 Capturing household beliefs, preferences and attitudes in ECA supply side the renewables needed for the power sector’s decarbon- ization and, on the demand side, sustainable transport, heat pumps for buildings, and improved waste management.2 These new technologies and the reduction in the consumption of fossil-fuel-intensive products offer no-regret investments in decarbonization. The World Bank’s 2021 ECA Climate Roadmap highlights “transition risk” as the biggest chal- lenge facing the region,3 with several ECA countries among the world’s most emission intensive due to high heating needs and fossil fuel dependence. The 2021 ECA Climate Roadmap calls for the prioritization of five interdependent systems transitions to support climate adaptation and mitigation in ECA: (1) energy, (2) industry, (3) transport, (4) urban, and (5) food, landscapes, and water, underpinned by a “whole of economy” agenda encompassing the broad themes of “policy,” “finance,” and “people.” Along with the need for careful attention to distributional impacts and policy trade-offs that will inevitably arise during implementation, adequate consideration of political economy factors and citizen engagement will be key, as these are likely to make or break the viability and sustainability of climate policy. Behavior changes of citizens are critical to the effectiveness of these decarbonization mea- sures. The transition to a climate-friendly system will entail introduction of both enforcement mechanisms and improved engagement with the public to motivate climate transitions. Changing behavioral practices and promoting public participation are thus crucial and will require the design of incentives and engagement (in particular, awareness) systems. Behavioral science has proven effective in identifying critical determinants for sustainable behaviors and providing cost-effective interventions to tackle those determinants (Vlasceanu et al. 2024). By better understanding the system, physical, social, and individual factors (Nielsen et al. 2024) that affect the decision-mak- ing process regarding environmentally harmful (or positive) behaviors, behavioral science can assist with the design of effective interventions and policies that promote sustainable actions. Furthermore, given that specific environmental problems such as energy efficiency, clean energy, transportation, and solid waste pollution are complex and require changes in behavior across indi- viduals, communities, and organizations, tools and methods provided by behavioral science can develop evidence-based policies tailored to the needs and preferences of different populations. 2 The waste sector contributes 3.3 percent to global greenhouse gas emissions, most of it from waste landfills (Climate Watch, GHG Emissions, https://www.climatewatchdata.org/ghg-emissions). Although this percentage may seem low, waste management is a key area where behavioral insights can mitigate climate impacts. Illegal waste dumping and open waste burning most strongly impact vulnerable and informal communities, who often work or are situated near illegal dumpsters and do not have access to proper waste collection. Perceived inconvenience, cost, social norms, and apathy can all impede effective waste management practices. 3 For more information, see: https://www.worldbank.org/en/news/press-release/2021/06/22/world-bank-group-increases-support-for- climate-action-in-developing-countries. 9 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION This report seeks to inform activities of the Whole-of-Economy (WoE) Climate Change Pro- gram, a World Bank Group (WBG) operation designed to deepen the understanding of climate change trends and their development implications in the ECA region and address the challenges through development-compatible policies. The Whole of Economy approach is an integrated strategy involving all sectors and stakeholders in the green transition. Adopting sustainable sys- tems and climate action is a complex problem that also involves a range of stakeholders such as the government and private sector because of their relation to citizens’ behavior. In this report we focus on the perspective of households and how we can help improve uptake of climate-friendly actions. We focus on 16 countries across Eastern and Central Europe and Asia. They are grouped into six regions: European Union (Bulgaria, Croatia, Poland, and Romania), the Western Balkans (Albania, Bosnia and Herzegovina, Kosovo, and Serbia), Central Asia (Kazakhstan, Kyrgyz Re- public, Tajikistan, and Uzbekistan), the South Caucasus (Armenia and Georgia), Türkiye, and Mol- dova. Our results highlight behavioral barriers that influence people’s sustainable behaviors and attitudes and inform government-led dissemination activities under component 2 of Integrating the People, HD, and Social Dimensions in the Climate Agenda. We present solutions that leverage tailored information, effective messengers, and communication between peers to increase take- up of environmentally friendly behavior. 10 Capturing household beliefs, preferences and attitudes in ECA A behavioral approach to climate policy Behavioral science can support the mitigation of climate change and environmental challenges and foster sustainable development.4 Sev- eral studies suggest that context-specific components and individual preferences interact to form decision-making patterns and behaviors in relation to sustainability. Climate change beliefs are shaped by a wide array of socio-structural and psychological factors, including age, gender, level of education, socioeconomic status, political orientation, and value systems (Echavarren 2017; Milfont et al. 2015; Poortinga 4 Following Hallsworth (2023), behavioral science as a discipline focuses on analyzing human behavior at the individual or group level. Based on this analysis, it uses scientific methods to explain and predict behavior. 11 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION et al. 2019). Cross-nationally, countries whose populations are most supportive of higher taxes on fossil fuels are not those who are more aware and concerned about climate change; rather, they are those with the highest levels of political trust (Fairbrother, Sevä, and Kulin 2019). Social norms are another factor that affects proenvironmental behav- iors.5 These tend to be more influential in individualistic than collectivist countries (Bergquist, Nilssona, and Schults 2019). Traditional policies operate under the assumption that people will make good decisions if they are fully informed and are presented with the right financial incentives. However, this approach fails to account for the fact that human behavior is influenced by various factors that go beyond socioeconomic characteristics, notably emotions, biases, and institutional and group contexts. As a result, public-service delivery outcomes, behavior, and decision-making may be negatively impacted. Research on how people think and behave shows that several interconnect- ed factors influence individuals. According to Nielsen et al. (2024) these factors can be classified into four categories. The first is the system level, which addresses the influence of elements such as institutions and organizations that form part of the system, as well as policies, legislation, and economic incentives. The second is the physical level, which considers crucial infrastructure and components that facilitate or obstruct desired behaviors. The third is the social level, which includes collective behaviors, perceptions, and habits; and the fourth is the individual level, which aims to capture people’s awareness, motivations, and values. Based in this framework, we pro- pose four relevant categories by means of which to analyze the people component from the ECA Climate Roadmap (figure 1). 5 These refer to the influence of the perceived informal rules that define acceptable and appropriate actions within a given group or community (Farrow, Grolleau, and Ibanez 2017). 12 Capturing household beliefs, preferences and attitudes in ECA Figure 1 Behavioral A pproach to Informing Climate Change Policies Infrastructure Institutional Social Individual Note: The Individual Level focuses on the individual person, their behaviors, attitudes, and personal characteristics. Examples include personal beliefs, knowledge, skills, and actions. The Social Level encompasses the immediate social environment of the individual, including relationships and social networks. Examples include family, friends, peer groups, and community interactions. The Institutional Level includes the organizations and institutions that influence the individual and social levels. Examples include government units, schools, workplaces, religious institutions, and organizations. The Infrastructure Level refers to the physical and organizational structures needed for the operation of a society. Examples include transportation systems, communication networks, water supply, and energy grids. Understanding what motivates people and influences their behavior is central to success- ful policy making. A behavioral approach for sustainability includes focusing on individual de- cision-making and its determinants in the context of climate change. Insights from behavioral studies have shown behaviorally informed interventions can affect social practices in the absence of legislation, changing prices, or the restriction of choices (Allcott and Kessler 2015) through tools like information provision, reminders, and social comparisons. One of the challenges in tackling climate change is that the cost of action is immediate but the benefit spans across time. Because individuals are shortsighted and biased toward the present, it can be difficult for people to develop a strong sense of cause and effect to motivate their actions. One of the factors we identified in willingness to take action was risk attitudes. Given the uncertainty of climate change outcomes, we observe that individuals willing to take more risks are likely to be more willing to upgrade to sustainable energy systems. Social norms are another factor that affects proenvironmental behaviors. This refers to the perceived informal rules that define the social acceptability of actions. These tend to be more influential in individualistic than collectivist countries (Bergquist, Nilssona, and Schults 2019). There can be pluralistic ignorance (misunderstanding others’ opinions) on norms related to cli- mate action. Correcting such misperceptions by providing information about the beliefs of other 13 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION people can help accelerate climate-friendly actions. Altruism is another important factor that af- fects environmental behavior. This refers to individuals who have a higher concern for the well-be- ing of others. We see this translate into higher willingness to take steps toward the adoption of efficient energy systems. In ECA, previous work has found that support for climate change policies varies according to age and level of political trust (Cojocaru, Lokshin, and Nikandrova 2024). Older respondents reported being as (or even more) concerned about climate change than younger respondents. However, the former are also more likely to object to tax increases to finance climate adaptation and mitigation policies, which likely reflects the general decline in willingness to pay taxes with age rather than a specific lack of willingness to finance climate change policies. On the other hand, Europeans with a high level of political trust tend to be much more supportive of fossil fuel taxes if they also believe in the reality and dangers of anthropogenic climate change. Research on household energy efficiency and behavior in the ECA region highlights the challenges of balancing individual cost with fiscal and environmental concerns. Studies have shown that reducing explicit energy subsidies disproportionately affects the poor, necessitating targeted social assistance (Lampietti, Banerjee, and Branczik 2006; Stuggins, Sharabaroff, and Semikolenova 2013). Governments struggle to maintain low energy tariffs while addressing bud- get pressures and environmental issues (Olivier, Ruggeri Laderchi, and Trimble 2013). Changes in energy-related behaviors are needed to implement even modest policies for efficiency and use of renewable energy. These range in their targets from daily habits and attitudes to acceptance of transitions and innovations (Gynther, Mikkonen, and Smits 2012). Successful energy efficiency policies in some countries have led to reduced energy intensi- ties, offering valuable lessons for the region (Stuggins, Sharabaroff, and Semikolenova 2013). Recent research in Poland reveals that households in fossil-dependent regions have adopted no- tech and low-tech energy-saving practices in response to price increases, with socioeconomic fac- tors influencing behavior (Kola-Bezka and Leki 2024). These findings suggest that policy makers must consider diverse household characteristics and behaviors when designing inclusive ener- gy-transition strategies that balance affordability, efficiency, and environmental goals. Liobikienè and Minelgaitè (2021) expand on these findings by analyzing energy-saving behaviors across EU countries, revealing that individuals who are environmentally responsible and concerned about climate change are more likely to adopt energy-efficient practices. Their study emphasizes that personal responsibility and environmental concern are key drivers of energy-saving behaviors and that interventions targeting these attitudes are essential for increasing energy efficiency. 14 Capturing household beliefs, preferences and attitudes in ECA For sustainable household waste management, the existing literature has identified key behaviors and their determinants. Ong, Fearnley, and Chia (2019) find that future orientedness, environmental identity, and knowledge of recyclable materials influence waste minimization be- haviors. Similarly, Miliute-Plepiene et al. (2016) emphasize the importance of situational variables and ecological awareness in waste sorting and recycling in a case study in Sweden. Expanding on this, Wang et al. (2021) find that environmental regulation had the highest effect on behavioral intention in household waste sorting in China, followed by behavioral control factors such as knowledge, awareness, and moral responsibility. Johansson (2016), in a systematic review of European and American literature, examines the factors that influence individuals to engage in recycling activities and emphasizes the need for a well-designed recycling infrastructure, specific recycling knowledge, and a general understanding of environmental aspects. Highlighting gaps in existing research, Raghu and Rodrigues (2020) point to psychological factors and the theory of planned behavior as critical areas for future investigation in sustainable solid waste management. In ECA, several factors influence solid waste management, including economic, social, and environmental considerations. For example, Zaikova et al. (2022) find that the factors influencing household waste separation behavior differ significantly between Russia and Finland. In Russia, the inconvenience of waste collection is a significant barrier, while intentions and information availability are vital drivers. In contrast, individuals’ intentions primarily influence waste separa- tion behavior in Finland. These findings show the critical implications of the context for developing recycling practices. Minelgaitè and Liobikienè (2019) highlight the complex relationship between waste generation and management behaviors in the EU. They find that while interventions tar- geting reducing and reusing behaviors had little impact on waste generation, recycling behaviors were positively correlated with it. This suggests that a greater focus must be paid to interventions promoting sustainable consumption and production and enhancing recycling efforts, which are crucial in minimizing waste generation. Minelgaitè and Liobikienè (2019) also emphasize the role of individual attitudes, with those who understand their contribution to the waste problem being more likely to engage in waste management behaviors. 15 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Climate change perceptions The questionnaire that was used was structured into six main sec- tions: sociodemographics, climate perceptions, support for gov- ernment policies, energy efficiency and home heating, solid waste management, and information treatment. We present results on each of the sections below. More details on the survey and sampling can be found in the appendix. Perceptions around climate change in ECA reflect an acceptance of climate change, al- beit with room for more progressive views. Individuals in ECA believe climate change is real (three out of four adults) and that the earth is warmer than it was one hundred years ago. These results are also supported by other surveys such as the World Bank’s Life in Transition survey 16 Capturing household beliefs, preferences and attitudes in ECA (LITS) IV.6 However, most underestimated how concerned others in their country are about this, and almost half think the government overestimates its impact. Out of the 16 surveyed countries, only Kazakhstan and Poland show lower shares of agreement with the idea that the earth is warmer, and only in Poland do the majority of people believe this is not true (figure 2). This is similar to the proportion seen in other middle- and high-income countries (Dechezleprêtre et al. 2022). Figure 2 Perc ep tions o f Global Warming Most places on Earth are warmer than they were 100 years ago. Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% 75% Share of respondents 50% 99% 95% 94% 96% 94% 94% 93% 92% 85% 85% 92% 92% 74% 91% 59% 25% 42% 0% Tajikistan Kyrgyz Rep. Uzbekistan Kazakhstan Bulgaria Croatia Romania Poland Moldova Georgia Armenia Türkiye Albania BiH Kosovo Serbia Yes Note: The figure presents the percentages of people who agree with the statement that Earth is warmer than 100 years ago across the ECA countries; “Yes” and “No” were the options provided to the respondents. There is a gap between people’s actual beliefs and what they believe are the concerns of other citizens when it comes to the seriousness of the impact of climate change (figure 3). People are skeptical about how much concern for climate change there is in their country. The survey asked individuals to estimate the proportion of citizens concerned that climate change is going to have an impact in their lifetime. These responses are compared to those of the LITS IV,7 where the citizens self-reported whether they were personally concerned about the impact of climate change; specifically, the latter survey posed the question: “Do you think climate change seriously affects or will seriously affect you during your lifetime?” In all but one country, estimates 6 For more information about the Life in Transition survey, see https://lits-visualizer.ebrd.com/. 7 For more information, see https://www.ebrd.com/what-we-do/economic-research-and-data/data/lits.html. 17 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION on concerns over climate change as revealed through the survey data on which this report is based are substantially lower than the results from the LITS IV. Only in Bosnia and Herzegovina is the result overestimated, by around 15 percent. In countries such as Türkiye, Kosovo, Armenia, and Moldova this share is underestimated by over 30 percent. Figure 3 S ec ond- Order Belie fs about the S eriousness o f Climate Change among Citi zens Out of 100 people in your country, how many do you think consider that climate change will have serious impact in their lifetime. Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% 75% Average Response 50% 70.4 61 58 56.6 53.7 53.7 53.7 53.1 25% 47.3 49.3 50.6 45.2 44.5 43.4 40.7 38.7 0% Tajikistan Kyrgyz Rep. Kazakhstan Uzbekistan Romania Bulgaria Poland Croatia Moldova Georgia Armenia Türkiye BiH Albania Serbia Kosovo LITS survey Note: The figure presents the belief of respondents as to the percentage of people in their country who believe climate change will have a serious impact in their lifetime. We compare these results to those of the LITS IV, which captured the beliefs of respondents themselves on this topic. Despite the majority’s agreeing that climate change is a concern, people still believe that the impacts are overstated by their governments. An important share of adults in ECA hold the belief that climate impacts are overstated by their governments, although there is variation across countries. Around 50 percent of individuals believe their governments are overstating climate change (figure 4). Only in Tajikistan and Albania do a substantially higher share of adults believe the government is inflating the importance of climate change. 18 Capturing household beliefs, preferences and attitudes in ECA Figure 4 Levels o f Belie f that the Government Frequently Overstates the Impac t o f Climate Change Climate change and its likely impacts are frequently overstated by the government Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 80% 21% 60% Share of respondents 28% 40% 26% 19% 38% 32% 33% 27% 28% 57% 29% 29% 27% 25% 34% 38% 20% 20% 37% 25% 26% 17% 19% 14% 19% 16% 17% 15% 13% 14% 6% 10% 6% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Somewhat agree Strongly agree Note: The figure presents the perceptions of respondents concerning the proportion of people in their country who believe climate change will have a serious impact in their lifetime. We compare these results those of the LITS IV, which captured the beliefs of respondents themselves on this question. Most adults in ECA report being willing to take action to safeguard the environment regard- less of whether their peers are doing the same and most are willing to make specific lifestyle changes. These results are in line with the results of the LITS IV that show that a relevant propor- tion of the population in these countries is willing to pay more in taxes to prevent environmental pollution. In most countries, between 60 and 80 percent of the population is willing to act and change their lifestyles for the environment (figure 5). Apart from Türkiye and Poland, at least three out of four adults are willing to make lifestyle changes for the environment. This is true even if people perceive that their friends, family, or neighbors are not taking environmentally friendly actions, with the sole exception of Uzbekistan. 19 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Figure 5 Willingness to Make Lif estyle Changes I am willing to change my current lifestyle for the benefit of the environment Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% 15% 21% 75% 26% 24% Share of respondents 18% 59% 29% 40% 30% 39% 43% 44% 74% 50% 42% 33% 82% 47% 70% 64% 60% 62% 25% 46% 36% 40% 37% 31% 30% 28% 29% 31% 18% 14% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Somewhat agree Strongly agree Note: The figure presents the percentages of people who agree strongly or agree with the statement, “I am willing to make changes to my current lifestyle for the benefit of environment.” Across the ECA region, social influence, risk-taking attitudes, and household composition are generally associated with a higher willingness to change lifestyles for the good of the en- vironment. Previous studies conducted in ECA found that those who are risk averse and those living in larger households are more convinced that climate change is real (Cojocaru, Lokshin, and Nikandrova 2024). In contrast, our results suggest that this belief does not translate into willingness to change behavior, but rather the contrary. Those who describe themselves as more prone to taking risks and those who state that more than 5 out of 10 of their peers have engaged in sustainable behavior or upgraded to sustainable technologies have a higher propensity to change. On the other hand, families with children living in the household have a lower propensity to engage in proenvironmental behavior. These results suggest that the determinants of awareness may not be the same as those of willingness to change behaviors. 20 Capturing household beliefs, preferences and attitudes in ECA Figure 6 Fac tors Influencing Willingness to Change Lif estyle Factors influencing willingness to change lifestyle for the environment Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.05 0.00 0.05 0.10 Marginal effects Note: The figure shows the coefficients from a probit regression where the dependent variable has a value of 1 if individuals are willing to change their lifestyles. This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value of 1 if a person is above the median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out to 100 euros) to a charity of the person’s choice. Social influence is a dummy variable that has a value of 1 if the person believes more than 5 of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. Trusted sources of information also appear to influence the willingness to change lifestyle choices (figure 6). People who trust information on new sustainable technologies and practices that comes mainly from governmental institutions have a lower propensity to change their life- style when compared to those who trust academia, NGOs, and commercial banks for the same information. This result is in line with the literature that shows trust plays a significant role in shaping climate change mitigation and adaptation behaviors (Fairbrother, Sevä, and Kulin 2019; Gür 2020) and that in particular trust in scientists and NGOs correlates with climate-friendly be- haviors (Cologna and Siegrist 2020). 21 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Figure 7 Fac tors Influencing the Perc ep tion o f Climate Change Being Overstated by the Government Factors influencing perceptions of climate change being overstated by government Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.10 -0.05 0.00 0.05 Marginal effects Note: The figure shows the coefficients from a probit regression where the dependent variable takes a value of 1 if individuals agree to strongly agree that government frequently overstates the impact of climate change. This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value of 1 if a person is above median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out to 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes more than 5 out of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. 22 Capturing household beliefs, preferences and attitudes in ECA Figure 8 Fac tors Influencing the Perc ep tion o f Climate Change’s Being Real Factors influencing perceptions of climate change being real Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.04 -0.02 0.00 0.02 0.04 Marginal effects Note: The figure shows the coefficients from a probit regression where the dependent variable has a value of 1 if individuals respond “Yes” to the statement “Most places are warmer than they were 100 years ago.” This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value 1 if a person is above median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out of 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes more than 5 out of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. The results presented above highlight the importance of risk attitudes in determining per- ceptions on climate change (figure 7, 8). People with a tendency to take risks might be early adopters of sustainable behaviors who can serve as examples for those who lag behind. Social influence through new messengers and champions can be a channel to impact perceptions in order to drive changes in individual household behaviors. Furthermore, there is a need for new messengers and champions to deliver news related to sustainable alternatives. Engagement ap- proaches to improve the take-up of sustainable behaviors and technologies can harness these nontraditional messengers and messages effectively to address low willingness to upgrade be- haviors. Such solutions will be elaborated in section 9. 23 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Support for government policies and awareness In all countries of the ECA region, more than half of adults are aware of subsidies for energy-efficient appliances, modern heating and cooling systems, and solar panels.8 Overall, awareness of such subsidy pro- grams is associated with higher education and income groups, as well as some behavioral factors such as social pressure and willingness to make lifestyle changes for the environment. Awareness of solar panel support programs is highest in the EU, Moldova, Bosnia and Herzegovi- na, Serbia, and Armenia (figure 9). For waste management, this number is much lower across all countries. On average around 15 percent of respondents are aware of all four support programs. 8 This question did not differentiate between unawareness and unavailability of subsidies. Conclusions should consider this limitation. 24 Capturing household beliefs, preferences and attitudes in ECA Figure 9 Awareness o f Government Suppor t Programs Awareness of Government Support Programs by Country Central Asia European Union Moldova South Caucasus Türkiye Western Balkans Waste management Solar Panels Modern heating or cooling systems Energy efficient appliances Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Proportion 20% 40% 60% 80% Note: The figure shows the proportions of people who report being aware about the four government programs mentioned above. The question asked was ”Which of the following government support programs are you aware about?” Education is an important factor influencing levels of awareness, as more-educated indi- viduals are between 10 and 15 percent more likely to be aware of support programs. Adults who have completed secondary or tertiary education are more likely to be aware of programs supporting the purchases of heating devices, solar panels, or improved waste management. Higher-income individuals are significantly more aware of support programs for all but solar pan- els compared with lower-income groups. Figure 10 shows an index that takes into consideration awareness of all four support programs. Section B in the annex presents these graphs individually. Age and gender are also linked to different awareness levels on support programs. Although previous studies in ECA have not found evidence of an “age gap” in perceptions of climate change (Cojocaru, Lokshin, and Nikandrova 2024), this study found that younger respondents are less likely to be aware of support programs, specifically those for solar panels and appliances, and women are more aware of public programs in the case of waste management. These results reflect common findings in the literature that women and younger people tend to be more concerned with environment and willing to contribute (Bush and Clayton 2023; Torgler, García Valiñas, and Macintyre 2008). There is also literature that suggests that sociodemographic correlations with environmental concerns often lead to mixed results (Park, Choi, and Kim 2012; Sargisson, De Groot, and Steg 2020). 25 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Awareness is also a function of social factors and the mental models of adults across ECA. Social influence and personal attitudes have been shown to affect awareness in meaningful ways. Being surrounded by people who have taken up sustainable energy practices is an indicator of higher awareness of programs for all type of technologies, while our study shows that willingness to make lifestyle changes for the environment is significant in the case of solar panels and waste management. Figure 10 Fac tors Influencing Awareness o f Government Suppor t Programs in Europe and C entral A sia, 2024 Factors influencing awareness of government support programs Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.10 0.00 0.10 0.20 Marginal effects Note: The dependent variable is an index calculated as a standardized weighted index following a generalized least squares (GLS) weighting for awareness on four government support programs. The question was “Which of the following government support programs are you aware about?” with the options being a purchase of energy efficient appliances, purchase of modern heating, purchase of solar panels, and acquisition of solid waste bins. This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value 1 if a person is above the median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out of 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes more than 5 out of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. 26 Capturing household beliefs, preferences and attitudes in ECA Support for government policies that promote sustainable practices is rather high in all ECA countries, with a preference for incentives and access improvements over taxes and bans. All government actions regarding sustainable practices, behaviors, and technologies are backed by at least half of adults across all the ECA countries. The pricing of polluting technologies through taxes is the least supported measure, backed by around 50 percent of adults in the EU, Armenia, Türkiye, Albania, and Kosovo. Bans on outdated and polluting technologies are preferred to taxes in all countries. Incentives and access improvements to these technologies are the most desired action, supported by 75 to 95 percent of adults in each country (figure 11). Figure 11 Suppor t for Inc enti ves, Ta x es, and Bans to Promote Sustainability in Europe and C entral A sia, 2024 Perception on climate change and actions Central Asia European Union Moldova South Caucasus Türkiye Western Balkans Improve access Bans Taxes Incentives Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Proportion 50% 60% 70% 80% 90% Note: The figure shows the percentages of people who say they support or strongly support the above-mentioned government policies. The question was “What is your level of support for the following government actions to favor sustainable practices, behaviors and technologies?” Willingness to make lifestyle changes for the environment is positively correlated with sup- port for government actions that foster sustainable practices. This proenvironmental preference is connected to a higher likelihood (around 20 percentage points) of being supportive of any of the government actions mentioned above, particularly the most restrictive ones—taxes and bans (figure 12). The size of the effect is more than three times the size of any other factor. Risk takers are also more likely to support government actions of all types (by around 4−5 percentage points) 27 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION compared to risk-averse respondents. This is consistent with the previous results on willingness to change lifestyle. Risk takers might be less dependent on the status quo and may be more willing to accept government actions that promote changes in households’ lifestyles. Social influence leads to higher support for more-restrictive government actions only. Re- spondents who are surrounded by a majority of peers engaging in sustainable behavior are sig- nificantly more likely to support government actions of bans and taxes. The result does not hold in the case of incentives and improved access where social pressure is not significant. Figure 12. Fac tors Influencing Suppor t for Diff erent Government Policies Fostering Sustainability in Europe and C entral A sia, 2024 a. Taxes Factors influencing support for taxes on polluting practices Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.05 0.00 0.05 0.10 0.15 0.20 Marginal effects Note: The figure shows the coefficients from a probit regression where the dependent variable has a value of 1 if individuals strongly support or somewhat support increasing the price of polluting technologies or practices through taxes. This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value 1 if a person is above the median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out to 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes more than 5 out of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. 28 Capturing household beliefs, preferences and attitudes in ECA b. Incentives Factors influencing support for incentives schemes Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.05 0.00 0.05 0.10 0.15 0.20 Marginal effects Note: The figure shows the coefficients from a probit regression where the dependent variable is 1 if individuals strongly support or somewhat support the introduction of incentive schemes to support the adoption of sustainable practices, behaviors and technologies. This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value 1 if a person is above the median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out to 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes more than 5 out of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. Older respondents and women are also more favorable toward government actions to sup- port sustainable practices. People under 30 years of age are less likely to support any of the four proposed government actions. Women and people over 55 are significantly more supportive of all actions apart from taxes, where significance does not hold. Again, this is in line with the literature about higher concern and involvement from women (Bush and Clayton 2023; Torgler, García Valiñas, and Macintyre 2008), but contradicts the claim that younger people are more open to initiatives to combat climate change (Gray et al., 2019). Income plays a role in all government actions apart from improved access, with higher-income groups being more supportive of taxes 29 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION while lower-income groups are less supportive of bans and incentives. In contrast to what was found regarding awareness of programs, education does not seem to play a role in levels of policy support for sustainable practices. Overall, the results show that social influence, risk attitudes, and age are correlated with willingness to change lifestyle and awareness and support for proclimate programs. There is a specific need for a more targeted communication to young people to increase their awareness of and support for government initiatives. Social influence is a relevant factor that can be used as a tool to promote sustainable behaviors among those who present low support for programs and willingness to change. These results underscore the importance of selecting the right messenger and channel to reach different audiences and increase the awareness of available support programs. Young people and households at a low socioeconomic level show a decreased awareness of these pro- grams, which could be tackled by implementing targeted messages through trusted sources, peers, and social influencers. Beyond communications, it is important to design support programs that are accessible and to contemplate the needs and motivations of specific populations. 30 Capturing household beliefs, preferences and attitudes in ECA Energy efficiency and heating and cooling practices Energy efficiency is key to keeping energy consumption in check amid rising demand. It can also stabilize energy bills during the en- ergy transition by lowering household consumption by volume. The ECA region’s energy system is vulnerable to supply shocks as well as to seasonal demand shocks arising from summer droughts and heat waves and winter cold spells. The Western Balkans and Central Asia are especially vulnerable to potential natural gas shortages because of their infrastructure deficits and low levels of cross-border trade. Elec- trification and heat pumps, alongside bioenergy, are a cornerstone of decarbonization in the region’s buildings. Households need to commit 31 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION to energy-efficiency measures such as heating technology upgrades and improved thermal insulation. Traditional heating systems are used by a significant share of the population in ECA, reach- ing 45 percent of our sample (figure 13). These systems mostly refer to solid fuel stoves and boilers and to a lesser extent to electric heaters and fireplaces. By contrast, high-efficiency solid fuel stoves and boilers, natural gas boilers, split units, and heat pumps are considered modern heating systems. In EU countries and the South Caucasus traditional heating is less common, but is still used by around a third of the respondents. Türkiye and Kazakhstan have the lowest rates of traditional systems. District heating is used by a substantial part of the population in Kazakhstan and Romania; it is the least common system in the rest of the countries. Figure 13 Heating Technolog y Use by Type and C ountry Share of population by current heating technology Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% Share of respondents 75% 50% 25% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Doesn’t own/ District Modern Traditional Unsere/ Others heating heating heating Note: The figure presents the percentages of respondents as based on their answers to the question “What is the main system/device you currently use to heat your home?” In all countries, most respondents feel that their home is at least somewhat adequately insulated, but there is room for improvement, as one out of four respondents reported having inadequate or no insulation (figure 14). The lowest satisfaction with home insulation is found in Tajikistan and Türkiye and the highest satisfaction in Bosnia and Herzegovina and Kosovo. Insulation is an important factor to ensure energy efficiency in buildings. 32 Capturing household beliefs, preferences and attitudes in ECA Figure 14 A dequac y o f Home Insulation Level by C ountry How adequately insulate is your home Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% Share of respondents 75% 50% 25% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Don’t know/ No simulation Indequate Somewhat adequate Very adequate No answer present insulation insulation insulation Note: The figure shows the percentages of respondents based on their answers to the question “How adequately insulated is your home (thinking about ceiling and wall insulation and double pane windows to eliminate cold or warm air seeping into your home)?” In most countries, there is room for improvement in terms of energy efficiency, as only 30 percent of the respondents own an energy-efficient stove and 10.5 percent own solar panels (figure 15). Solar panels are most prevalent in the EU, Türkiye, and Serbia, where between 15 and 25 percent of respondents stated they own solar panels. In all other countries, solar panel use is rather limited, with shares of respondents reporting ownership below 5 percent. Air conditioners are most prevalent in Kazakhstan, Albania, and Bulgaria, where more than 50 percent of respon- dents own one, while in most other countries, they are owned by 25 to 45 percent of respondents. Energy-efficient stoves are most prevalent in Croatia and Bosnia and Herzegovina where half of respondents have one, while in most other countries the share is around 30 percent. 33 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Figure 15 O wnership o f Energ y -Efficient Technologies by C ountry Share of Population owning efficient stoves, air conditioners and solar panels by Country Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% Share of respondents 75% 50% 25% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Air Solar Energy Efficient Technology Conditioner Panels Stove Note: The figure shows the percentages of responses based on their answers to the question “Which of the following appliances do you own along with energy efficiency level?” More than 50 percent of those surveyed were willing to upgrade their heating system to a more efficient one (figure 16). The proportion is particularly high in Central Asia. This is also true for upgrading the current insulation levels at home. These are encouraging numbers, given the urgent need for sustainable heating solutions in the region due to the reliance on inefficient, car- bon-intensive fuels and the significant impact on household budgets and pollution levels (World Bank and ESMAP 2023). 34 Capturing household beliefs, preferences and attitudes in ECA Figure 16 Intention to Upgrade to Energ y -Efficient Technologies in the Nex t Two Years by C ountry Intentions to upgrade to energy efficient heating and technologies Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% Share of respondents 75% 50% 25% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Energy efficient Modern heating Solar Variable appliances or cooling systems Panels Note: The figure shows the proportions of people willing to upgrade technologies based on their responses to the question “Do you plan to upgrade your home’s heating to a more efficient system, improve home’s insulation and plan to install a solar panel?” The options available were 1. “Yes, within the next 2 years,” 2. “Yes, but not in the next 2 years,” 3. “No, I am not interested in upgrading,” 4. “I already have an efficient heating system,” and 5. “Don’t know.” Limited access to technology is a significant barrier to the upgrading of heating technology and insulation and to the installing of solar panels. With respect to financial constraints, having a lower income is not linked to a lower willingness to upgrade, but the income security provided by being a full-time worker does positively influence willingness with regard to each of the four options. The higher income group has a higher likelihood only in the case of willingness to pay for improved waste collection. Not surprisingly, people living in rented houses are less likely to consider any upgrade to their home, but do not differ from homeowners with respect to willingness to pay for improved waste collection. People living in urban areas are less likely to be willing to upgrade their heating system or install solar panels, although this result may be connected to a higher share of people living in apartment buildings where district heating is more common and the installation of solar panels is more complicated. People willing to change their lifestyle, risk takers, and those with more friends undertaking en- vironmentally friendly actions are generally more willing to upgrade to more-sustainable energy use (figure 17). Those willing to make lifestyle changes to protect the environment are generally 10 percent- age points more likely to be willing to upgrade their insulation or install solar panels and 6 percentage points more likely to upgrade their heating system. A risk-taking attitude as well as social influence are associated with a higher willingness (by 5 to 10 percentage points) to engage in these behaviors. 35 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Figure 17 Fac tors Influencing Intentions to Upgrade to Sustainable Technolog y a. Heating system Factors influencing willingness to upgrade heating system Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.10 -0.05 0.00 0.05 0.10 Marginal effects Note: The figure shows the coefficients from a probit regression where the dependent variable has a value of 1 if individuals are willing to update their technology within 2 years. This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value of 1 if a person is above the median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out to 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes than 5 out of their 1o closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. 36 Capturing household beliefs, preferences and attitudes in ECA b. Insulation Factors influencing willingness to upgrade insulation Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.10 -0.05 0.00 0.05 0.10 Marginal effects Note: The figure shows the coefficients from a probit regression where the dependent variable has a value of 1 if individuals are willing to update their technology within 2 years. This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value of 1 if a person is above the median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out to 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes more than 5 out of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. 37 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION c. Solar panels Factors influencing willingness to install solar panels Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.10 -0.05 0.00 0.05 0.10 0.15 Marginal effects Note: The figure shows the coefficients from a probit regression where the dependent variable has a value of 1 if individuals are willing to update their technology within 2 years. This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value of 1 if a person is above the median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out to 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes more than 5 out of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. Demographic characteristics play a marginal role in intentions to change behavior in favor of sustainability. This is in line with previous studies that show that the influence of sociodemo- graphic factors on intentions to reduce energy use is not significant (Abrahamse and Steg 2011). Education, which we noted above is a major determinant of awareness of subsidies, is not sig- nificantly connected with any of the intentions analyzed. Similarly, the age differences that were present with regard to awareness of programs and support for sustainable policies do not translate into differences with regard to intentions to change behavior. Also, the characteristics of a larger household or a household with children are mostly insignificant in determining these intentions as well. Women show a lower propensity to change their heating system or install solar panels, which is surprising, given their more pronounced proenvironmental attitude. This may indicate a more conservative behavior compared to men when it comes to innovative technologies. 38 Capturing household beliefs, preferences and attitudes in ECA Solid waste management A majority of people in the ECA region believe that recycling and sort- ing are the most effective actions to mitigate climate change and helping the environment. This is particularly true in Central Asia, Moldo- va, the South Caucasus, and the Western Balkans. In the EU responses were more scattered across energy efficiency and recycling. Only in Tajikistan and Türkiye does a substantial share of the population gives importance to energy efficiency. The ECA region generated 392 million tons of waste in 2016, or 1.18 kg per person each day (Kaza et al. 2018). It is the region with the second-highest annual waste generation rate, after East Asia and Pacific with 468 million tons. Globally, the highest per capita waste generators are a few island states with high levels of tourism and the economic hubs in Western Europe. In con- trast, the countries generating the least waste are largely in Eastern Europe or Central Asia and typically have a lower gross domestic product per capita. Across the ECA region, waste collection 39 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION coverage at 90 percent is relatively high, though the urban waste collection coverage rate of 96 percent greatly exceeds the rural waste collection rate of 55 percent. In urban areas, waste collec- tion typically occurs through a mix of door-to-door curbside collection and drop-off at centralized bins. However, there still exists important inequalities between rural and urban areas. Household behaviors that contribute to sustainable waste management practices involve reusing, reducing, recycling, and composting waste, as well as reducing litter. For solid waste management, less than half of the respondents in the Western Balkans, the South Caucasus, Moldova, and Kazakhstan reported they consistently separated waste. The practice is most prevalent in EU countries, Türkiye, Tajikistan, and Uzbekistan, where the major- ity of the population reported being in the habit of doing so although in most of these countries recycling is not enforced by the government. Poland, Romania, Bulgaria, and Türkiye are the only countries where some sort of separate waste streams collection has been implemented.9 The highest shares are found in Tajikistan and Türkiye, where three out of four respondents said they did separate waste collection most of the time (figure 18). Figure 18 S or ting and S eparate Disposal o f Waste in the Europe and C entral A sia Region by Subregion and C ountry, 2024 How often do you sort waste and dispose of it separately in their corresponding bin Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% 75% Share of respondents 50% 25% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Always Most of the time Occasionally Never Note: The figure shows proportion of people willing to sort waste based on their responses to the question “How often do you sort waste and dispose of it separately?” 9 Information extracted from: https://www.eionet.europa.eu/etcs/etc-ce/products/country-profiles-on-the-management-of-municipal-waste and Kaza et al. (2018). 40 Capturing household beliefs, preferences and attitudes in ECA As seen in figure 19, in most countries individuals are willing to pay a higher amount for better waste collection. In Central Asia, the South Caucasus, and Moldova, this is particularly true, with 62 percent people saying they would pay more. In the EU, the proportion is lower. This could be due to the existing waste collection service’s being satisfactory. Figure 19 Willingness to Pay More for Improved Waste S er vic es in the Europe and C entral A sia Region by Subregion and C ountry Would you be willing to pay more for any or improved waste collection? Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% 75% Share of respondents 50% 25% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Yes, I am willing to pay Yes, but not in the next 2 years more within the next 2 years Note: The figure shows the percentages of people willing to pay for improved waste collection based on their responses to the question “Would you be willing to pay more for any or improved waste collection?” The options available were 1. “Yes, within the next 2 years,” 2. “Yes, but not in the next 2 years,” 3. “No, I am not interested in upgrading,” 4. “I already have an efficient heating system,” and 5. “Don’t know.” Willingness to change lifestyle and more friends being involved in environmentally friendly actions matter most for willingness to pay for improved waste management, as was the case with technology adoption. As seen in figure 20, People willing to change their lifestyles are 12 percentage points more willing to pay for improved waste collection than the rest of the respon- dents. Furthermore, people who are risk loving and altruistic are more likely to be willing to pay more for a better waste collection service. This could be because altruistic individuals are better able to account for the social cost of better waste collection correctly. 41 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Demographic characteristics do not to play role in willingness to pay more for waste manage- ment. Only education, which mattered in relation to upgrading to energy efficient technologies, mat- ters in relation to waste management. Being highly educated and having full-time employment also help to improve the willingness to pay more. As one would expect, higher-income individuals are more likely to agree to pay for better waste collection. Other household characteristics such as household size and number of children are not significant when it comes to the expression of this intention. Figure 20 Fac tors that Influenc e Willingness to Pay More for Improved Waste Management S er vic e Factors influencing willingness to pay for improved waste collection Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.05 0.00 0.05 0.10 0.15 Marginal effects Note: The figure shows the coefficients from a probit regression where the dependent variable has a value of 1 if individuals are willing to pay more for waste management within 2 years. This is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value 1 if a person is above the median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value 1 if a person is willing to give above the median (out to 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes more than 5 out of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. Overall, the results suggest that people perceive solid waste management to be important for climate change. This explains a higher share of the population’s sorting of their waste and will- ingness to pay for improved waste collection. In particular, people with positive social influence, risk takers, and educated people express an increased willingness to pay for improved waste management services. 42 Capturing household beliefs, preferences and attitudes in ECA Barriers to the adoption of climate- friendly behaviors and technologies Financial constraints and lack of trust in sustainable technologies are the main barriers to the adoption of climate-friendly behavior. In Cen- tral Asia responses are more scattered, with excessive effort or hassle and limited access being more strongly perceived as barriers than in Eastern Europe. Lack of trust is particularly relevant in the EU and the Western Balkans (figure 21). 43 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Figure 21 Barriers to the A dop tion o f Environment-Friendly Technolog y in the Europe and C entral A sia Region by Subregion and C ountry, 2024 What is the main reason that makes adoption of climate friendly behavoir challenging? Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% 75% Share of respondents 50% 25% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Excessive effort Financial Lack of Lack of trust Limited or hassle constraints information in technology access Note: The figure shows the percentages of respondents citing different reasons when asked what made adoption of climate friendly behavior difficult. A single-choice question was used in order to capture what people tended to think of as the main barrier. Most respondents who lack the intention to engage in sustainable energy practices claim the main reason is the cost associated with the technology or service. Being too expensive is by far the main reason put forth for not having the intention to engage in sustainable energy practices, being offered by 35 to 45 percent of respondents depending on the specific behavior (figure 22). However, because income in most cases does not significantly influence intentions to upgrade, this means the uptake of behaviors is seen as expensive regardless of the actual income status of the respondents. The lowest-income group has a proportionally higher share of people citing financial costs as a barrier, but this factor is also the most prominent factor for medium- and higher-income groups. The major barriers to adoption are similar across all income groups. 44 Capturing household beliefs, preferences and attitudes in ECA Figure 22 Barriers to the A dop tion o f Energ y -Efficient Technologies and Waste Management Reasons not to engage with sustainable energy practices 50 45 40 35 30 25 20 15 10 5 0 Heating Insulation Solar panels Waste It is too expensive for me If I upgrade now, I will need to change I don’ have those options in my communi It involves too much hassle and effort I don’t know how to do it Others Note: The figure shows the proportions of people citing different reasons as to what makes adoption of climate friendly behavior difficult in response to the question “Why are you not interested in improving your appliances or modifying current practices in the next 2 years?” These findings indicate the necessity for support programs that emphasize the cost ben- efit of sustainable transitions. Such programs should provide clear guidelines on application procedures, technology management, and the short-, medium-, and long-term benefits of the upgrades. Additionally, they should showcase examples of how others (neighbors, influencers, etc) are successfully implementing these technologies and alternative practices. 45 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Information treatment Two different messaging strategies were tested in the survey and both led to increases in the percentage of respondents who intended to engage in sustainable behavior. Given the policy priority of sustainable energy in the ECA region, we designed the treatment to understand the best way to incentivize uptake of sustainable behavior. Respondents were divided into three groups, a control group who received no mes- sage and two treatment groups. The first treatment group received a descriptive-norm-activating message highlighting the share of people in their country who believed climate change will have a serious impact 46 Capturing household beliefs, preferences and attitudes in ECA during their lifetime, with the share being taken from the LITS IV. The second group received an awareness-raising message on the contri- bution of households to greenhouse gases emissions relative to other sources. Both treatments also included a sentence intended to motivate participants to take action by engaging in sustainable energy behavior. The control group was given no message and was asked if they would update their exist- ing heating technology, add solar panels, and improve the insulation of their homes. The two treatment groups received the following messages and were asked their willingness to adopt sustainable technologies and practices. • Treatment 1: Descriptive norms T1: “X percent of people in *Country considers that climate change will have serious im- pact in their lifetime. Helping mitigate environmental issues is important. Help [Country] by upgrading to more efficient heating devices, avoiding the use of cars, and reducing plastic waste!” • Treatment 2: Attribution of responsibility T2: “Households are among the top three contributors of polluting gases in Europe and Central Asia. Change your behaviors and make a difference in [country]. Help *Country by upgrading to more efficient heating devices, avoiding the use of cars, and reducing plastic waste!” Both treatments led to significantly higher percentages of persons expressing their inten- tion to engage in these actions, with treatment effects varying in size across the different actions between 3 and 6 percentage points. The highest impact was found for solar panels, with treated respondents being between 5 and 6 percentage points more likely to consider the upgrade in the future compared to the control group. The treatments also resulted in an increase of 4 percentage points in the propensity to upgrade heating system and of 2.5 percentage points to upgrade insulation. Between the two treatments, the first (leveraging social norms) had a slightly higher impact than the second (raising awareness), although the difference across the two treat- ments is in most cases lower than 1 percentage point and not significant. Figure 23 below reports the impact of the two treatment effects on the willingness to upgrade heating and insulation, install solar panels, or pay more for improved waste collection. 47 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Figure 23 Impac ts o f the Information Treatments Survey experiment results 85% 80% 75% 70% 65% 60% 55% 50% 45% 40% CG T1 T2 CG T1 T2 CG T1 T2 CG T1 T2 Heating Insulation Solar panels Waste Note: The figure shows probit regression coefficients on people’s willingness to upgrade technologies. The question was “Do you plan to upgrade your home’s heating to a more efficient system, improve home’s insulation and plan to install a solar panel?” The options available were 1. “Yes, within the next 2 years,” 2. “Yes, but not in the next 2 years,” 3. “No, I am not interested in upgrading,” 4. “I already have an efficient heating system,” and 5. “Don’t know.” For the probit, the dependent variable has a value of 1 if people answer with either 1 or 2 and 0 if they answer with 3 or 5. People who reported already having an updated technology were excluded from the regression. Overall, the ECA population presents individual, social, institutional, and infrastructure-re- lated barriers to sustainable behaviors. On the individual level, risk aversion seems to be a dis- positional trait that hinders lifestyle changes and limits support for climate programs. In addition, we found a generalized present bias, which entails the tendency to settle for a smaller present reward rather than wait for a larger future reward in a trade-off situation. In this case the bias was evidenced by the lack of willingness to pay for an upgrade that would yield savings in the future. On the social level, we found low awareness of support programs, a lack of social examples of sustainable behaviors, and social misconceptions, where people tend to believe others are less concerned about the environment than they actually are. On the institutional level, we found a lack of trust in institutions and technologies and finally at the system level, people reported limited access to technologies and support programs. 48 Capturing household beliefs, preferences and attitudes in ECA Policy recommendations This section presents recommendations drawn from the findings of this diagnostic study and informed by evidence from behavioral science. These solutions aim to address the key barriers to adopting sustainable systems such as upgrading a home’s heating systems, improving insulation, installing solar panels, and paying for improved waste collection. This study has identified several barriers associated with individual, social, institutional and infrastructural barriers. In this context, we present several ideas based on international evidence that are divided into communication-related and system-related solutions. Communication-related solutions consider behavioral tools that can 49 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION make communications clearer, more memorable, and more engaging, provided structural barriers have been resolved. These solutions target mainly individual and social barriers. Policy-related recommendations on the other hand focus on the contextual or structural-related factors that can be altered to facilitate the desired behavior. These solutions focus on physical and systemic barriers. Table 1 summarizes the diagnosed barriers and proposed recommendations, divided into individual, social, institutional or infrastructure related solutions. Table 1. Barriers to Sustainable Behaviors and their C orresponding S olutions Individual Social Institutional Infrastructure Barrier → Lack of Lack of social Social Risk aversion Present bias Lack of trust Lack of access Solution ↓ awareness examples misconceptions Communication-related solutions Use trusted messengers Use the preferred communication channel Leverage current social norms Clarify policy mechanisms Highlight future gains Leverage social pressure Harness innovative channels and platforms to address existing narratives Targeted messages System and program-level solutions Increase access sustainable choices Consider the distributional angle in the design and implementation of support programs Improve choice architecture Reduce sludge 50 Capturing household beliefs, preferences and attitudes in ECA 9.1) Communication-related solutions A. Use trusted messengers Academic and nongovernmental institutions are seen in most countries as highly trustworthy messengers of information related to climate change. Respondents in Central Asian countries, apart from Uzbekistan, have substantially higher trust in local and national government, while they are the least trustful of NGOs. Conversely, in Eastern Europe the majority of respondents in each country, with the exception of Kosovo, prefer academic institutions and NGOs as sources of information regarding climate change. Based on these results, messages and recommendations on programs and climate actions should come not only from the government, but should involve other types of authorities such as relevant academics and public figures from each country (fig- ure 24). A study on the persuasive features of messengers showed that likability, expertise, and relevance positively influence messenger trustworthiness, which in turn enhances the attitudes toward recycling and intentions to recycle among consumers (Jain et al., 2022). Figure 24 Trusted Messengers o f Climate Change Information Which of the following institution wpould you most trust with information about climate change? Central Asia European Union Moldova South Caucasus Türkiye Western Balkans 100% 75% Share of respondents 50% 25% 0% Kazakhstan Kyrgyz Rep. Tajikistan Uzbekistan Bulgaria Croatia Poland Romania Moldova Armenia Georgia Türkiye Albania BiH Kosovo Serbia Academic Commercial Local government National Non-governmental institutions banks (e.g. municipality) government institutions B. Use the preferred channels for different segments of the population Television and social media dominate as sources of information about climate change (figure 25). In most countries in the region, these two are followed by either websites or radio. In Albania 51 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION and Kosovo phone calls also seem to be an effective way of communicating about potential cli- mate change issues. Other channels include SMS, billboards, leaflets, emails, and phone calls. Figure 25 Pre f erred C ommunication Channels Preferred Communication Modes by Region (Proportional) 100% Percentage of Total Responses 75% 50% 25% 0% Central Asia European Union Moldova South Caucasus Türkiye Western Balkans Communication Mode Other channels Radio Social Media Television Websites Note: Respondents were asked to report the top three most-used communication channels. The percentages represent the frequency of selection of a specific channel. C. Raise awareness about the broad support for climate action using social norms Social norms can be a powerful tool to promote proenvironmental behavior. The survey results suggest people underestimate how much other people in their country are worried about climate change impacts. This has been named in the literature as “pluralistic ignorance” (Andre et al. 2024). Messages could leverage this information to show social concerns and influence individual behavior. To test the effectiveness of such messages, we conducted an experiment in which one group received a social norm message highlighting the share of other people in their country who believed climate change would have a serious impact during their lifetime and a control group, who did not receive such a message. The message resulted in the members of the treatment group’s expressing the intention to upgrade heating or insulation, install solar panels, and pay for improved waste management at a much higher rate compared to the control group. A relevant policy tool consists of social comparisons interventions that highlight other peo- ple’s proenvironmental behaviors or attitudes in order to increase proenvironmental behaviors of individuals. One successful example is from the United States, where a company called OPOW- ER sent “Home Energy Report” letters to residential utility customers in which their electricity use was compared to that of their neighbors (figure 26). As a result energy consumption was 52 Capturing household beliefs, preferences and attitudes in ECA reduced by 2 percent (Allcott 2011). Our study tested this type of intervention through a message embedded in a survey. We found that highlighting that other people in a person’s country agree with taking action in relation to climate change issues did impact positively people’s willingness to upgrade to and pay for sustainable devices and services. Figure 26 “Home Energ y Repor t ” with S ocial Norms Source: Jachimowicz et al. (2018). D. Highlight future gains over present costs The main reported barrier for all income segments was financial constraints. Behavioral science has shown that people tend to show a present bias,10 whereby they prefer present benefits over future ones. To counteract this bias, communications need to stress the benefits and convenience of investing today in sustainable systems and practices to obtain higher rewards in the future. This is especially relevant for risk-averse households who tend to show less willingness to change their lifestyles for the benefit of the environment. We did a survey experiment in Central Asia where we observed that simply highlighting the annual savings from adopting an energy-efficient technology improved the willingness to adopt the technology by 20 percent. In another example where the goal was to promote uptake of solar panels, Google developed “The Solar API,”11 which contains information such as the size and slope of the roof and the modeled energy production of a rooftop array (figure 27). This information can be used to assess 10 The present bias refers to the tendency of people to give stronger weight to payoffs that are closer to the present time when considering trade-offs between two future moments (O’Donoghue and Rabin 1999). For example, a present-biased person might prefer to receive 10 dollars today over receiving 15 dollars tomorrow, but would not mind waiting an extra day if the choice were for the same amounts one year from today versus one year and one day from today. 11 For more information, see https://mapsplatform.google.com/resources/blog/powering-future-our-new-solar-api/. 53 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION the potential benefits of installing solar and can be used by solar professionals to help the owners of homes and builders explore and compare different solar configurations. Figure 27 “ The S olar API” Source: https://mapsplatform.google.com. Note: The API provides detailed information to help households understand the potential benefits of installing. E. Clarify policy mechanisms The survey result confirm that although most of the population in the ECA region is aware of climate policies in their countries, an average of 24 percent do not trust sustainable technolo- gies. Previous research has shown that educational videos with clarifying policy mechanisms can increase support for climate policies (Dechezleprêtre et al. 2022). Communications, thus, should clarify the purpose of support programs and the potential impacts if the population participates in them. F. Leverage different types of information or messages The content of the information is relevant to the promotion of the willingness to change behav- ior. There is limited understanding on the economic returns of different energy-efficient options. Clear and concise messaging focusing on the behavior to be changed is necessary. For example, in Zhang et al. (2018), the authors tried to solidify the awareness of the scientific consensus in the United States on climate change with a simple, clear message: “Ninety-seven percent of climate scientists have concluded that human-caused global warming is happening.” This led to 54 Capturing household beliefs, preferences and attitudes in ECA improvement in the perception of the degree of the scientific consensus on climate change by 16 percentage points nationally. The messages also need to be tailored to specific population segments. For example, our results suggest that tenants are less willing to upgrade their heating systems. Providing specific types of information for homeowners and renters can improve the take-up of these programs for these segments of the population. This is especially relevant for Türkiye, where 42.8 percent of our sample rent their homes. In all other countries the proportion of renters varied from 10 to 18 percent. G. Harness innovative channels and platforms to address existing narratives surrounding climate Using nontraditional engagement platforms can influence the climate narrative in meaningful ways, including by addressing intergenerational concerns. Engaging with students to promote the intergenerational transfer of beliefs, attitudes and behaviors can address persistent attitudes among rigid demographic groups. School competitions provide an opportunity for students to en- gage around climate issues in a meaningful way and bring other community and family members in the exercise. Such competitions challenge students to come up with ideas on how to support green transitions. The ideas could be put to a vote in the larger society and outreach on the pro- cess and winners could encourage engagement on the issue among nonparticipants (family and community members, business leaders, etc.). Such interventions can have a meaningful impact on decision-makers in the energy transition: a study conducted in the United States showed that providing climate curricula to students had a significant impact on parents’ climate change con- cerns, particular for those with more-conservative values (Lawson et al. 2019). Harnessing digital citizen engagement platforms is another potentially low-cost, high-im- pact opportunity to promote inclusion in the dialogue around climate at the country or even local level. Citizen engagement platforms can encourage dialogue and action on green transi- tions in meaningful ways, including by sparking open discussion that includes typically excluded stakeholders on viable solutions in the green transition and encourage ideation. In Cyprus,12 a World Bank team designed an online civic engagement platform that connected people and de- cision-makers and empowered them to interact (via automatic translation), ideate, and decide jointly how to solve societal problems at the neighborhood, village, city, or island level. The first campaign reached over 200,000 people and an impact evaluation found that participation led to increases in interpersonal trust. 12 https://www.cypruspusula.org/ 55 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION 9.2) System and program-level solutions Around 15 percent of those surveyed mentioned that they lacked access to sustainable tech- nologies and infrastructure. The percentage was slightly higher in Central Asian countries and Moldova. An important action that needs to accompany communicational efforts is the increasing of access by simplifying the steps and requirements for and developing infrastructure to support sustainable practices. This involves developing separate collection streams, promoting a circular economy, and creating a market for solar panels and sustainable heating devices. A. Use “choice architecture” to constrain unsustainable behaviors Prepaid meters and smart meters for energy consumption13 and different-sized bins for mixed and sorted waste are some examples that have shown positive results with regard to constrain- ing service use when access is guaranteed and where low motivation or financial constrains are barriers to behavior change. This type of intervention is typically referred to as “choice ar- chitecture,”14 where specific behaviors are favored over others based on the context’s design. For example, in South Africa postpaid electricity meters were replaced with prepaid versions. In- home displays provided timely feedback on usage as this increased the salience of electricity costs. Prepaid meters reduced electricity consumption 12−15 percent and increased revenue recovery at lower costs, making it more feasible for providers to offer electricity (Jack and Smith 2016). In the waste management space, in Romania reducing the size of general waste bins and increasing the size of bins for recyclables nudged people to sort their waste to avoid overflowing bins (Ionkova et al. 2023; figure 28). 13 Smart meters, which are different from prepaid meters, can also be an interesting option to add to reduce energy consumption/improve energy efficiency. They provide detailed, real-time consumption data to both users and utility providers, often via online portals or apps. They feature two-way communication, enabling remote monitoring, billing, and even disconnection or reconnection by the utility provider. 14 The term “choice architecture” coined by Thaler and Sunstein (2008) refers to the practice of influencing choice by “organizing the context in which people make decisions” (Thaler et al. 2013, 428; see also nudge). A frequently mentioned example is how food is displayed in cafeterias, where offering healthy food at the beginning of the line or at eye level can contribute to healthier choices. 56 Capturing household beliefs, preferences and attitudes in ECA Figure 28 Diff erentiation o f C ontainers for Waste in Romania Source: Ionkova et al. (2023). Note: The different colors and sizes provided physical cues to guide waste disposal behavior and prioritize recycling and segregation over residual waste disposal. B. Consider the distributional angle in the design and implementation of support programs The design and implementation of support programs should consider the unique barriers faced among low-income groups to access and use climate-friendly technologies. The lowest-income group across ECA countries has a proportionally higher share of people citing financial costs as a barrier to upgrading technology or changing their behaviors. The failure to address distributional concerns in the design of support programs can lead to an inefficient allocation of resources (for example, support programs’ primarily providing financial benefits to those who face few- er affordability constraints). These programs should consider the distributional implications of subsidy and grant amounts and for example vary grant amounts by level of income, such that the lowest-income groups receive the highest amount of financing to support the adoption of more-climate-friendly technologies. The design of support programs should also consider the preferred communication channels of low socioeconomic households, use language that is easy to understand, and provide transparency on the upfront costs that may act as barriers to invest- ment. This is particularly important, given that awareness of these support programs is generally lower for lower-income groups. 57 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION C. Reduce sludge in accessing support programs Support programs can fall short of their objectives if they are not communicated well (as de- scribed above) but also if accessing them is a challenge in and of itself. Many support programs are less accessible (or even inaccessible) because of complex processes, platforms, and cogni- tive-overloading outreach. “Sludge,” a term used to refer to intentional and unintentional frictions in choice architecture, can constrain access to support programs and even reduce demand for the benefits. Such processes, platforms, and communications can be simplified and made more user-friendly by reducing the number of steps that need to be taken and eliminating opportunities for favoritism in the allocation of benefits. Preparing a how-to practical guide can increase uptake of energy-efficient technologies. ECA already has relevant climate programs that could benefit from this perspective, such as the Strengthening Financial Resilience and Accelerating Risk Re- duction in Central Asia (SFRARR) Program,15 the Climate Adaptation and Mitigation Program for Aral Sea Basin (CAMP4ASB),16 and the ECA Climate Roadmap mentioned previously in this report. For the past six years, the World Bank team has been supporting the government of Poland in the design, implementation, and financing of the Clean Air Priority Program (CAPP). A behav- ioral diagnostic was conducted that pointed to the complexity of accessing program benefits and hassle factors associated with sustainable-heating upgrades, as well as ingrained beliefs about heating. Insights from the diagnostic (figure 29) were operationalized by the government, who simplified program structures and application processes (to make the program more appealing to lower-income households in particular) as well as engaging supply side actors and local gov- ernments to improve the implementation of the program.17 15 For more information, see: https://www.gfdrr.org/en/program/SFRARR-Central-Asia. 16 For more information, about this project see: https://projects.worldbank.org/en/projects-operations/project-detail/P151363. 17 https://blogs.worldbank.org/en/climatechange/clean-air-and-heating-choices-how-change-homeowners-behavior-poland. 58 Capturing household beliefs, preferences and attitudes in ECA Figure 29 The Clean Air Priority Program (C APP) Bene ficiary Journey in Poland Black Gold. Blessing or Burden? Understanding Behavioral Barriers to Switch to New Heating Technologies in Poland IT’S SUBSIDIZED BY IT’S EASY! THE GOVERNMENT g just fine, I'll d orkin o it is w in fi ler boi y neighbor doin ve y ears old c oal is m g ?" ” “My hat "W 1 et” 2 ut e y CONSIDERING d b ng oo cha LEARNING ABOUT REPLACING THE BOILER k s g to y loo ad THE PROGRAM “It t re no I am tion looks complicated!” applica “This sulation? Windows? ers? In "Boil nt? Forms? Agh!!!" Gra 3 ” ings sav he SETTING-UP ny ll t e a td a ” THE APPLICATION hav rstan forms t d ae ’ un lo n 4 l ’t ia I d on on anc FINANCING d n “I “ fi APPLICANT’S SHARE "Winter i s over, d o I re a l l y ne e d t his?" 5 WAITING FOR APPROVAL e?" mor ges har ake” ut or c mist so ll a 6 a ck a sa or b is w ct th STARTING ink g t tra THE PROJECT tin con th MANUAL st if the o am at ar “I "Wh The replacement of coal-fired boilers can help protect the 7 health and save the lives of ADAPTING TO THE NEW TECHNOLOGY thousands of people in Poland. Source: World Bank (2020). 59 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Conclusion Countries in the ECA region are significantly exposed to climate-relat- ed risks and possess limited capacity for adaptation. The most impov- erished nations in the region are particularly vulnerable as they face the most severe threats from climate change, with adverse impacts being unevenly distributed and poorer regions suffering the most. Addition- ally, the legacy of the “unfinished transition” following the dissolution of the Soviet Union has resulted in a more constrained private sector and weaker public institutions in many ECA countries compared to their counterparts in other regions. 60 Capturing household beliefs, preferences and attitudes in ECA Our survey shows that individuals are aware of these climate-change risks and are willing to adopt environmentally friendly behaviors. Looking at the status quo, however, a significant share of the population is dependent on fossil fuels and 45 percent use traditional stoves for heating. Similarly, only 30 percent of the respondents own an energy-efficient stove and around 10.5 percent have solar panels. About half of the population does not consistently sort domestic waste. Tailored policies based on country conditions would help place ECA on a path of sustainable and greener long-term growth. These policies should account for the costs of implementation and the trade-offs between immediate environmental objectives and long-term social benefits. In addition to pricing reforms, especially the reduction or phase-out of explicit energy subsidies, the private sector needs to be motivated to implement decarbonization measures. Complex challenges can involve more than one behavior that needs to change. Because behaviors tend to be interrelated, a thorough analysis of how the system works needs to be done before planning an intervention. Addressing systemic gaps with a behavioral lens can help gov- ernments to move the climate agenda forward and improve the take-up of public programs. 61 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION 62 Capturing household beliefs, preferences and attitudes in ECA References Abrahamse, Wokje, and Linda Steg. 2011. “Factors Related to Household Energy Use and Intention to Reduce It: The Role of Psychological and Socio-Demographic Variables.” Human Ecology Review 18 (1): 30−40. Allcott, Hunt. 2011. “Social Norms and Energy Conservation.” Journal of Public Economics 95 (9−10): 1082−95. Allcott, Hunt, and Judd B. Kessler. “The welfare effects of nudges: A case study of energy use social comparisons.” American Economic Journal: Applied Economics 11, no. 1 (2019): 236-276. Andre, Peter, Teodora Boneva, Felix Chopra, and Armin Falk. 2024. “Globally Representative Evidence on the Actual and Perceived Support for Climate Action.” Nature Climate Change 14 (3): 253−9. Bergquist, M., A. Nilssona, and W. Schultz. 2019. “A Meta-Analysis of Field-Experiments Using Social Norms to Promote Proenvironmental Behaviors.” Global Environmental Change 59: 10194. Bush, Sarah Sunn, and Amanda Clayton. 2023. “Facing Change: Gender and Climate Change Attitudes Worldwide.” American Political Science Review 117 (2): 591−608. Cojocaru, Alexandru, Michael Lokshin, and Arina Nikandrova. 2024. “Questioning the Climate Change Age Gap.” Policy Research Working Paper 10918, World Bank, Washington, DC. Cologna, Viktoria, and Michael Siegrist. 2020. “The Role of Trust for Climate Change Mitigation and Adaptation Behaviour: A Meta-Analysis.” Journal of Environmental Psychology 69: 101428. Dechezleprêtre, Antoine, Adrien Fabre, Tobias Kruse, Bluebery Planterose, Ana Sanchez Chico, and Stefanie Stantcheva. 2022. “Fighting Climate Change: International Attitudes toward Climate Policies,” OECD Economics Department Working Papers 1714, OECD Publishing, Paris, https://doi.org/10.1787/3406f29a-en. 63 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Echavarren, J. 2017. “From Objective Environmental Problems to Subjective Environmental Concern: A Multilevel Analysis Using 30 Indicators of Environmental Quality.” Society and Natural Resources 30 (2): 145−59. Fairbrother, M., I. J. Sevä, and J. Kulin. 2019. “Political Trust and the Relationship between Climate Change Beliefs and Support for Fossil Fuel Taxes: Evidence from a Survey of 23 European Countries.” Global Environmental Change 59: 102003. Gray, S., K. Raimi, R. Wilson, and J. Árvai. 2019. “Will Millennials Save the World? The Effect of Age and Generational Differences on Environmental Concern.” Journal of Environmental Management 242: 394−402. Gür, Nurullah. 2020. “Does Social Trust Promote Behaviour Aimed at Mitigating Climate Change?” Economic Affairs 40 (1): 36−49. Gynther, L., I. Mikkonen, and A. Smits. 2012. “Evaluation of European Energy Behavioural Change Programmes.” Energy Efficiency 5 (1): 67−82. Hallsworth, Michael. “A manifesto for applying behavioural science.” Nature Human Behaviour 7, no. 3 (2023): 310-322. Ionkova, K. M., A. A. Tomio, E. De Roy, S. Yadav, S. Kaza, J. G. Karver, and R. Vakis. Behavior Change in Solid Waste Management—A Compendium of Cases. Washington, DC: eMBeD World Bank. http://documents.worldbank.org/curated/en/099091423124016666/P177344 1302811082184c8156db86923f14. Jachimowicz, J. M., O. P. Hauser, J. D. O’Brien, E. Sherman, and A. D. Galinsky. 2018. “The Critical Role of Second-Order Normative Beliefs in Predicting Energy Conservation.” Nature Human Behaviour 2 (10): 757−64. Jack, B. K., and G. Smith. 2016. “Charging Ahead: Prepaid Electricity Metering in South Africa.” National Bureau of Economic Research Working Paper 22895, http://www.nber.org/papers/ w22895. Jain, Ritesh, Edwina Luck, Shane Mathews, and Lisa Schuster. “Creating Persuasive Environmental Communicators: Spokescharacters as Endorsers in Promoting Sustainable Behaviors.” Sustainability 15, no. 1 (2022): 335. Johansson, K. 2016. “Understanding Recycling Behavior: A Study of Motivational Factors behind Waste Recycling.” WIT Transactions on Ecology and the Environment 202: 401−14. 64 Capturing household beliefs, preferences and attitudes in ECA Kaza, Silpa, Lisa Yao, Perinaz Bhada-Tata, and Frank Van Woerden. 2018. What a Waste 2.0: A Global Snapshot of Solid Waste Management to 2050. Washington, DC: World Bank Publications. https://hdl.handle.net/10986/30317. Kola-Bezka, M., and K. Leki. 2024. “Household Energy Behaviour in the Times of Crisis: Lessons for Policy Initiatives from Peripheral, Fossil-Dependent Regions of the European Union.” Energy Policy 188: 114113. Lampietti, Julian A., Sudeshna Ghosh Banerjee, and Amelia Branczik. 2007. People and Power: Electricity Sector Reforms and the Poor in Europe and Central Asia. Directions in Development, Energy and Mining. Washington, DC: World Bank. http://hdl.handle. net/10986/717. Lawson, D. F., K. T. Stevenson, M. N. Peterson, S. J. Carrier, R. L. Strnad, and E. Seekamp. 2019. “Children Can Foster Climate Change Concern among Their Parents.” Nature Climate Change 9: 458−62. Minelgaitè, Audronè, and Genovaitè Liobikienè. 2019. “Waste Problem in European Union and its Influence on Waste Management Behaviours.” Science of the Total Environment 667: 86−93. Milfont, T., E. Zubielevitch, P. Milojev, and C. Sibley. 2021. “Ten-Year Panel Data Confirm Generation Gap, But Climate Beliefs Increase at Similar Rates across Ages.” Nature Communications 12: 4038. Miliute-Plepiene, Jurate, Olle Hage, Andrius Plepys, and Algirdas Reipas. 2016. “What Motivates Households Recycling Behaviour in Recycling Schemes of Different Maturity? Lessons from Lithuania and Sweden. Resources, Conservation and Recycling 113: 40−52. Nielsen, Kristian S., Viktoria Cologna, Jan M. Bauer, Sebastian Berger, Cameron Brick, Thomas Dietz, Ulf JJ Hahnel et al. “Realizing the full potential of behavioural science for climate change mitigation.” Nature Climate Change 14, no. 4 (2024): 322-330. O’Donoghue, T., and M. Rabin. 1999. “Doing It Now or Later.” American Economic Review 89 (1): 103−24. Olivier, Anne, Caterina Ruggeri Laderchi, and Christopher P. Trimble. 2013. Balancing Act: Cutting Energy Subsidies while Protecting Affordability. Washington, DC: World Bank Publications. 65 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Ong, Corinne, Lyle Fearnley, and Siow Boon Chia. 2019. “Towards a Sustainable Future: A Holistic Inquiry of Waste Management Behaviors of Singapore Households.” International Journal of Sustainable Development & World Ecology 26 (7): 583−96. Park, Sang-June, Sungchul Choi, and Eun-Jeong Kim. 2012. “The Relationships between Socio- Demographic Variables and Concerns about Environmental Sustainability.” Corporate Social Responsibility and Environmental Management 19 (6): 343−54. Poortinga, W., L. Whitmarsh, L. Steg, G. Böhm, and S. Fisher. 2019. “Climate Change Perceptions and Their Individual-Level Determinants: A Cross-European Analysis.” Global Environmental Change 55: 25−35. Raghu, Sumana Jagadeshi, and Lewlyn L. R. Rodrigues. 2020. “Behavioral Aspects of Solid Waste Management: A Systematic Review.” Journal of the Air & Waste Management Association 70 (12): 1268−1302. Sargisson, Rebecca J., Judith I. M. De Groot, and Linda Steg. 2020. “The Relationship between Sociodemographics and Environmental Values across Seven European Countries.” Frontiers in Psychology 11: 2253. Stuggins, Gary, Alexander Sharabaroff, and Yadviga Semikolenova. 2013. Energy Efficiency: Lessons Learned from Success Stories. Washington, DC: World Bank. Sunstein, Cass R., Cass Sunstein, and Richard H. Thaler. Nudge. Vuibert, 2022. Thaler, R. H., Sunstein, C. R., & Balz, J. P. (2013). Choice architecture. In E. Shafir (Ed.), The behavioral foundations of public policy (pp. 428-439). Princeton, NJ: Princeton University Press. Torgler, Benno, María A. García Valiñas, and Alison Macintyre. 2008. “Differences in Preferences towards the Environment: The Impact of a Gender, Age and Parental Effect.” FEEM Working Paper 18, Fondazione Eni Enrico Mattei, Milan. Vlasceanu, Madalina, Kimberly C. Doell, Joseph B. Bak-Coleman, Boryana Todorova, Michael M. Berkebile-Weinberg, Samantha J. Grayson, Yash Patel et al. “Addressing climate change with behavioral science: A global intervention tournament in 63 countries.” Science advances 10, no. 6 (2024): eadj5778. Wang, Yixuan, Xingle Long, Liang Li, Qinglin Wang, Xiping Ding, and Sijia Cai. 2021. “Extending Theory of Planned Behavior in Household Waste Sorting in China: The Moderating 66 Capturing household beliefs, preferences and attitudes in ECA Effect of Knowledge, Personal Involvement, and Moral Responsibility.” Environment, Development and Sustainability 23: 7230−7250. http://documents.worldbank.org/curated/ en/099092023140527206/P1777440fed3230ce089060ff8ce59c9f5e World Bank. 2024a. Net Zero Energy by 2060: Charting the Path of Europe and Central Asia toward a Secure and Sustainable Energy Future. Washington, DC: World Bank. https://hdl. handle.net/10986/41089. World Bank. 2024b. Unleashing The Power Of The Private Sector. Europe and Central Asia Economic Update. Washington, DC: World Bank, Office of the Chief Economist. World Bank and ESMAP. 2023. Toward a Framework for the Sustainable Heating Transition in Europe and Central Asia (English). Washington, DC: World Bank. http://documents.worldbank.org/curated/en/099092023140527206/ P1777440fed3230ce089060ff8ce59c9f5e. Zaikova, Anna, Ivan Deviatkin, Jouni Havukainen, Mika Horttanainen, Thomas Fruergaard Astrup, Minna Saunila, and Ari Happonen. 2022. “Factors Influencing Household Waste Separation Behavior: Cases of Russia and Finland.” Recycling 7 (4): 52. Zhang, B., S. van der Linden, M. Mildenberger, J. Marlon, P. Howe, and A. Leiserowitz. 2018. “Experimental Effects of Climate Messages Vary Geographically.” Nature Climate Change 8: 370−4. 67 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Appendix Appendix A The survey and sample The findings in this report are based on a cross-country analysis of population beliefs, at- titudes, and behaviors as captured through a quantitative survey implemented across 16 countries of the ECA region. The selection of countries was purposeful: the goal was to capture representative countries across the various subregions of ECA so that generalizations for the re- gion could be made.18 The survey instrument was designed to collect data that would provide accurate pictures of adults in each of the 16 countries, striving for regional representativeness within countries. In the remainder of this section, we discuss details of the survey, the forms it took according to local conditions, and characteristics of the respondents. Three methods were used: Computer Assisted Webpage Interviewing (CAWI), implemented in Poland, Romania, Serbia, Croatia, Bulgaria, and Turkiye; Computer Assisted Telephonic Interviewing (CATI), implemented in Moldova, Albania, Kosovo, Georgia, Armenia, Bosnia and Herzegovina, Uzbekistan, Tajikistan, Kazakhstan, and the Kyrgyz Republic; and Computer Assisted Personal Interviewing (CAPI), im- plemented in Uzbekistan, Tajikistan, Kazakhstan, and the Kyrgyz Republic. A. Survey data collection and sample Data Collection: The survey data were collected between May and August, 2024, with 13.0 per- cent captured through CAPI, 54.2 percent through CATI, and 32.8 percent were administered through CAWI. The data collection timeline encountered several postponements, attributed to various fieldwork challenges. These included accessibility issues in remote and weather-affect- ed areas and instances of unsuccessful interviews. Despite these hurdles, the survey sought to gather timely and relevant insights within the adjusted project timeline. Sample: The survey focused on a nationally representative sample of adults in selected countries in the ECA region, covering urban and rural areas to ensure a comprehensive under- standing of household variabilities. Using probabilistic sampling, the survey was administered to a total of 16,500 households: in Eastern Europe, a total of 10,302 households in Albania, Armenia, Bosnia and Herzegovina, Bulgaria, Croatia, Georgia, Kosovo, Moldova, Poland, Romania, Serbia, 18 The selection of countries in Eastern Europe was sensitive to the current geopolitical situation involving Russia, and thus the Russian Federation, Ukraine and Belarus were excluded from the study. 68 Capturing household beliefs, preferences and attitudes in ECA and Türkiye were covered; and in Central Asia, a total of 6,198 households across Uzbekistan, Tajikistan, Kazakhstan, and the Kyrgyz Republic were covered. Samples were designed to be representative of the national adult population with respect to age, gender, and urban-rural residency. Within each stratum, primary sampling units were randomly selected across the municipalities of each country, ensuring that an equal number of interviews were conducted in each municipality. We sampled 800 respondents per country in Eastern Europe, totaling 10,302 participants across the 12 countries, and 1600 respondents per country in Central Asia, totaling 6198 participants across the 4 countries. The selection was based on administrative regions and adult population data ensured a representative sample within each country. Random sampling methods were employed to minimize bias and enhance the general- izability of findings. Most of the surveyed population were between 30 to 65 years old. Almost half of the sample worked full-time and had either secondary or tertiary education. The share of people owning their homes was also dominant (figure 30). Figure 3 0 S ociodemographics o f the Sur vey S ample Descriptive statistics Age Group Education Employment Gender 60% 50% 60% 40% 40% 30% 40% 40% 30% 20% 20% 20% 20% 10% 10% 0% 0% 0% 0% Under 30 30 to 65 Over 65 Lower Primary Secondary BSc/MAster PhD NA Full-time Part-time Retired Other inactive NA Female Male NA Share of Total Hose Type Income Renting Residence 40% 40% 60% 60% 30% 30% 40% 40% 20% 20% 20% 20% 10% 10% 0% 0% 0% 0% Detached Semi-detached Apartment (<10) Apartment (>10) NA Low Medium High NA Yes No Urban Rural Note: The figure shows the distribution of our sample (a total of 16,500 households) across 16 ECA countries. 69 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Table A1. Summary o f S ample Numbers, Data C ollec tion Dates, and Methodolog y Country Collected data point Survey started Survey finished Method Albania 809 April 08, 2024 May 05, 2024 CATI Armenia 800 April 22, 2024 April 30, 2024 CATI BIH 807 April 05, 2024 April 29, 2024 CATI Bulgaria 957 April 11, 2024 April 19, 2024 CAWI Croatia 801 April 04, 2024 May 05, 2024 CAWI Georgia 812 April 09, 2024 April 29, 2024 CATI Kosovo 825 April 13, 2024 May 03, 2024 CATI Moldova 842 April 16, 2024 April 29, 2024 CATI Poland 846 April 04, 2024 April 24, 2024 CAWI Romania 1,004 April 12, 2024 April 25, 2024 CAWI Serbia 804 April 04, 2024 May 01, 2024 CAWI Türkiye 995 April 16, 2024 April 18, 2024 CAWI Tajikistan 1,552 June 26, 2024 August 20, 2024 CATI/CAPI Uzbekistan 1,555 July 01, 2024 August 29, 2024 CATI/CAPI Kyrgyz Rep. 1,514 June 09, 2024 July 29, 2024 CATI/CAPI Kazakhstan 1,577 June 13, 2024 August 28, 2024 CATI/CAPI TOTAL 16,500 B. Descriptive statistics The questionnaire The questionnaire was structured into six main sections: sociodemographics, climate per- ceptions, support for government policies, energy efficiency and home heating, solid waste management, and information treatment. We measured willingness to upgrade technology and change behaviors in relation to energy efficiency and home heating and solid waste management. For the Central Asia countries, the survey also sought to understand willingness to pay (WTP) for enhancements to heating appliances. We also captured household characteristics and basic demographic information of the respondents. Sociodemographics: We captured an individual’s gender, age, type of household, income, employment, and decision-making roles. We also measured psychological predispositions such as risk taking, prosociality, and trust. 70 Capturing household beliefs, preferences and attitudes in ECA Climate change perceptions: We captured individuals’ beliefs concerning climate change and its impacts. The questions focused on both primary beliefs (for example, whether the world is warmer than it was 100 years ago) and secondary beliefs (for example, what most citizens in a country believed about the seriousness of the impact of climate change). In addition, we asked whether the government of the country in which an individual lived frequently overstated the impacts of climate change. Finally, we measured willingness to act depending on the climate perceptions. Support for government policies: We sought to elicit individuals’ support for government policies to mitigate climate change. These included four main policy types: taxes, an absolute ban on polluting technologies, incentive schemes (such as subsidies), and provision of sustain- able infrastructure. Energy efficiency and heating: We captured the current use of energy appliances and heat- ing behavior. Respondents made of a range of energy types, from solid fuels to more-sustainable ones. We also measured willingness to change current behaviors. Solid waste management: We captured current practices on waste burning and sorting among respondents. This was followed by eliciting whether respondents were willing to change their behavior and pay more for a better waste management system. Information experiment: We explored whether different ways of presenting information could change people’s views and shift actions in a more sustainable direction. The control group did not get any information and the two treatment groups were presented with messages highlighting either social norms or attribution of responsibility. The two messages were 1) country-specific social norms in relation to the seriousness that the impact of climate change implies and 2) the household contribution (as a percentage) to pollution in Europe. 71 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Table A2. Summary o f Descrip ti ve Statistics o f the S ample Weighted sample Unweighted sample N % N % N of observations 16,500 16,500 Gender Male 8,019 48.70 7,639 46.40 Female 8,431 51.30 8,829 53.60 Age group Between 35 and 55 7,590 46.00 7,226 43.80 Under 35 6,936 42.00 5,915 35.80 Over 55 1,974 12.00 3,359 20.40 Income group Medium 6,838 47.30 6,258 44.60 Low 2,883 19.90 3,747 26.70 High 4,737 32.80 4,018 28.70 Education Primary (or less) 980 6.00 1,043 6.30 Secondary 8,427 51.40 9,040 54.90 Tertiary (BA, MA, PhD) 6,988 42.60 6,375 38.70 Employed 0.589 0.492 0.572 0.495 Household size Small/medium 12,852 77.90 12,282 74.40 Large (>5 members) 3,648 22.10 4,218 25.60 Children No 5,994 36.30 6,546 39.70 Yes 10,506 63.70 9,954 60.30 Renting home No 12,571 76.20 13,663 82.80 Yes 3,929 23.80 2,837 17.20 Area Rural 6,364 38.60 6,735 40.80 Urban 10,136 61.40 9,765 59.20 House Type Detached 8,744 53.40 10,262 62.40 Semidetached 644 3.90 548 3.30 Apartment 6,994 42.70 5,644 34.30 72 Capturing household beliefs, preferences and attitudes in ECA Table A3. Experiment S ample Distribution Survey experiment samples Control group Treatment 1 Treatment 2 N (weighted) 64,476,279 (30.5%) 64,306,953 (30.5%) 82,389,642 (39.0%) Heating upgrade No 1,391 (37.3%) 1,329 (35.4%) 1,660 (33.7%) Yes 2,338 (62.7%) 2,427 (64.6%) 3,264 (66.3%) Insulation upgrade No 1,181 (34.9%) 1,234 (35.4%) 1,410 (31.7%) Yes 2,202 (65.1%) 2,250 (64.6%) 3,041 (68.3%) Solar panels No 2,591 (56.0%) 2,550 (53.8%) 3,160 (51.6%) Yes 2,033 (44.0%) 2,193 (46.2%) 2,959 (48.4%) Improved waste man. No 1,824 (42.8%) 1,685 (39.6%) 2,139 (39.7%) Yes 2,435 (57.2%) 2,575 (60.4%) 3,251 (60.3%) Gender Male 2,446 (49.3%) 2,520 (50.2%) 3,052 (47.2%) Female 2,520 (50.7%) 2,495 (49.8%) 3,416 (52.8%) Age group Between 35 and 55 2,244 (45.0%) 2,465 (49.1%) 2,881 (44.4%) Under 35 2,080 (41.7%) 2,010 (40.0%) 2,847 (43.9%) Over 55 661 (13.3%) 550 (10.9%) 762 (11.7%) Income group Medium 2,017 (45.6%) 2,079 (46.9%) 2,743 (49.0%) Low 925 (20.9%) 953 (21.5%) 1,003 (17.9%) High 1,485 (33.5%) 1,404 (31.6%) 1,848 (33.0%) Education Primary (or less) 299 (6.0%) 338 (6.8%) 342 (5.3%) Secondary 2,479 (50.1%) 2,404 (48.1%) 3,545 (55.0%) Tertiary (BA, MA, PhD) 2,168 (43.8%) 2,255 (45.1%) 2,563 (39.7%) 73 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION Survey experiment samples Control group Treatment 1 Treatment 2 Employed 0.605 (0.489) 0.594 (0.491) 0.574 (0.495) Household size Small/medium 3,912 (78.5%) 4,064 (80.9%) 4,874 (75.1%) Large (> 5 members) 1,073 (21.5%) 961 (19.1%) 1,616 (24.9%) Children No 1,871 (37.5%) 1,823 (36.3%) 2,299 (35.4%) Yes 3,114 (62.5%) 3,202 (63.7%) 4,191 (64.6%) Renting home No 3,720 (74.6%) 3,780 (75.2%) 5,072 (78.2%) Yes 1,265 (25.4%) 1,245 (24.8%) 1,418 (21.8%) Area Rural 1,952 (39.1%) 1,799 (35.8%) 2,614 (40.3%) Urban 3,033 (60.9%) 3,226 (64.2%) 3,876 (59.7%) House Type Detached 2,525 (51.0%) 2,575 (51.5%) 3,646 (56.8%) Semidetached 238 (4.8%) 203 (4.1%) 203 (3.2%) Apartment 2,192 (44.2%) 2,225 (44.5%) 2,574 (40.1%) 74 Capturing household beliefs, preferences and attitudes in ECA C. Awareness of support programs of the government Figure 31 Fac tors Influencing Awareness o f Suppor t Programs a. Solar panels Factors influencing awareness of support programs (Solar Panels) Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.10 0.00 0.10 0.20 Marginal effects 75 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION b. Heating and cooling Factors influencing awareness of support programs (Heating&Cooling) Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.10 -0.05 0.00 0.05 0.10 Marginal effects c. Energy efficient appliances Factors influencing awareness of support programs (EE Appl.) Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.10 -0.05 0.00 0.05 0.10 Marginal effects 76 Capturing household beliefs, preferences and attitudes in ECA d. Waste management Factors influencing awareness of support programs (Waste) Probit Regression Coefficients with 95% Confidence Intervals  Age Under 35 Over 55  Gender Female  Education Secondary Tertiary (Ba,Ma,PhD)  Income and Employment Low income group  High income group Employed (Full or Part-time)  Household characteristics Large household (>5) HH with children Renting house: Yes Urban Limited access  Behavioral factors Willingness to change lifestyle Climate change impacts overstated Citizens' belief Preference for risk Preference for altruism Social influence: >5/10 Trust for info: government -0.10 -0.05 0.00 0.05 0.10 Marginal effects Note: The dependent variable is an index calculated as a standardized weighted index following a GLS weighting for awareness of four government support programs. The question was “Which of the following government support programs are you aware about?” The response choices were a. “Purchase of energy efficient appliances,” “Purchase of modern heating,” “Purchase of solar panels,” and “Acquisition of solid waste bins.” The index is regressed on sociodemographic characteristics, household characteristics, and behavioral factors with country fixed effects. Preference for risk is a dummy variable that has a value of 1 if a person is above the median in their country on the risk scale (between 0 and 10). Preference for altruism is a dummy variable that has a value of 1 if a person is willing to give above the median (out to 100 euros) to a charity of their choice. Social influence is a dummy variable that has a value of 1 if a person believes more than 5 out of their 10 closest friends are adopting environmentally friendly behaviors. Bars represent a confidence interval of 95 percent. 77 BEHAVIORAL INSIGHTS FOR THE GREEN TRANSITION 78