Getting the Public on Side

2. Developing a Public Acceptability Tool: How can the insights from perceptual and behavioural data be harnessed to improve reforms?

Abstract

Building on the lessons learnt from the specific but paradigmatic case of inequality-reducing policies (Chapter 1), this chapter explores how the insights from perceptual and behavioural data can be harnessed to inform the design and communication of reforms across policy areas. It focuses on the policy tools and frameworks that can help integrate considerations of public acceptability ex-ante into the reform process.

As a main outcome, a proposed Public Acceptability Tool is outlined. It is designed to provide policymakers with a “compass” they can use to better understand the conditions for public acceptability and ensure that reforms are well aligned with people’s key perceptions, attitudes and preferences. The Public Acceptability Tool provides a general framework for identifying and articulating the key dimensions that contribute to public acceptability across reform areas. This general framework is accompanied by complementary tools to help operationalise it in specific reform areas and national contexts.

Objective and contribution to the OECD’s work on the political economy of reform

Making Reform Happen: What has been done

The OECD has a longstanding tradition of work on the political economy of reform across a wide range of policy areas. A core part of the OECD’s activity consists in helping countries (i) identify priority areas for reform;1 and (ii) achieve their objectives through the formulation and design of effective policies. To put its policy advice into action, the OECD seeks to provide countries with a better understanding of the conditions that are necessary for the successful design and implementation of reforms and policies. This implies learning from the experience of other countries, leveraging the OECD’s capacity for international comparison and knowledge sharing. It has also led the OECD to pay closer attention to background factors that are specific to each national context (most notably historical, cultural and institutional factors) and the role they play in influencing the outcomes of and responses to policy. As a result, political economy considerations feature prominently in sectoral studies and recommendations across many of the areas covered by the OECD.

Recent examples include: agricultural policy reform and food system change (OECD, 2021, pp. 81-118[1]); aid policies and the implementation of development assistance (Hout, 2015[2]); biodiversity policy reform (OECD, 2017[3]); environmental policy reform and climate change mitigation (D’Arcangelo et al., 2022[4]; Dechezleprêtre et al., 2022[5]; OECD, 2023, pp. 15-50[6]); inequality and the implementation of policies for inclusive growth (de Mello and Dutz, 2012[7]; OECD, 2021[8]); institutional and territorial reforms in the context of multi-level governance (OECD, 2017[9]); labour market reform in the context of the new OECD Jobs Strategy (OECD, 2018, pp. 329-342[10]); the implementation of automatic adjustment mechanisms in the context of pension reform (OECD, 2021, pp. 83-117[11]); and political economy challenges for tax policy reform, including in areas where public opposition tends to be high such as inheritance taxation (OECD, 2010, pp. 47-80[12]; OECD, 2021, pp. 128-131[13]).

Alongside its sectoral work, the OECD has conducted broader transversal analysis on the political economy of reform. It has sought to identify the economic, social and political determinants of reform implementation through cross-country econometric analysis (Duval and Elmeskov, 2005[14]; Høj et al., 2006[15]; OECD, 2019[16]). Quantitative analysis of reform determinants has been complemented by finer-grained qualitative analysis of the reform process, which has helped bring a number of crucial elements into the picture including the role played by political actors, institutional settings and contextual variables.2 This has been done through case studies of successful and failed experiences of structural reform in different policy areas, most notably in the context of the 2007-2010 Making Reform Happen exercise (Tompson, 2009[17]; OECD, 2010[18]).

Through its Making Reform Happen exercise, the OECD has identified public acceptability as a key condition for the successful implementation of reforms. Making Reform Happen provided a series of general and sector-specific lessons on the successful conduct of reform, based on the quantitative and qualitative analysis of actual reform episodes in member-countries over several decades. The importance of public acceptability is reflected (i) in the general conclusion that the case for reform must be built on a broad consensus among the public, rather than simply on technical grounds; and (ii) through evidence that a “reform by stealth” approach is less likely to be successful in the long-term (Tompson, 2009, p. 49[17]; OECD, 2010, pp. 16-17[18]). In the decade following Making Reform Happen, the issue of public acceptability has been increasingly taken up in OECD work (OECD, 2017[3]; Dechezleprêtre et al., 2022[5]; OECD, 2023[6]). Its relevance has also grown as the challenges associated with the climate, demographic and digital transitions, as well as rising public debt in the aftermath of successive crises, require ambitious reforms in difficult areas including tax, spending, pensions, labour markets and education (OECD, 2024[19]). In parallel and adding urgency to the issue, reform activity in the decade following Making Reform Happen has slowed significantly following a brief post-crisis surge (see Figure 2.1).

Figure 2.1. The pace of structural reform has fallen from its post-2009 crisis peak

Responsiveness to OECD Going for Growth reform recommendations across countries

Note: Reform activity is measured using the responsiveness rate indicator (RRI) that captures the extent to which Going for Growth policy recommendations are taken up. “Advanced economies” comprise all OECD member-countries excluding Chile, Colombia, Costa Rica, Mexico and Turkey. “Emerging-market economies” include these 5 OECD countries as well as Argentina, Brazil, China, Indonesia, India, Russia and South Africa. The chart shows a simple average of the RRI across countries for both groups.

Source: OECD (2019[20]), Economic Policy Reforms 2019: Going for Growth, OECD Publishing, Paris, https://doi.org/10.1787/aec5b059-en.

Several of the recommendations from Making Reform Happen focus on public acceptability and on the conditions through which governments can build broad support for necessary reforms.3 These recommendations notably include:

• Having a clear electoral mandate for reform;

• Communicating effectively to demonstrate the need for change;

• Establishing a strong evidence base to support the case for reform;

• Understanding the real and perceived conditions of the policy regime to be reformed;

• Engaging with the groups threatened by the proposed reform through consultative processes;

• Considering appropriate compensatory measures at the design stage to address the social and distributional impacts of reform.

This report seeks to reassess and deepen these recommendations by taking account of the new opportunities for measuring subjective factors (including public perceptions, attitudes and policy preferences) created by recent methodological advances (see Box 1.1 in the previous chapter). In this respect, it adds a third experimental and behavioural element to the quantitative and qualitative approach developed in Making Reform Happen. This added element can shed valuable light on the “demand for reform” and help better understand the conditions for public acceptability. The aim in doing so is to take better account of people’s underlying views, attitudes and understanding of issues when designing reforms – in other words, to better align the “supply-side” and “demand-side” of reform.

Deeper analysis of “demand-side” factors is also crucial for understanding how the political economy of reform has evolved and what this implies for public acceptability. The political economy of reform has changed in significant ways in the decade and a half since Making Reform Happen. In particular, a succession of socio-economic shocks, including the Global Financial Crisis and the COVID-19 pandemic, have had an impact on people’s expectations and demands (OECD, 2010, pp. 30-32[18]; OECD, 2021[8]). Over the same period, structural transformations, including the advent of social media and more fragmented political landscapes, have contributed to reshape modes of political action, communication and relations between citizens and policymakers (OECD, 2020[21]; OECD, 2021[22]). In this context, it is important to understand how these changes have affected the conditions for public acceptability.4 This research agenda is not limited to the OECD. Governments increasingly recognise the role that perceptual and behavioural data can play in informing policy and have strengthened their capacity to collect these data (see Box 2.1). Furthermore, other international organisations, including the IMF and the World Bank, have put similar emphasis on public acceptability and its determinants and are making increasing use of perceptual and behavioural data in order to better understand them (see Box 2.2).

Box 2.1. Making public views part of the process of change: The example of the Australian Public Service Reform Agenda

The Australian government is using perceptual data to promote a people-centred public service

The Australian Public Sector (APS) reform agenda seeks to strengthen the integrity of the public service, as measured by the OECD Public Integrity Indicators, and contribute to deliver more effective, fair and inclusive outcomes. To support this agenda, the Australian Public Service Commission (APSC) – a specialised policy agency within the portfolio of the Department of Prime Minister and Cabinet – conducts two regular national surveys to evaluate its effectiveness and ensure the APS “places people and businesses at the centre of policy and services”:

• The Trust and Satisfaction in Australian Democracy survey

• The Survey of Trust and Satisfaction in Australian Public Services

The surveys cover a range of issues that are highly relevant for understanding public views of and responses to the APS reform agenda. This includes levels of trust and satisfaction in democracy; public satisfaction (and dissatisfaction) with public services, including social welfare, public health and taxation services; social capital; trust in information; and how well-being and life events affect trust in institutions. The results of both these surveys are shared with the public through reports.

Perceptual data can help manage change, ensure greater transparency and build public support

The primary use of these data consists in “managing change” within the APS by integrating public views into the existing reform process, rather than in “making the case for reform” directly to the public. However, the collection, use and reporting of these data can help build support for the reform agenda by ensuring greater transparency, demonstrating that community experience and feedback are taken into account and identifying challenges and gaps in trust that policymakers and civil servants can respond to. For example, results from the surveys show higher than average gender gaps in terms of trust in government and in public institutions compared to OECD countries, with a lower proportion of women expressing high or moderately high levels of trust than men (OECD, 2024[23]).

Other tools are designed to support the implementation of change. For example, the APS Agency Engagement Maturity Assessment Tool helps agencies and departments identify engagement capability gaps and foster a culture of continuous improvement within the public service. The collection and use of data on lived experience is also promoted to inform policies and programmes.

By shedding new light on the “demand-side” of reform, the work of the OECD Expert Group on New Measures of the Public Acceptability of Reforms helps complement and update the analysis conducted in Making Reform Happen. While different in nature, the contributions made by the Expert Group to the OECD’s work on inequality and on the political economy of reform are connected. The activities of the Expert Group focus on understanding the subjective determinants of public acceptability from a multidisciplinary and policy-oriented perspective. In doing so, the Expert Group has proceeded in two stages and sought to draw specific implications for two distinct strands of OECD work: (i) on inequalities; and (ii) on the political economy of reform. As detailed in Chapter 1, the Expert Group’s contribution to the OECD’s work on inequalities has consisted in establishing a conceptual framework that can underpin the research programme and data collection efforts outlined in (OECD, 2021[8]) and carried forward by the OECD Observatory on Social Mobility and Equal Opportunity (see Annex 1.A). The aim of this conceptual framework is to help identify and analyse the role that subjective factors play in shaping the public acceptability of inequality-reducing policies. As such, lessons relating to the analysis and measurement of key subjective factors in the area of inequality also provide the basis for a broader reflection on public acceptability that can complement and help advance the OECD’s work on the political economy of reform.

The main outcome of this work has consisted in developing new policy and measurement tools for assessing the public acceptability of reforms along the key dimensions that matter to people. The second stage of the Expert Group’s activities was practical in nature. A “Public Acceptability Tool” was designed to support policymakers’ efforts to integrate considerations of public acceptability ex-ante into the design and communication of reforms. In order to do so, the proposed Public Acceptability Tool seeks to better harness relevant insights on public acceptability from the available perceptual and behavioural data, building on the lessons drawn from the study of inequality. The rationale for developing a Public Acceptability Tool and its expected role are presented in Section 2.1.2.

Box 2.2. A common challenge and shared research agenda – Analysis of the public acceptability of reforms and its drivers by the IMF and the World Bank

The issue of public acceptability features increasingly in the analysis and advice on reform provided by the IMF and World Bank.

As with the OECD, this growing focus has led both international organisations to explore (i) the role played by perceptions, attitudes and policy preferences; and (ii) the lessons that can be drawn from perceptual and behavioural data about the determinants of public acceptability, in a country-specific and a comparative cross-country perspective.

Analysis of patterns and trends in reform activity across countries and policy areas confirms that public acceptability constitutes a significant barrier to reform. For example, (Alesina et al., 2023[24]) draws on two extensive IMF datasets to study the macroeconomic and electoral effects of a wide range of reforms. These datasets cover respectively (i) reforms in domestic finance, external finance, trade, product and labour markets for 90 advanced and developing economies over the period 1973-2014 (Structural Reform Database); and (ii) information on the results and timing of elections for a subset of 66 democracies among the countries included in the reform dataset over the same period (Electoral Dataset).

Results from this study suggest that the electoral effects of reform are real and depend on the timing of reform in relation to both the electoral cycle and the business cycle. In this respect, governments that implement reforms towards the end of an electoral term (i.e. with insufficient time to deliver visible results) or in periods of economic contraction tend to be penalised at the ballot box. Voters do not seem to fully integrate the delay that is needed for the economic effects of reform to materialise and are more likely to associate reforms with current economic conditions at the time of their implementation. The implications of this finding are particularly challenging given that (i) growth-enhancing liberalising reforms often take time to deliver economic benefits; (ii) a majority of reforms tend to be implemented during recessions, confirming the “reform under crisis” hypothesis; and (iii) tighter monetary policy and the need for fiscal consolidation may further reduce governments’ margins for building public support and therefore their capacity to carry out necessary reforms (Duval, Furceri and Miethe, 2018[25]).

Both the IMF and the World Bank are using randomised survey experiments to assess public acceptability and analyse its drivers across a range of policy areas.

Climate policies

Dabla-Norris et al. (2023[26]) conducted an online survey covering 30 000 respondents across 28 advanced and emerging economies to analyse the individual characteristics and beliefs associated with climate risk perceptions and preferences for three major types of climate policies: carbon-pricing; regulation; and subsidies for low-carbon technology and renewable energies. Results show that beliefs regarding the effectiveness of policies in reducing emissions, their distributional impact and their associated “co-benefits” in terms of improved air quality, health outcomes and job creation play an important role in translating concerns about climate change into actual support for policies to address it. Providing salient information about the effectiveness and “co-benefits" of climate policies also contributes to strengthen public support. In the case of carbon-pricing, the study confirms that using the revenue raised to address distributional concerns or subsidise green infrastructure and low-carbon technologies increases public acceptability.

Hoy et al. (2023[27]) provides evidence on the drivers of public support for energy subsidy reform – another important lever in the transition to a net-zero economy – using a large-scale experimental survey covering 37 000 respondents across 12 middle-income countries where fossil fuel subsidies are both large and regressive. The study highlights (i) limited public knowledge about energy subsidies, with just over 50% of respondents aware that these subsidies exist in their country; (ii) the scope for strengthening public support for reform through the provision of information (particularly on the negative environmental impacts of fossil fuel subsidies); and (iii) the extent to which support for energy subsidy reform is conditional on accompanying measures that redirect revenue towards investment in priority areas, such as education, health and infrastructure. Interestingly, the study also suggests that public support is not driven primarily by individual self-interest or endowments: respondents who stand to lose the most from energy subsidy reform (i.e. high-income individuals, car owners…) are just as likely to increase their support for reform when presented with information than those with little to lose.

Labour regulation

Duval et al. (2024[28]) uses an online survey covering 7 000 respondents across 14 countries to explore the relative importance of individual characteristics (gender, age…) and endowments (income, education, labour market status…) compared to beliefs (political views, trust in government, level of information…) in explaining people’s preferences for stricter/ looser employment protection legislation (EPL). This question has practical implications for the design of EPL reforms as the effectiveness of different levers will depend on what drives public acceptability in this area (i.e. compensatory measures may help build public support if endowments are the key driver, while behaviourally-informed communication campaigns may be more effective if beliefs are the key driver).

The analysis shows that an individual’s beliefs contribute 2 to 3 times more to their views on EPL reform than their own endowments and personal payoffs. As a result, individuals who would stand to gain from looser EPL (i.e. labour market outsiders) are just as likely to support strict EPL as others and are also more likely to state that they do so on the basis of societal concerns (e.g. that deregulation will not create jobs and will increase inequality) rather than individual concerns (e.g. that deregulation might hurt them personally). Furthermore, among respondents who support strict EPL, only around one-third of them would change their opinion if adequate monetary compensation were provided to address their specific concerns (either individual or societal) about deregulation. While beliefs seem to play a key role, the survey experiment also shows (i) that people have a limited understanding of EPL; and (ii) that beliefs about EPL reform can change when information is provided about how labour regulation works in practice. Insights from this study were notably used in the 2024 Article IV consultation with Spain to support IMF advice on EPL reform.

Tax policy

Hoy (2022[29]) analyses the impact that information about the progressivity of the tax and transfer system has on people’s willingness to pay taxes, based on a randomised survey experiment covering 30 000 respondents across 8 developing countries. Results confirm that tax compliance is influenced by the perceived progressivity of the tax system and not just by individual incentives. They also show that providing information about the impact of taxes and transfers on income inequality can increase people’s willingness to pay taxes, with this effect largely driven by respondents who have a stated preference for redistribution and see their prior expectations regarding the progressivity of the tax and transfer system in their country confirmed or exceeded in practice. These results have several practical implications for tax policy. First of all, reforms which aim to strengthen a country’s fiscal position through greater equity in the tax and transfer system may potentially also deliver an added benefit by increasing people’s willingness to pay taxes (and vice-versa for reforms that reduce progressivity). Secondly, there may be scope to increase tax compliance through information campaigns focusing on the progressivity of the tax system and how tax revenue is used to fund government transfers that support a range of households and notably the most vulnerable.

2.1.2 Developing a Public Acceptability Tool: What for?

The Public Acceptability Tool responds to policymakers’ need for better guidance on the conditions that contribute to building broad support for necessary reforms. The aim in doing so is to improve the design and communication of reforms by providing (i) an ex-ante assessment of how well the case for reform aligns or not with citizens’ understanding of the issue; and (ii) an opportunity to incorporate their views and demands into the reform process. This guidance is based on the growing capacity to identify, measure and analyse the perceptions, attitudes and preferences that shape public views on policy and constitute the “demand-side” of reform. It will enhance the advice provided to policymakers by drawing on existing efforts to collect relevant perceptual and behavioural data and by putting this body of evidence into action to develop more effective strategies for reform – i.e. strategies that take better account of the conditions for public acceptability throughout the reform process and better reflect citizens’ expectations and demands.

The role of the Public Acceptability Tool is to act as a “compass” that can help policymakers design reforms with public acceptability in mind and strengthen the case for reform by ensuring it is well aligned with public views. As a compass, the tool will contribute to organise the relevant perceptual and behavioural data, harness the key insights on public acceptability from these data and integrate them more systematically into the reform process. As such, the tool does not seek to provide a “recipe” or template for a successful reform. Its role consists in supporting policymakers' decisions throughout the reform process by helping them keep track of the main determinants of public acceptability and apply the insights gathered from perceptual and behavioural data in a more systematic and effective manner.

The development of a Public Acceptability Tool presents several expected benefits for policymakers. These benefits include:

• Improving policy recommendations by taking better account of the public perceptions, attitudes and preferences that determine citizens’ views on the relevant issues and shape their responses to proposed reforms;

• Selecting between different options for reform (first-best, second-best…) by assessing their “feasibility” in terms of how well aligned they are with the “demand side” of reform and what this implies for the prospects for building sufficient public support; and

• Designing reforms that are acceptable to the public and communicating the case for reform more effectively, notably by identifying mismatches between proposed reforms and citizens’ underlying views and beliefs on the relevant policy issues.

In doing so, the Public Acceptability Tool aims to complement and enhance existing OECD policy advice, including its sectoral and general recommendations on the design and implementation of reforms as well as its recommendations on public communication around reform. It also seeks to align with OECD statistical standards, including possible future guidelines on the measurement of subjective factors.

As a policy compass, the Public Acceptability Tool has two main functions. These functions have implications for how and at what stage the tool should be used.

A distinction can be made on this basis between:

• A listening function, through which the tool can be used to harness insights from perceptual and behavioural data in order to better understand “what kind of reform is likely to be acceptable to the public”. This “strategic” use of the tool can support policymakers at the design stage of reform.

• A translation function, through which the tool can help harness the data to shed light on “how a proposed reform can be presented more effectively”. This “tactical” use of the tool is mainly relevant for the communication and implementation stages of reform.

The Expert Group underlined the fact that the Public Acceptability Tool’s main value will derive from its capacity to inform the design and content of reforms in a way that better reflects public demands and can contribute to build broader support. As such, primary focus should be put on the tool’s listening function.

The Public Acceptability Tool is designed to provide a general framework that can be applied across different reform areas. Further details on the implications of this approach, as well as possible alternative approaches for assessing the public acceptability of reforms, can be found in Box 2.3. The structure and content of the proposed Public Acceptability Tool are discussed in Section 2.2, while Section 2.3 outlines other complementary tools that can help operationalise the Public Acceptability Tool for specific reform areas, national contexts and histories.

Box 2.3. Conceptual approaches for assessing the public acceptability of reforms

A two-stage approach was followed in developing the Public Acceptability Tool

This involved:

1. Establishing a general framework for assessing the public acceptability of reforms through identification of the key dimensions that matter across reform areas and contexts; and

2. Adapting this general framework to specific areas and contexts to ensure that it is relevant and operational for assessing concrete reform proposals.

This approach has several advantages. First, developing a general framework extends the tool’s scope for application as it can be used to address issues of public acceptability and support reform strategies across a wide range of policy areas. This also opens up the possibility for comparative analysis on public acceptability and how its drivers may differ across reform areas and countries, as well as for drawing further general lessons on public acceptability that can complement the analysis conducted in Making Reform Happen. Secondly, the comprehensive nature of the framework may increase its strategic value for policymakers and the relevance of the insights it can provide. In this respect, a general framework is more likely to capture the different dimensions that matter for public acceptability and can help account for the ways in which they interact in specific policy areas, national contexts and reform episodes. In doing so, it also covers the different elements that policymakers may rely on to build their case for reform and can help them determine how this case should be adjusted along different dimensions in response to public views.

Potential limits to this approach also need to be considered. The effectiveness of the tool depends first of all on the degree to which it is able to map and cover the key dimensions of public acceptability. Secondly, these dimensions are set at a general level and the tool must be tailored for specific policy areas and country contexts in order to be applied. Tailoring the general framework for specific applied cases notably implies that proper account be taken of differences between policy areas in terms of (i) the relative importance of the key dimensions of public acceptability; and (ii) the determinants of successful reform implementation. It also implies that focus be put on understanding differences in public perceptions, attitudes and policy preferences across reform areas, national contexts and groups, to the extent allowed by the available perceptual and behavioural data.

Other approaches are possible to help assess the public acceptability of reforms.

Alternative approaches identified in the literature include:

• A “sectoral” approach which seeks to identify a single key dimension of public acceptability for specific types of reform. An interesting illustration of this type of approach can be found in Bierbrauer, Boyer and Peichl (2021[30]) for a significant and widespread type of tax reform designed to make personal income tax systems more linear (i.e. the changes introduced are a “monotonic” function of income, making the income tax system more progressive or more regressive). Here, economic impact and changes in the individual tax burden are taken as the key determinant of public acceptability. Under these conditions, economic voting is likely to predominate and public acceptability can be assessed using a “median voter” model: reforms will be politically feasible if a majority of individuals benefit financially from the measure introduced. Sectoral approaches of this kind have defined areas of application and may give rise to more precise policy recommendations. Conversely, they are narrower in scope and may leave out of the picture dimensions of public acceptability that become relevant in given reform episodes or particular levers that policymakers may use to build their strategy for reform.

• An “inductive” approach that does not define public acceptability a priori but starts instead from the analysis of public attitudes and preferences to determine a reform strategy. This approach tends to rely on the use of survey experiments and focus groups to explore targeted questions regarding the grounds on which people support or oppose reform and provide tailored guidance on that basis. Duval et al. (2024[28]) offers an example relating to employment protection legislation (see Box 2.2). Inductive approaches of this kind can provide valuable insights and feedback on specific reform proposals by guiding and improving on more traditional survey work.

Conversely, because these approaches are more limited in scope and do not seek to define public acceptability either at a general or sectoral level, the insights that can be drawn from them are also likely to be more limited and less strategic insofar as they do not connect the different dimensions of public acceptability and may leave some of them out of the analysis. In this respect, inductive approaches may be best thought of as complements to the two previous approaches rather than as substitutes.

The Public Acceptability Tool (PAT)

The Structure of the PAT

The proposed Public Acceptability Tool (henceforth PAT) aims to identify and organise the different dimensions through which citizens assess reforms. The multidimensional nature of the PAT is a corollary of its role as a general framework for assessing public acceptability. The scope of the PAT implies that it be able to identify and cover a set of key dimensions which determine public support across reform areas.5 It also implies that the PAT be able to support a broader analysis of the ways in which these different dimensions interact. Analysis of these interactions is essential given that the relative importance of the different dimensions will vary between reform areas and may be influenced by contextual factors that are specific to countries and to particular reform episodes. These contextual factors include for example communication by government and by opponents of reform and their effectiveness in “activating” different dimensions of public acceptability in public and policy debates. Section 2.2.1 identifies the key dimensions of public acceptability that compose the PAT and provides a general structure for organising the insights drawn from these different dimensions.

First, economic and distributional impacts constitute key dimensions through which citizens assess the outcomes of reform. Economic and distributional impacts are of central importance in the literature on the political economy of reform, where a core challenge consists in ensuring (i) that reforms are effective and contribute to increase the aggregate welfare of society; and (ii) that they are perceived as fair in terms of how the costs and benefits of reform are shared across groups, taking account of their relative political weight.6 For policymakers, an essential first step towards building a strong case for reform must consist in seeking to understand these impacts and to clearly articulate equity and efficiency considerations. This conclusion is reflected in the general guidance on the political economy of reform provided in Making Reform Happen (OECD, 2010, pp. 12-19[18])7 and supported by empirical analysis of the trade-offs between equity and efficiency implied by structural reforms (Causa, de Serres and Ruiz, 2015[31]; Ostry, Berg and Kothari, 2018[32]). It is also confirmed by recent experimental evidence on the important role played by perceived efficiency and equity as determinants of public support for policies, notably in areas relating to climate change (Maestre-Andrés, Drews and Van Den Bergh, 2019[33]; Dechezleprêtre et al., 2022[5]; Hoy et al., 2023[27]) and taxation (Stantcheva, 2021[34]). As a result, the PAT should include Economic and Fairness dimensions and seek to connect these dimensions as key elements through which citizens assess the outcomes of reforms.

Secondly, the PAT also needs to take account of behavioural barriers that may undermine the case for reform and limit public support. The Expert Group drew attention to the fact that the behavioural aspects and implications of reform play an important role in determining public acceptability across a wide range of reform areas. This issue is well studied in the literature on the public acceptability of health interventions, where promoting behavioural change is the main objective and identifying and addressing the barriers that may prevent it constitutes a key challenge (Diepeveen et al., 2013[35]; Reynolds et al., 2019[36]). Its importance has also been highlighted in the area of climate change and environmental policy (OECD, 2025[37]). Furthermore, the Expert Group provided examples suggesting that failure to take account of these behavioural dimensions may prevent the passage of reform even in situations where the case for reform is strong on both equity and efficiency grounds and levels of public support are high. When designing a general framework, as is the case with the PAT, a broad approach is needed as the range of behaviours and potential barriers to be considered will be more diverse and often specific to particular reform areas and national contexts.

On this basis, the PAT includes a focus on the Behavioural Dimensions of Social Change. Reforms imply a broader process of social change in that they require individuals to adjust their behaviour to achieve a desired aggregate outcome. Taking account of the informational, cognitive and behavioural aspects of this process can provide a more realistic picture of the “demands” that reforms make on individuals, as well as of their capacity and willingness to make the behavioural changes demanded of them.8 Doing so can help bring to light important additional considerations that may influence people’s perceptions of and responses to reforms, beyond their immediate economic and distributional impacts. By paying attention to these dimensions, the PAT can help policymakers better understand how individuals should be accompanied through the broader process of social change implied by reform and how to foster greater public acceptability by addressing the behavioural barriers that may emerge in this context. For example, informational and cognitive barriers tend to be particularly significant in areas where the underlying issues are complex and levels of “literacy” are low, such as tax reform and pension reform. Figure 2.2 provides an example illustrating how limited knowledge and perception gaps may prevent the case for pension reform from being made effectively on economic and fairness grounds. Section 2.2.2 outlines some of the main behavioural dimensions of social change that may apply across reform areas and should therefore be covered by the PAT.

Figure 2.2. Across OECD countries, perceived measures of pension adequacy are poorly correlated with objective measures such as impact on poverty and net replacement rate

Note: Respondents were asked about their degree of agreement with the statement “I think that the government would (or does) provide my family and me with adequate income support in the case of income loss due to old-age”. Response options were “agree”, “strongly agree”, “disagree”, “strongly disagree” and “undecided”. Data on the relative income poverty rate for 66- to 75-year-olds are based on equivalised household disposable income, i.e. income after taxes and transfers adjusted for household size. The poverty threshold is set at 50% of median disposable income in each country. Data refer to 2018 or 2017, except for Mexico and the Netherlands (2016). Data on the future net pension replacement rate are based on entitlements for a man on average earnings and refer to voluntary and mandatory pensions. Data refer to 2018.

Source: OECD (2021[38]), Main Findings from the 2020 Risks that Matter Survey, OECD Publishing, Paris, https://doi.org/10.1787/b9e85cf5-en.

Finally, the PAT includes a Process Dimension to capture the sense of legitimacy that a given reform can draw, or not, from the institutional and political processes used to design, implement and justify it. Institutional and political factors linked to the reform process also play an essential role in shaping policy options, public support and ultimately the outcomes of reform across different areas. This conclusion is well-established in the literature on the political economy of reform (Williamson, 1994[39]; Drazen, 2000[40]; Persson and Tabellini, 2000[41]) and is supported by empirical evidence.9 In addition to the expected outcomes of reform and the necessary conditions for enabling the implied behavioural change, the PAT therefore needs to cover the distinct role that procedural elements play as determinants of public acceptability. An additional question to consider is how the legitimacy drawn from these procedural elements (covered under the Process dimension) relates to the legitimacy drawn from the outcomes of reform (covered under the Economic and Fairness dimensions). In this respect, the Expert Group noted that procedural elements are likely to be particularly important in contexts where policymakers are not able to articulate the equity and efficiency dimensions of reform in a clear and convincing manner or build sufficient public support on these grounds.

Based on these guiding principles, Figure 2.3 offers a visual representation of the possible structure around which the PAT can be designed as a general framework for assessing the public acceptability of reforms. In doing so, it identifies the key dimensions that contribute to public acceptability across reform areas and organises these dimensions in a way that is relevant to policymakers and can support their decisions throughout the reform process.

Figure 2.3. The structure and key dimensions of the PAT

The Content of the PAT

As discussed in the previous section, the PAT covers 4 dimensions that have been identified as relevant for understanding public acceptability across reform areas based both on a study of the political economy literature and on the inputs of the Expert Group. These 4 dimensions are:

• The economic impacts of reform and how they are assessed;

• Fairness considerations relating to the distributional impacts of reform;

• The informational, cognitive and behavioural dimensions of social change; and

• Procedural features relating to the way in which a reform is designed, implemented and adapted over time.

This section describes the main sub-components that could be included under each dimension, as well as possible perceptual and behavioural data that may be used to assess them (see the boxes at the end of each sub-section). By organising the relevant data and harnessing their insights, the PAT presents an ex-ante overview of how well a proposed reform aligns with public perceptions, attitudes and preferences along the key dimensions that shape citizens’ views of and responses to reform. This overview can help policymakers identify areas of relative “strength” on which to build the case for reform (i.e. dimensions where the reform is well aligned with public views), as well as areas of relative “weakness (i.e. dimensions where the proposed reform may be out of line with public views and more easily contested). It would also help identify key perceptions and attitudes that need either to be shifted to secure public support or taken as “hard” constraints that should guide the design of reform. The insights provided thereby could be used to (i) inform the design of proposed reforms (listening function of the PAT); and (ii) develop more tailored communication strategies and engagement efforts (translation function of the PAT).

A visual representation of the content of the PAT can be found below (see Figure 2.4). It includes the different dimensions of the PAT and their sub-components. Focus has been put on harnessing policy-relevant insights on the key political economy challenges associated with each dimension, as well as possible means and solutions for addressing these challenges. The Economic and Fairness dimensions share a common sub-component labelled “Purpose of Reform” which affects the assessment of both the economic and the distributional impacts of reform.

Figure 2.4. A framework for assessing the public acceptability of reforms – The main dimensions of the PAT and their sub-components

Economic and Fairness Dimensions

The Economic Dimension covers the expected economic impact of a proposed reform and how it is perceived by the public.10 The Fairness Dimension covers the expected distributional impact of a proposed reform and citizens’ views on its overall fairness, including broader evaluations on deservingness and burden-sharing across groups.

Sub-component #1: Prospects for Reform

Assessing the prospects for reform implies looking at the perceived need for reform (based on citizens’ evaluation of the status quo)11 and at the public’s understanding of the impacts of reform in terms of both equity and efficiency (i.e. citizens’ evaluation of the outcome of reform). Objective and subjective data are used under this sub-component to provide insights on how well the need for reform is understood, to what extent the reform is likely to be supported or opposed and on what grounds.

These insights can help policymakers:

• Better understand how people think of the efficiency and equity aspects of reform in a particular policy area. In this respect, building public support is likely to be more difficult in contexts where (i) the status quo is well established; and (ii) the impacts of reform are not well understood. This assessment would provide policymakers with a preliminary overview of how the public sees the case for reform. Insights can be drawn from there on the extent to which a proposed reform is aligned with public perceptions (i.e. the general prospects for reform in a particular area), on the relative areas of strength and weakness in the case for reform and on the grounds on which this case should be built to ensure it is acceptable, notably when articulating the economic and fairness dimensions of this case. Figure 2.5 provides an illustration using a simple binary indicator that signals whether the economic need for and impacts of reform are clearly understood or not.

• Identify possible perception gaps that can be acted on to strengthen the case for reform. Comparison between (i) the perceived and actual performance of existing policy settings; and (ii) the perceived and projected impacts of reform, along both economic and fairness dimensions, can help highlight areas where public perceptions are misaligned with actual or projected outcomes. Information campaigns targeting these perception gaps may be considered as a possible way to strengthen the case for reform by changing the public’s evaluation of the status quo and/or of the impacts of reform. Behavioural and procedural elements can provide further insight on the conditions under which information campaigns are likely to be effective or not and should also be taken into account (see below). In addition to gaps in public perceptions, possible insights could also be drawn from the comparison between citizens’ views on the need for reform and those of experts and policymakers. Understanding to what extent these views align or not and how they differ can help select policy priorities and determine what type of information to provide.12

• Help operationalise the notion of “ripeness for reform”. The notion of “ripeness for reform” captures the perceived state of the policy setting to be reformed and to what extent it is viewed as well-functioning or not (Ross, 2007[42]). Its importance as a determinant of successful reform implementation was notably highlighted by Making Reform Happen, where it was measured using a simple indicator reflecting the presence or absence of previous reform attempts in a given policy area (Tompson, 2009[17]). Subjective measures of perceptions and attitudes towards the status quo can yield additional insight on the extent to which an existing policy setting is “ripe” for reform. Similarly, comparison with the actual state of this policy setting may provide a lever to act on by helping identify a relevant perception gap.

Figure 2.5. Assessing the prospects for reform

Note: "+/-" indicates whether the need for reform and impacts of reform are well understood or not along both dimensions.

Sub-component #2: The Purpose of Reform

The perceived purpose of a reform is another significant element to consider, as it can affect the salience and relative importance given to the different dimensions (Nelson, Oxley and Clawson, 1997[43]). Some general rules may be identifiable here, with certain types of reform being intrinsically more likely to be perceived as “economic reforms” designed to promote fiscal sustainability or enhance the flexibility of product and labour markets for example. Furthermore, as with the prospects for reform, the perceived purpose of a reform could be compared with the actual framing intended by policymakers to assess how well aligned they are, including in terms of the accuracy of policymakers’ understanding of the public’s views on the proposed reform.13 In Figure 2.5, it is assumed that a reform will be perceived and framed as being driven either by economic goals or by fairness goals. It is also conceivable that a reform may be interpreted as being driven by political goals, in which case its purpose would fit under the Process Dimension.14

Sub-component #3: Key challenges for Reform (General and Sector-specific)

Assessing the prospects for reform can help better identify general challenges linked to the public’s understanding and evaluation of:

i. The status quo (i.e. to what extent the existing policy setting is “ripe” for reform); and

ii. The economic and distributional impacts of the proposed reform (i.e. whether reform will improve the status quo or not, for the individual and for society).

In particular, it may bring to light areas where the case for reform is not well aligned with public perceptions and attitudes (either because the need for reform is not clearly perceived on economic/fairness grounds, or because the impacts of reform are not well understood) and may therefore be more difficult to make or easier to oppose. Strengthening the case for reform by addressing potential areas of weakness identified by the data constitutes a first key challenge. This type of challenge applies across reform areas and should therefore feature as a separate element in the PAT.

Many of the key challenges for building public support will be specific to particular policy areas, types of reform and instrument. Desk research, consultation with experts and surveys may be used to identify key sector-specific or reform-specific challenges. Appropriate indicators can then be selected or designed to assess the extent of the challenge and monitor how effective the case for reform is in addressing it. As noted previously, the OECD Framework for Assessing Decarbonisation Strategies identifies political economy challenges associated with the use of specific instruments (D’Arcangelo et al., 2022[4]). The Expert Group highlighted examples of sector-specific challenges in the context of two case studies (reforms of tax and benefits systems and reforms of pension systems) and how these challenges relate to the different dimensions of the PAT. For illustrative purposes, these examples included:

• Economic Dimension:

  • Making the link between revenue and spending clear for the public; Increasing willingness to pay (tax and benefits reform)

  • Long-term economic impact and sustainability of the reform design (tax and benefits reform)

  • Communicating the costs of the status quo (pension reform)

  • Limited understanding of tax and pension systems and of the expected aggregate/individual impacts of reform

• Fairness Dimension:

  • Deservingness of benefit recipients; Target-specific perceptions/beliefs (tax and benefits reform)

  • Communicating the costs of the status quo on fairness grounds (tax and benefits reform)

  • Making the link between longevity/health outcomes and retirement age (pension reform)

  • Long-term distributional impact and sustainability of the reform design; Indexation of pensions (pension reform)

Sub-component #4: Means and Solutions for addressing Key Challenges

Finally, a Means and Solutions sub-component is designed to indicate possible ways in which the main political economy challenges identified for a specific reform proposal can be addressed. Here again, the Expert Group drew attention to a number of possible solutions to the key challenges highlighted during the discussion of the case studies. These solutions included: choice architecture (to address challenges linked to limited understanding of policy issues); analysis of policy preferences, framing and mental accounting (to help strengthen the perceived link between revenue and priority areas for spending)15; leveraging the social dynamics of collective action to strengthen tax morale (to increase willingness to pay)16; information provision and the design of appropriate metrics (to address perception gaps); reform analogies…

Box 2.4. Possible measurement of the PAT’s Economic and Fairness Dimensions

The following measures can be used to assess the prospects for reform under both dimensions:

Economic Dimension

• Evaluation of the status quo in terms of the relevant economic objectives for reform (efficiency, financial sustainability, international benchmarking…)

  • Comparison between the perceived and actual economic performance of the existing policy setting.1

    • At individual level (for self) / At aggregate level (for the economy as a whole)

    • In static terms (strong/weak) / In dynamic terms (improving/getting worse)

Assessing the public’s evaluation of the status quo provides an indication of how clearly the need for reform is perceived on economic grounds. A possible additional element to consider here is how important the economic objectives of the proposed reform are for the public (in terms of salience or preference intensity).

• Economic impact of reform on the relevant economic variables

  • Comparison between the perceived and actual impact of reform on current economic outcomes:2

    • At individual level (will the reform improve economic conditions for me and my household?)

    • At aggregate level (will the reform improve economic conditions for society as a whole?)

Measuring these perceptions provides an indication of how and to what extent the economic impacts of reform are understood by the public.

Fairness Dimension

• Evaluation of the status quo in terms of fairness

  • Comparison between the perceived and actual distributional impact of the existing policy setting:

    • At individual level (for self) / At aggregate level (for society as a whole)

    • In static terms (fair/unfair) / In dynamic terms (improving/getting worse)

Assessing the public’s evaluation of the status quo provides an indication of how clearly the need for reform is perceived on fairness grounds. An important additional element to consider here is what type of fairness principles are used to assess the status quo?

• Distributional impact of reform

  • Comparison between the perceived and actual distributional impact of reform3:

    • At individual level (do I receive a fair share of the benefits and costs of reform?)

    • At aggregate level (are the benefits and costs of reform shared fairly across society?)

Measuring these perceptions provides an indication of how and to what extent the distributional impacts of reform are understood by the public. As noted, Distributional Impact Assessments (DIAs) constitute particularly useful tools that can be used in combination with perceptual data to help identify relevant perception gaps. In addition, attention could also be paid to differences between the imagined stakeholder matrix (i.e. the main groups that the public perceives as being impacted by the reform) and the actual stakeholder matrix (i.e. the main groups expected to be impacted by reform, as identified by DIAs or analysis), in line with the measurement recommendations made by the Expert Group (see Section 1.3.1 in the previous chapter). Doing so can highlight possible mismatches regarding the groups that will be most affected by reform, as well as the key constituencies that may mobilise for or against reform. In both cases, these groups may not necessarily align with the “winners” and “losers” of reform in simple material terms.

When measuring the perceived purpose of reform, surveys and other methods, including social listening, can help determine the general framing through which the public perceives and evaluates a reform – i.e. whether a reform is driven primarily by economic goals (fiscal sustainability, flexibility, competitiveness…), fairness goals (reducing inequalities…) or another type of goal (political considerations…). More targeted surveys, focus groups and interviews may be needed to elicit policymakers perceptions of the purpose of reform, though studies exist looking at how political leaders (de Geus, Thode and Weidenfeld, 2016[44]) and civil servants (Civil Service, 2020[45]) perceive reform.

When identifying means and solutions, a number of perceptual data highlighted in Chapter 1 offer insights that could be used to strengthen the case for reform and address some of the key challenges listed above. For instance, data on public perceptions of who is responsible for reducing inequality collected through the Opportunities module of the OECD Risks that Matter survey could inform efforts designed to increase willingness to pay.4 Similarly, developing appropriate indicators, for example on the quality of public spending, could also contribute to address key challenges, notably by strengthening the perceived link between revenue and spending.

Notes:

1The OECD Risks that Matter survey includes questions relating to social policy that can be used to measure many of the aspects of perceived economic performance identified here (see Annex 1.B). The OECD has also developed over many years a wide range of policy indicators relating to economic outcomes, which can be used to measure actual performance. By comparing these objective and subjective measures, the PAT would aim to shed light on the way in which the public evaluates the status quo and how this relates to actual performance – in line with what (OECD, 2021[8]) has done for perceived and actual inequality (see Section 1.1.1 in the previous chapter).

2The perceived economic impact of reform could be measured using dedicated surveys or other data collection tools. Projections by trusted institutions can be used to measure the expected actual economic impact of reform.

3The perceived distributional impact of the existing policy setting and of the proposed reform can be measured using surveys or other data collection tools. Some relevant perceptual data are already collected in the OECD Risks that Matter survey, for example on the perceived effectiveness and fairness of social benefits (see Annex 1.B). Distributional Impact Assessments (DIAs) represent an important tool for measuring the actual impacts of both the existing policy setting and proposed reform. Their importance is notably highlighted in the European Commission’s 2022 Communication on Better Assessing the Distributional Impact of Member-states’ Policies. Where relevant, DIAs can be complemented with other measures of inequality.

4See OECD (2023[46]) for analysis of this data.

As a summary, a more detailed visualisation of the Economic and Fairness Dimensions of the PAT is presented in Figure 2.6. This includes all of the main sub-components and relevant perception gaps that may be identified through the analysis of these dimensions.

Figure 2.6. Detailed visualisation of the Economic and Fairness Dimensions of the PAT

Behavioural Dimensions of Social Change

Understanding how citizens perceive economic and distributional impacts is essential to build public support for the outcomes of reform. However, policymakers must also consider the behavioural barriers that may affect citizens’ views of and ability to comply with the type of social change implied by reform. Failure to take account of these behavioural barriers may undermine the case for reform and limit public support. Based on the input from the Expert Group, the PAT aims to identify the main types of informational, cognitive and behavioural barriers that may be encountered as part of the broader process of social change. These main types of barriers apply across different reform areas and are associated with specific challenges. On this basis, the PAT would include the following Key challenges, designed to cover the main types of barriers that can be encountered, as well as possible Means and Solutions under the Behavioural Dimensions of Social Change:

Sub-component #1: Key challenges – The State of Knowledge on policy issues & reforms

Limited public understanding of policy issues constitutes a significant challenge across a number of reform areas. This challenge is particularly acute in the case of complex issues such as pensions, where there is evidence of information avoidance and engagement with information may be limited despite efforts to provide and facilitate access (Eberhardt et al., 2019[47]; Dutta-Powell, Court and Gilchrist, 2021[48]; Eling and Jaenicke, 2023[49]).17 More generally, citizens may find it difficult to clearly evaluate the economic and distributional impacts of reform. Again, this can be the case even when official estimates are provided, as different methodologies and models can lead to different and competing evaluations of these impacts (Bourguignon and Landais, 2022[50]). The implications for the public acceptability of reforms are significant. For instance, limited technical understanding of policy issues tends to reduce the effectiveness of economic arguments, which often underpin policymakers’ case for reform, in public debates. This also implies that citizens may view the trade-off between the equity and efficiency aspects of reform in very different terms from policymakers. Furthermore, a limited technical understanding of policy issues does not necessarily imply that people will be likely or willing to change their views when provided with new and more complete information. In addition to the public’s “literacy” (how well is an issue or the effects of a policy understood?), being able to assess the public’s openness to information would provide valuable insight regarding the relevance and possible effectiveness of information campaigns on a particular policy issue.

Possible means and solutions for improving engagement with information and addressing challenges linked to the state of knowledge include:

• Peer learning and good practices on public communication and information provision;18

• The experimental testing of messages and their impact on beliefs;19

• Exploring different forms of communication, including the use of representative “personas”, qualitative and narrative data, participatory and deliberative processes;

• Reform analogies (designed to demonstrate that similar types of reform have been successful in other areas, for example by using energy subsidy reform as an analogy for tax reforms…);

• Choice architecture (design of fair default options in the case of pension reform…).

A broader insight to be drawn on the state of knowledge is that the case for reform should start from – and to a degree be grounded in – the existing conditions of public debate and take account of relevant informational and cognitive constraints, as they shape public responses to and support for reform. Policymakers may consider possible means to lift some of these constraints and change the state of knowledge. If the relevant perceptions and attitudes cannot be shifted however, the state of knowledge may need to be viewed as a hard constraint to which the design and communication of reform must be adapted.

Sub-component #2: Key challenges – The effects of uncertainty on support for reforms

Even in situations where the public is well informed and has a good understanding of the issue and the impacts of reform, uncertainty can prevent the passage of reform. This barrier is well identified in the political economy literature, where it is often analysed as a source of status quo bias and a constraint for public acceptability (Fernandez and Rodrik, 1991[51]; Drazen and Grilli, 1993[52]; Jain and Mukand, 2003[53]; van Wijnbergen and Willems, 2016[54]; Potter, 2018[55]).

Different forms of uncertainty may affect public acceptability:

i. Uncertainty regarding who will “win” and “lose” from reform in economic terms;

ii. Uncertainty regarding the transition costs involved in moving from the established status quo to a new system or outcome;

iii. Uncertainty regarding the effectiveness of a reform in terms of achieving its stated objectives;

iv. Uncertainty regarding the government’s ability to put adequate and effective support measures in place as part of a package to ease the costs of transition.

These sources of uncertainty cover elements that are important for other dimensions of the PAT: perceived economic and distributional impacts of reform (Economic and Fairness dimensions); the government’s credibility (Process dimension). As such, uncertainty may constrain policymakers’ ability to make an effective case for reform on particular grounds, in similar fashion to a limited technical understanding of policy issues. Identifying and addressing sources of uncertainty may however open up opportunities for strengthening public support for reform through the provision of appropriate information, as highlighted by Dechezleprêtre et al. (2022[5]).

Sub-component #3: Key challenges – People’s “margins” for adapting to social change

The costs of adjusting to social change go beyond the immediate economic costs that reform imposes on individuals. This may include less immediate psychological and behavioural costs, as reform changes the environment in which individuals take decisions, modifies the incentives they face and mandates or prohibits certain types of behaviour. In this respect, a behaviourally informed assessment of the “winners” and “losers” of reform based on these broader costs may help better identify the key constituencies affected by reform, including potential sources of opposition to reform among populations that have limited margins to adapt their behaviour.20 Consideration of these costs may also influence the design of reforms and notably decisions regarding the need for compensation, support and transition measures, as well as the type of measures that should be put in place to accompany the process of social change. In this perspective, focus should be put on what is required to change actual behaviour, both at individual and collective level, beyond simply changing the incentives that individuals face. As a further point of interest, the Expert Group highlighted the fact that opposition to reform based on the unequal ability to adapt to social change is often expressed on equity grounds and refers to principles of fairness.21

Sub-component #4: Key challenges – Area-specific behavioural elements

In addition to these general themes, particular behavioural elements or traits will be relevant for reform design and communication in specific areas. The case studies discussed by the Expert Group highlighted several examples of area-specific behavioural elements to consider, including:

• Tax and benefits reform: Status quo bias and loss aversion (e.g. in the case of reforms designed to reduce or eliminate distortionary tax exemptions); Framing effects that influence the strength or salience of the link between areas of public revenue and spending (e.g. mental accounting in the case of a carbon tax with revenue recycling); and

• Pension reform: Heterogeneity in attitudes and preferences towards risk; Time-inconsistent preferences (myopia and hyperbolic discounting).

Box 2.5. Possible measurement of the PAT's Behavioural Dimensions of Social Change

The following measures can be used to assess the different sub-components for this dimension:

State of Knowledge

• Public literacy on policy issues (financial literacy, pension literacy, tax literacy…) can be measured using specifically designed tests or based on evidence from existing surveys (e.g. on tax perceptions) to determine how well people understand an issue and the effects of policy (overall and on them). For example, Dechezleprêtre et al. (2022[5]) construct a Knowledge Index that measures people’s level of understanding of the causes and impacts of climate change. Different OECD sources and surveys can be used in similar fashion to assess public perceptions and literacy in particular policy areas, including tax and pensions. Otherwise, public literacy can be assessed based on the literature or dedicated surveys and tests.

• People’s self-assessed knowledge of policy issues can be measured through surveys, if possible in conjunction with measures of policy literacy, to determine how well people feel they understand an issue and the effects of policy. Comparison between these two objective and subjective measures of literacy (actual and perceived knowledge) can provide an indication of people’s openness to information and through it of the likely effectiveness of information campaigns in shifting their perceptions and strengthening the case for reform.

• Existing evidence on public engagement with information can also provide useful insights on people’s openness to information in specific areas, notably in situations where forms of information avoidance are observed such as in the case of pensions (see above). More direct and onerous approaches can be used to measure openness to information through experimental surveys or protocols designed to measure willingness to pay for information (Alesina, Miano and Stantcheva, 2023[56])1 or to assess belief adjustment and the effectiveness of information (Alesina, Stantcheva and Teso, 2018[57]; Duch and Torres, 2023[58]). Evidence on people’s openness to information can be related in turn to possible perception gaps identified under the Economic and Fairness dimensions (see Figure 2.6 above). In this perspective, the existence of perception gaps combined with low openness to information would suggest that the impact of communications campaigns could be limited and that policymakers will need to consider appropriate methods for conveying evidence in order to shift public perceptions.

Uncertainty

This sub-component can be measured through added questions on people’s degree of uncertainty/ confidence in their views regarding key variables that affect support for reform.

As detailed above, these variables may include:

• Who will gain and lose from reform? (self; groups);

• The effectiveness of reforms;

• The government’s credibility in implementing reform and compensation measures.

Uncertainty and the impact of information can also be assessed using other methods (focus groups…).

Margins for adapting to social change

• Capacity to adapt can be measured first of all in material/financial terms, by looking at objective measures of income instability and at perceived income insecurity.2

• Assessing the non-financial margins required to adapt to social change implies that one determines first what behavioural change is asked of individuals and key groups of stakeholders.3 From there, the range of “losers” from and potential opponents to reform can be broadened by identifying populations with a limited capacity to change their behaviour or who face higher adjustment costs.

Additional elements to consider may notably include: (i) how the implied behavioural change is evaluated (positively/negatively), as this may also influence people’s willingness to support reform; and (ii) preference intensity could also possibly be measured as another indicator of willingness to change behaviour.4

Area-specific behavioural elements that are relevant for reform design or communication

The relevant elements for specific policy areas would be identified based on desk research and/or consultation with experts. These elements may be measured or assessed in applied contexts using surveys and experimental methods. OECD work on the application of behavioural insights can provide resources for understanding how reform design and communication should be adapted in light of the role played by behavioural elements.

Notes:

1. An interesting finding in this survey experiment is that, all things considered, those who have the largest misperceptions are less willing to pay to receive the correct information. This is either because they are more confident in their views or hold strong beliefs and are less open to new information.

2. See for example (OECD, 2023[59]).

3. See OECD (2025, pp. 20-21[37]) for a case study on how behavioural science has been used to identify the main barriers to the adoption of sustainability measures by small- and medium-sized enterprises in the Netherlands. The OECD BASIC and LOGIC frameworks provide guidance on how to apply and mainstream the ethical use of behavioural insights in public policy (OECD, 2019[60]; 2024[61]).

4. This point is in line with the Expert Group’s recommendation to focus on the Intensity of Policy Preferences as a key factor to measure (see Section 1.3.1 of the previous chapter for further detail).

As a summary, a more detailed visualisation of the Behavioural Dimensions of Social Change covered by the PAT is presented in Figure 2.7. This includes the main sub-components and some of the channels through which these behavioural factors may affect other dimensions of public acceptability.

Figure 2.7. How the Behavioural Dimensions of Social Change can affect the case for reform

Process Dimension

As highlighted in Section 2.2.1, a Process Dimension is needed to capture the sense of legitimacy that a given reform can draw, or not, from the political and institutional processes used to design, implement or justify it. Process legitimacy is an important determinant of public acceptability and is distinct from the support that can be derived from the outcomes of reform and from the presence or absence of behavioural barriers to social change. The current section seeks to identify the main sources of process legitimacy that apply across reform areas. Further questions on how the Process Dimension interacts with the other dimensions of the PAT and on its relative importance across different reform areas are also considered.

A number of key procedural elements contribute to the legitimacy and acceptability of reforms. These key elements can be used to specify the content of the Process Dimension. Building on the elements identified in Making Reform Happen and on input from the Expert Group, the following Key Challenges and possible Means and Solutions for addressing them should be considered.

Sub-component #1: Key challenges – Administrative and technical capacity

Administrative capacity is an essential condition for the successful implementation of reforms (Jain and Majumdar, 2016[62]; Khemani, 2019[63]). Insofar as it helps determine whether a reform delivers on its objectives and meets public expectations, administrative capacity constitutes one of the main sources of legitimacy both during and after the implementation stage of reform.22 This also covers the technical capacity to evaluate and monitor the economic and distributional impacts of reforms through the development of appropriate tools, including macro-econometric models, micro-simulation models and surveys.23 Administrative and technical capacity can be related to the uncertainty component presented under the Behavioural Dimensions of Social Change. Building and demonstrating sufficient capacity can help address issues of credibility and commitment that governments may face, notably in deploying compensatory measures. Additional and adapted capacity may also be necessary to address some of the barriers to reform identified under the margins to adapt component and facilitate behavioural change.

Sub-component #2: Key challenges – Evidence base

Establishing a strong evidence base to support the case for reform constitutes a key recommendation from Making Reform Happen (see Section 2.1.1 above). Particular emphasis should be put on two aspects: (i) the extent to which appropriate evidence is produced and integrated into the reform process, notably at the design stage; and (ii) the extent to which it is effectively communicated and used to set common terms for policy and public debates, as outlined for example in (OECD, 2023[64]). Trusted information providers can contribute on both counts, notably when they have a mandated role in the reform process and sufficient institutional weight to influence it.24 Developing and using effective metrics to inform policy and public debates – i.e. ones that are relevant to the policy issue and adapted to the public’s concerns and understanding of the issue – constitutes another important element to consider.25

Sub-component #3: Key challenges – Political conditions

Political conditions are often exogenous to the reform process but tend nonetheless to play an important role in ensuring public support and through it the success or failure of reform. The main political success factors identified in Making Reform Happen (Tompson, 2009[17]; OECD, 2010[18]) as contributing across reform domains include: (i) strong electoral mandates; (ii) government cohesion; and (iii) the timing of reform in the electoral cycle. These factors may be further complemented by additional indicators (general or sector-specific) based on analysis of the literature, by subjective indicators and by measures of trust in government, civic engagement or political efficacy as proxies for satisfaction with political processes.26

Sub-component #4: Key challenges – Engagement with stakeholders

Engagement with stakeholders represents an essential and well-identified condition for overcoming opposition to reform and ensuring broad legitimacy across reform areas. National institutional frameworks need to be taken into account, with effective consensus-building institutions and a culture of social dialogue often highlighted as factors that contribute to promote public acceptability and facilitate reform (Tompson, 2009, p. 55[17]). Even in the absence of established institutions of this kind, participatory processes may provide means for doing so. Discussions by the Expert Group identified several key characteristics for assessing engagement with stakeholders, notably (i) its inclusive character and capacity to ensure that all voices are heard27; and (ii) the capacity to reach out to and mobilise constituencies in favour of reform, as well as engaging with constituencies that are opposed to it (de Geus, Thode and Weidenfeld, 2016[44]). Engagement of this kind should be extended when additional stakeholders have been identified on the basis of their limited margins for adapting to the changes implied by reform.

Box 2.6. Possible measurement of the PAT’s Process Dimension

The following measures can be used to assess the different sub-components for this dimension:

Administrative capacity

This component would check, first of all, the extent to which capacity is considered and assessed throughout the different stages of the reform process. Benchmarking and existing tools and frameworks can be used to measure and assess capacity needs.1 Capacity-building measures introduced as part of the reform process (recruitment, technical training, peer learning, building or introduction of appropriate tools…) can also be assessed, as well as the pilot testing of these measures where applicable. The assessment of capacity can be complemented with subjective measures of satisfaction with public services (surveys, online tools, analysis of digital traces…), as done for example in the OECD Survey on Drivers of Trust in Public Institutions (OECD, 2024[23]) or in the case of the Australian Public Sector reform agenda (see Box 2.1).

Evidence base

Benchmarking, standards and good practices for evidence-based policymaking can help assess to what degree the available information is effectively integrated into the reform process. Similarly, existing recommendations and good practices on public communication can help assess how effectively evidence is communicated in policy and public debates (OECD, 2021[22]; 2023[64]). In addition to these procedural indicators, the reception of the evidence and its effectiveness in influencing policymakers and citizens on key issues or variables (e.g. effective communication of reform objectives, impacts or the costs of the status quo) can also be assessed using traditional methods (surveys, interviews, focus groups…) and social listening tools (applied to parliamentary texts for policy debates and to news media sources and digital traces for the public debate). On this topic, see for example (OECD, 2024[23]) and the case study on social listening provided in Section 2.3.5 below.

Political conditions

A series of key political variables are identified in Making Reform Happen, including having a strong electoral mandate for reform, government cohesion, a stable political horizon and sufficient time to prepare a reform (Tompson, 2009[17]). These variables and indicators can be used as a starting point and extended through further analysis of the literature in specific reform areas. Subjective measures can complement these indicators. For example, measures of perceived legitimacy could be used to assess the strength of the mandate for reform, in addition to the indicator proposed in (Tompson, 2009[17]) (presence or absence of a reform pledge in the government’s electoral programme), through surveys, focus groups or social listening tools.

Engagement with stakeholders

Making Reform Happen identifies a series of key variables associated with communication and engagement with stakeholders: use of consensus-building institutions, formal concertation with social partners, respect for acquired rights, role for opponents of reform, activation of constituencies in support of reform (Tompson, 2009[17]). These variables and indicators can be used as a starting point and extended through further analysis of the literature in specific reform areas. Ongoing OECD work on facilitating and promoting citizen participation throughout the policy process can also provide valuable insights and recommendations, as well as possible indicators, in this area (OECD, 2022[65]).

Note:

1See, for example, the OECD Administrative Capacity Building Self-Assessment Instrument in the area of regional development and structural cohesion (OECD, 2021[66]).

As a summary, a more detailed visualisation of the Process Dimension is presented in Figure 2.8. It includes the main sub-components and some of the ways in which elements of process may interact with the components identified under the PAT’s Behavioural Dimensions of Social Change.

Figure 2.8. Elements of process can help address some of the behavioural barriers to reform

A complementary toolkit to operationalise the PAT

The general framework detailed in the previous section can help improve the OECD’s analysis and advice on reforms by ensuring that considerations of public acceptability are taken into account more systematically. It contributes to do so by (i) “mapping” the key dimensions that matter for the public acceptability of reforms; (ii) providing a means to identify and organise the relevant perceptual and behavioural data; and (iii) harnessing the insights from this body of data to better understand how well a proposed reform aligns with public views and demands.

A set of complementary tools is needed to support policymakers in designing and implementing concrete reform proposals for specific areas and national settings. These tools aim to operationalise the general framework for use in applied contexts. Section 2.3 outlines the proposed elements for a toolkit that governments can use. These elements include:

• A general checklist to help ensure that all the elements relevant to public acceptability are taken into account during the reform process (Section 2.3.1);

• Sectoral roadmaps to determine which dimensions of public acceptability are likely to matter most for a specific type of reform and where the main challenges lie. These roadmaps will help “configure” the PAT for use in particular areas and contexts (Section 2.3.2);

• Data review and collection to identify “reform-breaking data” and indicators to assess how effective the case for reform has been in building public support. These indicators would contribute to deepen and contextualise standard opinion surveys measuring support for reform by providing a more detailed picture of the key factors that shape public responses to reform. The identification of reform-breaking data could help policymakers decide whether and when it may make sense to halt a particular reform effort in light of the failure to shift key public perceptions or attitudes (Section 2.3.3); and

• A social listening tool to help assess how the public is responding to the main aspects of the government’s case for reform.

The PAT is meant to be applied in conjunction with the existing tools and methods that governments use to inform the design, implementation and communication of reforms. As such, the general framework and complementary tools provided by the PAT can be integrated, as needed, into the regular processes for designing, implementing and communicating reforms. Doing so can help enhance these processes through the added insights the framework and tools offer on public acceptability. The PAT is notably conceived to work alongside and add to existing tools and methods that are widely used across OECD countries, notably: (i) macro-econometric and micro-simulation models for assessing the economic and distributional impacts of proposed reforms; and (ii) communication tools to disseminate information about the effects of reform to the public.

General checklist

A general checklist can provide policymakers with practical guidance on how to apply the PAT in a way that is effective and systematic. In doing so, it can help organise the different dimensions of public acceptability covered by the PAT and ensure that their implications are taken into account at relevant stages of the reform process in order to build a strong case for reform. A prototype checklist is presented in Box 2.7 below and detailed in Annex 2.A at the end of this chapter for further development and pilot application.28 It builds on and extends a similar “reformer’s checklist” from Making Reform Happen (OECD, 2010, p. 16[18]) by adding a specific focus on the conditions for public acceptability. It also takes the standardised checklist developed for the OECD Principles for Ethical Behavioural Science in Public Policy (OECD, 2022[67]) as a possible template.29

Figure 2.9. A general checklist can be built around the framework provided by the PAT

The checklist would reflect the structure and key features of the PAT’s general framework. On this basis, it would be organised around the 3 guiding questions previously identified as essential steps for assessing the public acceptability of reforms (see Figure 2.3). Figure 2.9 above provides an overview of the structure and content of the PAT.

The checklist should start by considering the intended purpose of the reform. Policymakers may have different motivations for reform, including reform on economic grounds; reform on equity grounds; and reform on political grounds. Understanding these motivations provides valuable preliminary insight as they are tied to particular objectives and will frame the case for reform by putting emphasis on specific dimensions of the PAT. For example, in the case of pension reform, objectives may include (i) ensuring the long-term financial sustainability of the pension system (Economic Dimension); (ii) increasing the adequacy of pensions to reduce old-age income poverty (Fairness Dimension); (iii) promoting greater individual responsibility for retirement choices or facilitating pension decisions (Behavioural Dimensions of Social Change); or (iv) strengthening the governance and stability of pension systems (Process Dimension). A clear understanding of the motivation and key objectives of a reform can help frame the case for reform, anticipate how the public is likely to respond to it and identify key challenges that will need to be addressed.

An additional issue to consider here is the extent to which public acceptability constitutes a relevant concern for policymakers. Reforms are not always designed with the general public in mind. Policymakers may for instance target particular groups as the relevant constituency in the case of reforms on political grounds. Even in this case, it can be argued that public acceptability should remain an important concern as it is an essential condition for the long-term political “viability” of a reform and for avoiding problems of reform reversal and “dead-end reforms” (i.e. reforms that are enacted but prevent any further reform activity in their area over the medium/long-term due to the negative impact they have on perceptions, attitudes and overall public acceptability).

From there, the general checklist would follow the 3 main stages set out in the PAT for assessing the public acceptability of reforms. This implies looking at the following broad questions:

1) To what extent are policymakers able to clearly articulate the equity and efficiency implications of reform? As argued in Section 2.2.1, insofar as the economic and distributional impacts of reform represent the main outcomes of interest to citizens, the Economic and Fairness dimensions can be taken to cover the “outcome legitimacy” of a reform and constitute key conditions for public acceptability.

2) How well are the behavioural barriers to reform understood? Again, as argued in Section 2.2.1, outcome legitimacy may not be enough in and of itself to bring about reform. A second essential step for policymakers consists therefore in identifying the relevant behavioural implications of reform and the ways in which they may undermine the public’s understanding of the case for reform and willingness to support or comply with reform. The PAT proposes to take proper account of these Behavioural dimensions. In order to do so, it considers the broader informational, cognitive and behavioural implications of the type of social change introduced by a reform, the barriers they may present and the measures that may be needed to accompany individuals in making this change.

3) What role do elements of process play in strengthening or undermining the case for reform? Reforms can also draw on a form of “process legitimacy” separate from their outcomes and a third and final step consists in assessing the conditions that contribute to it. These elements are considered under the Process dimension of the PAT. The relative importance of outcome legitimacy and process legitimacy will vary across reforms areas. This issue will be addressed when developing sectoral roadmaps for reform (see Section 2.3.2 below).

Finally, the checklist should also include general principles or good-practices rules that can help better integrate public acceptability into the design of reforms. For example, the discussion of pension reform by the Expert Group highlighted a series of “proportionality rules” that can help address several key challenges by exploiting the interactions between different dimensions of the PAT:

• Proportionality between the Margins to Adapt and the degree of Stakeholder Engagement: The more important the adjustment and behavioural change demanded, the greater is the need for consultation (and/or administrative capacity to facilitate the transition). This would reflect a general principle of “nothing about me decided without me”.

• Proportionality between the State of Knowledge and the Evidence Base: The less an issue is understood, the more focus should be put on providing information and on engagement with stakeholders. This would reflect the principle of “not taking people for fools”.

Further research could be conducted to identify general and sectoral good-practice rules of this kind, based on country experience.

To facilitate the application of the PAT, the checklist can be tailored to fit the key characteristics of a particular reform proposal and context, as well as the available data. Data constraints mean that it may not be possible or cost-effective to cover all the relevant dimensions of the PAT and implement the full checklist presented in Annex 2.A. Consideration would therefore be given in each specific case to possible streamlining and tailoring of the checklist to identify what are the areas of key importance for assessing the public acceptability of a reform and what indicators can be used to do so. Designing roadmaps for reform (Section 2.3.2) and strategies for data collection (Section 2.3.3) can help identify the key questions and indicators. Where perceptual or behavioural data are not available for specific dimensions of the PAT, alternative options can be explored for measuring and assessing public acceptability or ad hoc tools can be implemented (surveys, survey experiments…).

Box 2.7. Outline of a possible Reformer’s Checklist

1st step: Assessing outcome legitimacy – To what extent does the case for reform clearly articulate the equity and efficiency implications of reform for the public?

Purpose of reform

• What is the intended purpose of reform?

  • Is that purpose clearly established and understood by policymakers?

• How well is the intended purpose perceived by the public?

  • Is the case for reform effectively and coherently framed around its intended purpose?

  • Are good practices and OECD guidelines in terms of public communication being followed?

Prospects for reform

• Do policymakers clearly understand the expected economic & distributional impacts of reform?

  • Has the performance of the existing policy setting been properly assessed or benchmarked? Has the economic impact of reform been properly assessed?

  • Has the distributional impact of the existing policy setting and proposed reform been assessed?

  • Does the assessment indicate that:

    • A majority of the population is expected to benefit from reform (Economic dimension)?

    • The reform is expected to reduce inequality overall or for the key groups affected (Fairness dimension)?

• Does the case for reform take proper account of public views on the need for reform?

  • How well aligned is the case for reform with the public’s understanding of:

    • The need for reform?

    • The likely impacts of reform?

  • Have communication campaigns sought to address the key perception gaps identified? How effective have they been in shifting the relevant attitudes and perceptions?

2nd step: Taking account of the behavioural dimensions of social change – What behavioural barriers may undermine support for reform?

State of Knowledge

• Can the public’s level of knowledge and openness to information be assessed? Have the implications been taken into account when designing communication around the reform?

• How likely is the public to engage with quantitative evidence? Is this reflected in the type of information and modes of communication used to build public support for reform?

Uncertainty

• Who do people perceive to be the main “winners” and “losers” of reform?

• How well do these perceptions align with the expected “winners” and “losers”?

Margins to adapt

• What behavioural adjustments are demanded by reform? Have groups with limited margins to adapt been identified?

• Have appropriate support, compensation or transition measures been put in place? Is the public sufficiently aware of these accompanying measures?

3rd step: What elements of process are needed to support the case for reform?

Administrative and technical capacity

• Is there appropriate administrative and technical capacity in place to deliver on reform objectives?

  • Have capacity needs been assessed and integrated at all stages of the reform process?

Evidence base

• How well is the evidence used?

  • To what extent has the design of reform been informed by the evidence?

  • How well has the evidence been communicated to the public? Has focus been put on ensuring that the messaging is coherent and effective, in line with existing good practices?

Political conditions

• Have the key political conditions for reform been satisfied?

  • Is there an electoral mandate for reform and is it clearly perceived by the public?

  • Does the reform have clearly identifiable “owners” and how are they perceived by the public?

• Is the expected timeframe for design, adoption and implementation of reform realistic?

  • Does the timeframe for reform allow sufficient time for effective communication and consultation, in light of the challenges identified?

Stakeholder engagement

• How effective and inclusive has consultation with key stakeholders been?

  • Is there a strategy in place for engaging with the constituencies threatened by reform?

  • Have efforts been made to engage with and mobilise constituencies in favour of reform?

  • Have OECD guidelines and good practices on stakeholder engagement been applied?

• Has the role of stakeholder engagement been taken into account when designing the reform?

  • Can consultation increase support for reform by helping address challenges related to uncertainty about reform or identifying ways to facilitate adaptation?

Sectoral roadmaps for reform

To accompany the general checklist, sectoral roadmaps for reform can be drawn up to provide more detailed guidance and inform concrete reform proposals. The general framework is designed to help integrate considerations of public acceptability ex-ante into the reform process and organise the available data. The insights generated using the general framework can be used to develop applied roadmaps for reform that are adapted to the particular conditions encountered by policymakers. These sectoral roadmaps would notably reflect the relative importance of the PAT’s different dimensions and specific interactions between them within given policy areas and national settings. They would also focus on the key challenges that the case for reform will need to address, as well as on the appropriate means and solutions for addressing them.

The following resources can be used to help develop sectoral roadmaps that are tailored for specific reform areas and national settings:

• Insights from OECD analysis on the key challenges for reform in specific areas and countries, building on the general recommendations for reform contained in the Economic Policy Reforms: Going for Growth reports, as well as on the detailed economic surveillance of countries provided through the Economic Surveys, and on in-depth sectoral studies conducted in specialised OECD publications such as Pensions at a Glance or Tax Policy Reviews. Taking the area of tax policy as an example, Table 2.1 offers an illustration of the added concrete insights that can be drawn to guide reform efforts by applying the PAT (3^rd column on the Relative importance of the different dimensions of public acceptability), compared to a more traditional political economy analysis (2^nd column on Key political economy challenges).

• Lessons from previous reform episodes in the area considered and in cases that are comparable in terms of national context, challenges faced and/or institutional settings. These lessons can be collected through different means, notably:

  • Cross-country comparative analysis on the conditions for the public acceptability of particular types of reform. Similar work has already been conducted by the IMF on the political costs of different types of reform, combining data-sets on structural reforms and on electoral outcomes (Ciminelli et al., 2019[68]; Alesina et al., 2023[24]) (see Box 2.2 above). This work could be extended to cover issues of public acceptability, for example (i) through modelling of the possible relations between key subjective factors identified and the successful implementation of reforms; or (ii) by looking at the longer-term impact of past reforms on public perceptions and through this on the feasibility of future reforms in the same area.

  ❏ Another interesting approach could consist in looking at the “actual roadmaps” from previous reform episodes (i.e. what shape the government’s intended case for reform ended up taking in practice and what grounds dominated public debate). These actual roadmaps can help better understand what arguments and dimensions the public was receptive to.30 Identification and analysis of these actual roadmaps would be conducted through case studies of relevant reform episodes and interviews of policymakers and stakeholders involved as well as of experts.

Sectoral roadmaps also provide an opportunity to adapt and streamline the general checklist in a way that fits more closely with applied conditions. The sectoral roadmaps are designed to help policymakers specify and fine-tune their case for reform based on a more detailed assessment of key challenges and the relative importance of different dimensions for public acceptability in the policy areas and contexts considered. On this basis, a subset of relevant questions can be selected from the general checklist to produce a streamlined checklist that is adapted to the conditions encountered for a specific reform proposal. Additional questions that do not feature in the general checklist may also be included. These streamlined checklists may be easier to implement and less onerous in terms of the data required.

Table 2.1. Examples of sectoral insights on the public acceptability of tax reform

Selected representative examples of tax and benefit reforms	Key political economy challenges	Relative importance of the different dimensions of public acceptability	Possible analogies with other areas
Health taxes Ex: Increasing tobacco excise duties	Fairness concerns: The tax is regressive Behavioural aspects: The tax aims to change individual behaviour and reduce a negative externality	Efficiency tends to dominate fairness concerns in public debates The purpose of the tax matters: Its aim is viewed as legitimate	Other “sin taxes” (sugar-sweetened beverages…) Other taxes aimed at changing behaviour in situations with negative externalities
Property taxes Ex: Phasing out of mortgage tax relief	Equity/Efficiency trade-off: The impact differs across households (recent buyers lose out) Status quo bias: The tax advantage is capitalised in house prices; Some households face adjustment costs Salience of property as an issue	Status quo bias tends to dominate public debates, even when the existing system is poorly designed As a result, argument for reform is hard to make on equity grounds	Other areas that involve reforming an existing system to which people have adapted their expectations and behaviour
Financing of social protection Ex: Introduction of Social Security Contributions in systems based on Non-Tax Compulsory Payments	No equity/efficiency trade-off: Strong argument for reform on both grounds Behavioural aspects: Uncertainty over outcomes and transition costs (scale and distribution)	Behavioural aspects may dominate both equity and efficiency arguments for reform Reducing uncertainty and accompanying individuals throughout the transition process should be key priorities
Environmental taxes Ex: Increasing the price of carbon and recycling the revenue	Equity/Efficiency trade-off: Efficient instrument, but recycling is needed to compensate for the uneven costs of adjustment Behavioural aspects: The tax aims to change behaviour and margins to adapt may be limited for certain groups Salience of carbon tax	Economic and behavioural issues (cost of living; adjustment costs) tend to translate into fairness concerns Compensation is hard to do well and credibly, which may limit its effects (issues of capacity, uncertainty and trust) Risk that (i) recycled revenue gets discounted over time; and (ii) the carbon tax is associated with the overall increase in fuel prices (ex. of the Gilets Jaunes)	Limits of the analogy between health taxes and environmental taxes: In the case of the proposed increase of the carbon tax in 2018 in France, the expectation may have been that a logic similar to that seen with health taxes would apply, with the common benefits of behavioural change overcoming equity considerations (episode of the Gilets Jaunes)

Source: OECD Secretariat – Presentation to the OECD Expert Group on New Measures of the Public Acceptability of Reforms.

Data review and collection

Sectoral roadmaps can help identify the key subjective factors that are needed to ensure the case is well understood and can gather sufficient public support. In turn, measurement of these subjective factors can provide valuable insight on:

• How well aligned the government’s case for reform is with public perceptions, attitudes and preferences and how difficult it is likely to be to make this case;

• Which public perceptions, attitudes and policy preferences will need to be leveraged or shifted in priority to ensure sufficient support; and

• How the case for reform may need to be adapted in light of the constraints created by public perceptions, attitudes and preferences.

To facilitate the application of the PAT, a review would be conducted of the available relevant data collected through existing OECD tools (Risks that Matter survey; Drivers of Trust in Public Institutions survey…), as well as through the main national and international sources. Where key data gaps are identified, consideration would be given to the types of additional data that are needed and the most effective means through which they can be collected. Avenues for doing so may include the consolidation and merging of existing datasets, notably administrative data, or the setting up of an OECD Perceptions Database to compile key data (see Section 1.4 in the previous chapter).

Measurement of these factors can also help identify “reform-breaking data” – i.e. perceptions, attitudes or preferences that are of crucial importance for the case for reform and without which it is unlikely to be understood or accepted by the public. Once identified, these perceptions, attitudes and preferences can be used as a goal or indicator against which governments can assess whether and to what extent they have been successful in making the case for reform.31 Failure to shift or maintain enough of these key factors could also provide a useful indication of when it may make sense to bring a particular reform effort to a halt. This would be done on the grounds that the case for reform was not successfully made or understood, the conditions for public acceptability are not met and that a reform, even enacted in this context, would lack sufficient legitimacy and thereby run the risk of being reversed or blocking future opportunities for reform in this area (i.e. so-called “dead-end reforms”).

Social listening tool

Efforts to monitor the political and public debate around specific reforms could further increase the relevance of the proposed PAT. Similar strategies and capabilities for “social listening” have been developed during the COVID-19 pandemic (Alfonsi et al., 2022[69]). These strategies have generally aimed to improve public health communication through a better understanding of people’s perceptions of issues and risks, of the way in which communication is received by different communities and of the “communication needs” of these communities. They have also been used to analyse the public acceptability of sanitary measures during the pandemic, including vaccine mandates (McGowan, 2022[70]; Spitale, Biller-Andorno and Germani, 2022[71]). More broadly, countries’ experience with the use of social listening tools has been reviewed by the OECD and good practice principles have been established (OECD, 2021[22]; 2023[64]). Applied to the public acceptability of reforms, social listening would aim to assess the effectiveness of the case for reform and to better understand along which dimensions that case is helping build public support and conversely where it is less well received or more easily contested. In doing so, these efforts may also shed light on important questions that remain outside the framework of the PAT, including the degree of salience of the policy issue and reform proposal in the political and public debate, the use and effectiveness of normative vs. technical language in these debates and the impact that emotions may have on the relative importance of the different dimensions of public acceptability.

New or existing data tools can be used to collect relevant data and analyse the main “narratives” around specific reforms. In a number of OECD countries, social listening capabilities are already contributing to improve policymakers’ understanding of public views and inform their communication around key or sensitive issues (OECD, 2021, pp. 80-98[22]; 2023, pp. 25-26[64]). Existing data tools could support the application of the PAT through targeted forms of social listening and be directed towards one or more of its dimensions, as relevant for particular reforms and contexts. The PAT would in turn provide a framework for organising the data. It should be noted furthermore that narrative approaches have been used to better identify and characterise reform episodes in different policy areas, drawing on qualitative data and textual analysis.32

Various methods (machine learning…) and resources (existing databases, web-scraped data…) can extend these approaches and develop a dedicated data tool to help monitor the public acceptability of specific reforms. Section 2.3.5 below provides an example of the application of such a tool to assess the response to the government’s case for reform in public and policy debates. This illustrative case study uses a tool developed in collaboration with the ISI Foundation33 to provide a retrospective analysis of the policy and public debates around the recent French pension reform, focusing on news media articles and parliamentary texts. For more detail on this case study, see (Parazzoli et al., forthcoming[72]).

What can social listening tell us about public acceptability? An application to the case of the 2023 pension reform in France

Background

Social listening opens new and potentially significant avenues for assessing the public acceptability of reforms through the analysis of non-standard data sources, including large public datasets and digital traces. Leveraging these data, which are often collected at scale and in real-time, could help complement traditional methods for assessing public acceptability, such as opinion surveys or focus groups, and provide policymakers with additional insights that can be used to improve the design and communication of reforms.

To test the potential contribution of social listening for understanding public acceptability and improving reforms, the OECD and ISI Foundation have explored (i) the possible insights that can be drawn from the analysis of two relevant sources of data: news media articles (as a proxy for public debate) and parliamentary speeches (as a proxy for policy debate); and (ii) the value of the PAT as a framework for organising these data, taking the 2023 pension reform in France as a retrospective case study.

Case study: The 2023 French Pension Reform

The French pension reform studied here (Loi du 14 Avril 2023 de Financement Rectificative de la Sécurité Sociale pour 2023) was passed in March 2023 and ratified in April 2023. The main elements of the reform included a gradual increase in the minimum legal retirement age (from 62 years to 64 years) and the bringing forward of an already planned extension of the necessary contribution period for a full pension (from 42 years to 43 years). For more detail on the 2023 French pension reform, see notably (Boulhol and Queisser (2023[73]) and OECD (2023, pp. 15-57[74]).

In addition to its timeliness, a number of reasons made this case highly relevant for analysis:

i. The significance and salience of pension reform: The main elements of the 2023 French pension reform were in line with the trends observed and types of reform implemented in other countries in recent years, with the legal retirement age increasing in over half of OECD countries;

ii. The important role played by issues of public acceptability: Despite being in line with the experience of other OECD countries, the 2023 French pension reform generated large-scale opposition, including strikes and 14 days of national protest by trade unions; and

iii. The 2023 pension reform came after several recent reform efforts in the same area: This included two significant parametric reforms implemented in 2010 (Loi Woerth) and 2014 (Loi Touraine), as well as a proposal for systemic pension reform tabled in 2019 but withdrawn following the outbreak of the COVID-19 pandemic.

Furthermore, all four dimensions of the PAT featured prominently in the case for reform as well as in public debates around it:

• Economic dimension: One of the main stated objectives of the reform was to ensure the financial sustainability of the French pension system and eliminate its projected deficit by 2030. This objective was notably justified by the fact that total pension spending in France is among the highest in the OECD at around 14% of GDP (compared to 9% on average) and estimated to increase by 0.5% to 1% of GDP over the next decade according to projections by the French Pensions Advisory Council (Conseil d’Orientation des Retraites). Furthermore. pension system deficits could credibly be tied to broader concerns about growing public debt in the French case, as future gaps in pension revenue would likely have to covered by government contributions or give rise to difficult renegotiations with the social partners.

• Fairness dimension: The reform was also presented as a means to ensure the continued viability of the current “pay-as-you-go” system and preserve the French model of intergenerational solidarity. Issues of intragenerational fairness were also considered, as the reform included an increase in the minimum contributory pension as part of a larger package designed to improve financial sustainability. However, the case for pension reform tends to be complex and therefore difficult to make on fairness grounds. For example, when looking at intergenerational fairness, in France people aged 65 and older tend on average to be relatively well off with a level of disposable income similar to the rest of the population (though it is projected to fall in the future) and a rate of old age poverty that is among the lowest in the OECD and below the rate of poverty for the general population. Issues related to intragenerational fairness, such as the effects of the reform on vulnerable groups (including women) and inequalities in life expectancy between occupations (as well as their impact on pension wealth) also featured prominently in debates around the reform.

• Behavioural dimensions of Social Change: Due to the complexity of pension systems and pension decisions, behavioural barriers tend to play an important role in the context of pension reform. As in most cases, this was taken into account to a degree in the design of the reform and the key measures of the 2023 French pension reform are to be implemented gradually to allow workers (particularly those close to retirement age) time to adjust. As previously mentioned, complexity and limited knowledge often mean that it is difficult to make the case for pension reform on fairness grounds. Figure 2.2 shows for example that public perceptions in France do not reflect the pension system’s actual performance in terms of protecting pensioners from economic risks, which would be a key element to consider in debates on intergenerational fairness.

• Process dimension: The 2023 French pension reform also presents an interesting case from the point of view of “process legitimacy”. First of all, the government could point to a clear electoral mandate for reform, as reforming the pension system was a central element in the programme of President Emmanuel Macron’s re-election campaign in 2022 and had already been signalled as a priority during his first term (2017-2022). Secondly, the 2023 pension reform was more limited in scope than a previous reform proposal tabled in December 2019 and suspended in March 2020 due to the outbreak of the COVID-19 pandemic. Finally, the 2023 pension reform was adopted through a special procedure (Article 49.3 of the French constitution) rather than through the ordinary legislative process. Article 49.3 allows the government to commit its responsibility and have a legislative proposal adopted without parliamentary approval but subject instead to a vote of no-confidence in the government (motion de censure).

Research design

Research focused on the analysis of (i) news articles published online by leading media outlets as the main data source for studying public debates; and (ii) parliamentary speeches as the main data source for studying policy debates. These data were compared to assess the salience of specific topics relevant to the reform in the news media and in parliamentary discourse. Media articles were obtained from public big data sources (the Global Database of Events – GDELT) and parliamentary speeches from the open data portal of the French National Assembly (Assemblée Nationale – i.e. the lower house of the French Parliament). Similar analysis could potentially also be conducted for other relevant data sources including social media and digital traces.

The analysis proceeded to identify news media articles related to each of the four dimensions of the PAT, based on a series of significant keywords and indicators that were defined by OECD experts and used to characterise relevance to each dimension. For the purpose of the analysis, keywords relating to the PAT’s Behavioural Dimensions of Social Change were limited to a specific subset of dimensions relating to time and risk. Word and document embeddings were used to rank documents according to their similarity to each dimension and identify the most relevant ones (Mikolov et al., 2013[75]). The aim in doing so was to explore the relative importance of the different dimensions of public acceptability in public debates. The working hypothesis, given the general characteristics of pension reform, was that issues related to “process legitimacy” were likely to dominate issues related to “outcome legitimacy”, notably in public debates. This reflects the fact that the complexity of pension reform makes it difficult to build a strong and convincing case on economic and fairness grounds.

In addition to the relative importance of the different dimensions, efforts were made to explore what further possible insights could be drawn by analysing the content of public and policy debates. These efforts focused on (i) the salience and framing of the topics that dominated discussions around the reform, and (ii) the degree to which they may align or differ between public debates and policy debates. In order to do so, the sets of documents relating to the different dimensions of the PAT were analysed using topic modelling, a type of statistical modelling that employs unsupervised machine learning to identify clusters of similar words within a body of text.34 The relevant sets of documents for each dimension of the PAT were then compared across the two different sources of data – news media articles and parliamentary speeches – to assess the degree to which they aligned or not in terms of the main topics they covered.35

Results

Figure 2.10 shows the evolution over time of the volume of media articles related to each of the four dimensions of the PAT, aggregated at a weekly level and normalised over the total of articles published each week for the period between 1^st November 2022 and 15^th June 2023. This analysis tends to confirm the working hypothesis: elements relating to the Process Dimension are dominant in news media coverage of the reform and the focus on these elements increases once the law was introduced in the National Assembly. Conversely, focus shifts away from elements relating to the Economic Dimension and the Risk and Time Dimension, which were initially the most covered.

Figure 2.10. Weekly media articles relating to the different dimensions of the PAT

Note: Weekly number of documents classified as belonging to each dimension of public acceptability in news media articles from major French media outlets over the period 1st November 2022 to 15th June 2023. Articles are taken from 10 selected media sources: Le Monde, Libération, L’Equipe, Le Figaro, Les Echos, La Tribune, L’Humanité, La Croix, Le Point, and Le Parisien. The values are normalised over total documents published each week

Source: Parazzoli et al. (forthcoming[72]), “Taking the Pulse of Reform: What can a multi-channel analysis of media and parliamentary speeches tell us about public acceptability?”

Turning to the content of news media articles, the main topics specific to each dimension were identified. Figure 2.11 presents the results of this analysis for the Economic Dimension, showing the surfaced topics and their frequency (on the left), as well as their evolution over time (on the right). The topics are represented by the three most relevant words related to them – i.e. the so-called “topic descriptors”. The results indicate the salience of the issue of public debt – one of the main objectives of the reform and a key element in its rationale – in news media coverage. Other relevant topics include discussions about inflation, housing, private insurance, and billionaires. Furthermore, the evolution in the coverage of different topics can provide an indication of how the focus of public debate responded to external events and key moments in the legislative process.

Figure 2.11. Media coverage of selected topics for the Economic Dimension of public acceptability

Source: Parazzoli et al. (forthcoming[72]), “Taking the Pulse of Reform: What can a multi-channel analysis of media and parliamentary speeches tell us about public acceptability?”

Finally, the analysis sought to compare the content of news media articles and of parliamentary speeches to identify meaningful analogies and differences. Figure 2.12 shows differences in coverage between media articles and parliamentary speeches for all four dimensions of the PAT. Topics on the left (purple) were more discussed in parliamentary speeches, while those on the right (green) were more discussed in media articles. The analysis across the different dimensions indicates, for instance, that reference to academic research was more common in parliamentary speeches, while topics relating to individual pension decisions – such as pension contributions (cotisation), pension saving plans (Plan d’Épargne Retraite or PER) and common pension investment instruments (Société Civile de Placement Immobilier or SCPI) – featured more prominently in the news media. This may suggest that, while research and expert opinion feed directly into policy debates (notably through parliamentary hearings and reports commissioned by the government or other state institutions), they may have much less impact on public debates.

As a result, the practical and individual-level effects of the pension reform may have received more attention and carried more weight in the public debate (as proxied by news media), while aggregate-level considerations and important issues such as the role of social policy featured less prominently than in policy debates. Interestingly also, looking at the Fairness Dimension, the impact of the reform on women was discussed comparatively more in the news media than in parliament. Taking account of these types of “mismatches” may yield insights that can help strengthen the case for reform by better addressing citizens’ demands and concerns and thereby increase public support for reform. It also highlights the potential benefits that can be drawn from the analysis of non-standard data that are public and collected in real-time.

Figure 2.12. Comparison of the coverage of topics in news articles and parliamentary speeches

Note: Bars leaning on the left (in purple) indicate that the topic was more discussed in parliamentary speeches, while those on the right (green) indicate that the topic was more discussed in media.

Source: Parazzoli et al. (forthcoming[72]), “Taking the Pulse of Reform: What can a multi-channel analysis of media and parliamentary speeches tell us about public acceptability?”

Possible next steps for this work

The PAT provides flexibility in its use. It can be applied and operationalised in several ways, according to policymakers’ needs. The general framework presented in Section 2.2 covers the main dimensions of reform that matter to people and contribute to public acceptability across reform areas and across country settings. It is set at a high enough level to be used or customised by different governments for different types of reform. As such, it could potentially form part of a digital hub on reform that countries can access to carry out preliminary analysis on public acceptability based on the available data. This digital hub would be supported by the setting up of an OECD Perceptions Database, as proposed in Chapter 1. Furthermore, a complementary toolkit could be developed to help governments operationalise the insights on public acceptability provided by the general framework and design reforms that are adapted to their own national contexts and circumstances. Elements of this potential toolkit are presented in Section 2.3.

More specifically, the following applications of the PAT could be considered:

• First, specific implementation of the PAT could be undertaken as part of a concrete reform episode, in collaboration with interested countries. In this context, the general framework and toolkit would support policymakers by helping them assess the conditions for public acceptability and drawing relevant insights from the available data to improve the design, implementation and communication of necessary reforms. These case studies would provide a test of how operational the PAT is in an applied context; how it can be adapted to specific reform areas and national settings; and how suitable indicators can be developed.

• Second, the general framework could be integrated into OECD analysis on the political economy of reform to take better account of public acceptability in policy advice and recommendations. Here, the framework could be applied in a sectoral context (for example, to support the analysis of pension reform in the OECD Pensions at a Glance reports or the analysis of labour-market reform in the OECD Employment Outlook…) or integrated more fully into the economic surveillance conducted in OECD Economic Surveys or Going for Growth reports.

• Third, retrospective applications of the framework could be made to past reform episodes. Doing so would provide an opportunity to (i) draw additional insights from these episodes relating to public acceptability, in line with the work done in Making Reform Happen; and (ii) test the validity of the framework.

References

[79] Alesina, A., S. Ardagna and F. Trebbi (2006), “Who Adjusts and When? The political economy of reforms”, IMF Staff Paper Series, Vol. 53, pp. 1-29, https://www.jstor.org/stable/30036020.

[78] Alesina, A. and A. Drazen (1991), “Why are Stabilizations Delayed?”, The American Economic Review, Vol. 81/5, pp. 1170-1188, https://www.jstor.org/stable/2006912?seq=1.

[24] Alesina, A. et al. (2023), “Structural Reforms and Elections: Evidence from a world-wide new dataset”, Journal of the European Economic Association, https://doi.org/10.1093/jeea/jvad075.

[56] Alesina, A., A. Miano and S. Stantcheva (2023), “Immigration and Redistribution”, The Review of Economic Studies, Vol. 90/1, pp. 1-39, https://doi.org/10.1093/restud/rdac011.

[57] Alesina, A., S. Stantcheva and E. Teso (2018), “Intergenerational Mobility and Preferences for Redistribution”, American Economic Review, Vol. 108/2, pp. 521-554, https://doi.org/10.1257/aer.20162015.

[69] Alfonsi, C. et al. (2022), “Public Communication Trends after COVID-19: Innovative practices across the OECD and in four Southeast Asian countries”, OECD Working Papers on Public Governance No. 55, https://doi.org/10.1787/cb4de393-en.

[76] Aphecetche, T. et al. (2022), “Understanding the Political Economy of Reforms: Lessons from the EU”, European Economy Economic Brief Series, Vol. Economic Brief 070, https://doi.org/10.2765/04105.

[90] Atkinson, A. et al. (2012), “Lessons from National Pensions Communication Campaigns”, OECD Working Papers on Finance, Insurance and Private Pensions, No. 18, OECD Publishing, Paris, https://doi.org/10.1787/5k98xwz5z09v-en.

[86] Aumond, R., V. Di Tommaso and G. Rünstler (2022), “A Narrative Database of Labour Market Reforms in Euro-area Economies”, European Central Bank Working Paper Series No. 2657, https://www.ecb.europa.eu/pub/pdf/scpwps/ecb.wp2657~fed869894e.en.pdf.

[30] Bierbrauer, F., P. Boyer and A. Peichl (2021), “Politically Feasible Reforms of Nonlinear Tax Systems”, American Economic Review, Vol. 111/1, pp. 153-191, https://doi.org/10.1257/aer.20190021.

[80] Bojar, A. and T. Vlandas (2021), “Group-Specific Responses to Retrospective Economic Performance: A multilevel analysis of parliamentary elections”, Politics and Society, Vol. 49/4, pp. 517-548, https://doi.org/10.1177/0032329221989150.

[73] Boulhol, H. and M. Queisser (2023), “The 2023 France Pension Reform”, Intereconomics: Review of European Economic Policy, Vol. 3, pp. 130-131, https://doi.org/10.2478/ie-2023-0025.

[50] Bourguignon, F. and C. Landais (2022), “Micro-Simulating the Impact of Public Policies on Households: Why, how, and which ones?”, Notes du Conseil d’Analyse Economique, Vol. 74/5, pp. 1-12, https://doi.org/10.3917/ncae.074.0001.

[31] Causa, O., A. de Serres and N. Ruiz (2015), “Structural reforms and income distribution”, OECD Economic Policy Papers, No. 13, OECD Publishing, Paris, https://doi.org/10.1787/5js3777lbxzn-en.

[68] Ciminelli, G. et al. (2019), “The Political Costs of Reforms: Fear or reality?”, IMF Staff Discussion Notes, Vol. SDN 19/08, https://doi.org/10.5089/9781513512112.006.

[45] Civil Service (2020), Policymaker Perspectives on Reform, https://publicpolicydesign.blog.gov.uk/wp-content/uploads/sites/255/2021/05/Policymaking-Reform-Full-Report-1.pdf.

[4] D’Arcangelo, F. et al. (2022), “Estimating the CO2 emission and revenue effects of carbon pricing: New evidence from a large cross-country dataset”, OECD Economics Department Working Papers, No. 1732, OECD Publishing, Paris, https://doi.org/10.1787/39aa16d4-en.

[26] Dabla-Norris, E. et al. (2023), “Public Perceptions of Climate Mitigation Policies: Evidence from cross-country surveys”, IMF Staff Discussion Notes, Vol. SDN 2023/002, https://www.imf.org/en/Publications/Staff-Discussion-Notes/Issues/2023/02/07/Public-Perceptions-of-Climate-Mitigation-Policies-Evidence-from-Cross-Country-Surveys-528057.

[44] de Geus, A., E. Thode and C. Weidenfeld (2016), Europe Reforms Labour Markets: Leaders’ perspectives, de Gruyter, https://doi.org/10.1515/9783110365108.

[7] de Mello, L. and M. Dutz (eds.) (2012), Promoting Inclusive Growth: Challenges and Policies, OECD Publishing, Paris, https://doi.org/10.1787/9789264168305-en.

[5] Dechezleprêtre, A. et al. (2022), “Fighting climate change: International attitudes toward climate policies”, OECD Economics Department Working Papers, Vol. No. 1714, https://doi.org/10.1787/3406f29a-en.

[35] Diepeveen, S. et al. (2013), “Public Acceptability of Government Intervention to Change Health-Related Behaviours: A systematic review and narrative synthesis”, BMC Public Health, Vol. 13/756, https://doi.org/10.1186/1471-2458-13-756.

[40] Drazen, A. (2000), Political Economy in Macroeconomics, Princeton University Press, https://www.jstor.org/stable/j.ctv346rwp.

[52] Drazen, A. and V. Grilli (1993), “The Benefit of Crises for Economic Reforms”, The American Economic Review, Vol. 83/3, pp. 598-607, https://www.jstor.org/stable/2117535.

[58] Duch, R. and F. Torres (2023), “Government Audits of Municipal Corruption and Belief Updating: Experimental Evidence - Version 2.0”, CESS Working Paper, https://www.raymondduch.com/files/government-audits-municipal-corruption-belief.pdf.

[48] Dutta-Powell, R., C. Court and C. Gilchrist (2021), Engaging People with Pensions via Digital Dashboards: A rapid evidence review from the Behavioural Insights Team for the Pensions Dashboard Programme, https://www.pensionsdashboardsprogramme.org.uk/wp-content/uploads/2021/06/BIT_PDP_REA_01-06-21.pdf.

[14] Duval, R. and J. Elmeskov (2005), “The Effects of EMU on Structural Reforms in Labour and Product Markets”, OECD Economics Department Working Papers, No. 438, OECD Publishing, Paris, https://doi.org/10.1787/830757326248.

[87] Duval, R. et al. (2018), “A Narrative Database of Major Labor and Product Market Reforms in Advanced Economies”, IMF Working Papers Series, Vol. No. 2018/19, https://www.imf.org/en/Publications/WP/Issues/2018/01/25/A-Narrative-Database-of-Major-Labor-and-Product-Market-Reforms-in-Advanced-Economies-45585.

[25] Duval, R., D. Furceri and J. Miethe (2018), “The Needle in the Haystack: What drives labor and product market reforms in advanced countries?”, IMF Working Paper Series, Vol. No. 2018/101, https://www.imf.org/en/Publications/WP/Issues/2018/05/09/The-Needle-in-the-Haystack-What-Drives-Labor-and-Product-Market-Reforms-in-Advanced-Countries-45796.

[28] Duval, R. et al. (2024), “Preferences for Labour Regulation: Endowments vs. beliefs”, Economic Policy, https://doi.org/10.1093/epolic/eiae021.

[47] Eberhardt, W. et al. (2019), “The Retirement Belief Model: Understanding the search for pension information”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.3205085.

[42] Edwards, G. and D. King (eds.) (2007), Policy Histories and Partisan Leadership in Presidential Studies: The case of social security, Oxford University Press, https://doi.org/10.1093/acprof:oso/9780199217977.003.0013.

[49] Eling, M. and C. Jaenicke (2023), “Pension Literacy is Different”, Journal of Insurance Issues, Vol. 46/1, pp. 30-72, https://www.jstor.org/stable/48733775.

[51] Fernandez, R. and D. Rodrik (1991), “Resistance to Reform: Status quo bias in the presence of individual-specific uncertainty”, The American Economic Review, Vol. 81/5, pp. 1146-1155, http://www.jstor.org/stable/2006910.

[83] Flipo, A. (2021), “Que Pensent les Habitants des Territoires Peu Denses de la Mobilité? Une exploitation textuelle des données du Grand Débat National”, Flux, Vol. 123/1, pp. 54-73, https://doi.org/10.3917/flux1.123.0054.

[85] Garritzmann, J., E. Neimanns and M. Busemeyer (2023), “Public Opinion Towards Welfare State Reform: The role of political trust and government satisfaction”, European Journal of Political Research, Vol. 62/1, pp. 197-220, https://doi.org/10.1111/1475-6765.12501.

[82] Golman, R., D. Hagmann and G. Loewenstein (2017), “Information Avoidance”, Journal of Economic Literature, Vol. 55/1, pp. 96-135., https://doi.org/10.1257/jel.20151245.

[15] Høj, J. et al. (2006), “The Political Economy of Structural Reform: Empirical Evidence from OECD Countries”, OECD Economics Department Working Papers, No. 501, OECD Publishing, Paris, https://doi.org/10.1787/881353527404.

[29] Hoy, C. (2022), “How Does the Progressivity of Taxes and Government Transfers Impact People’s Willingness to Pay Tax?: Experimental evidence across developing countries”, World Bank Policy Research Working Paper Series, Vol. No. 10167, https://openknowledge.worldbank.org/entities/publication/1c090cd1-f029-5c9c-9ecd-474bd1af1000.

[27] Hoy, C. et al. (2023), “Building Public Support for Reducing Fossil-Fuel Subsidies: Evidence across 12 Middle-Income Countries”, World Bank Policy Research Working Paper Series, Vol. No. 10615, https://documents1.worldbank.org/curated/en/099623511222327179/pdf/IDU05ee7729604e8304a85088a8012e5028d7d24.pdf.

[62] Jain, S. and S. Majumdar (2016), “State Capacity, Redistributive Compensation and the Political Economy of Economic Policy Reform”, International Review of Economics and Finance, Vol. 42, pp. 462-473, https://doi.org/10.1016/j.iref.2015.10.025.

[53] Jain, S. and S. Mukand (2003), “Redistributive Promises and the Adoption of Economic Reform”, The American Economic Review, Vol. 93/1, pp. 256-264, https://www.jstor.org/stable/3132172?seq=1.

[63] Khemani, S. (2019), “What Is State Capacity?”, World Bank Policy Research Working Papers Series, Vol. No. 8734, https://doi.org/10.1596/1813-9450-8734.

[33] Maestre-Andrés, S., S. Drews and J. Van Den Bergh (2019), “Perceived Fairness and Public Acceptability of Carbon Pricing: A review of the literature”, Climate Policy, Vol. 19/9, pp. 1186-1204, https://doi.org/10.1080/14693062.2019.1639490.

[84] Martin, M. and M. Islar (2021), “The “End of the World’ vs. the ”End of the Month“: Understanding social resistance to sustainability transition agendas, a lesson from the Yellow Vests in France”, Sustainability Science, Vol. 16, pp. 601-614, https://doi.org/10.1007/s11625-020-00877-9.

[70] McGowan, B. (2022), “World Health Organization’s Early AI-supported Response with Social Listening Platform”, Journal of the Medical Library Association, Vol. 110/2, pp. 273-275, https://doi.org/10.5195/jmla.2022.1398.

[75] Mikolov, T. et al. (2013), “Efficient Estimation of Word Representations in Vector Space”, arXiv preprint, https://doi.org/10.48550/arXiv.1301.3781.

[81] Mus, M., H. Mercier and C. Chevallier (2023), “Designing an Acceptable and Fair Carbon Tax: The role of mental accounting”, PLOS Climate, Vol. 2/10, https://doi.org/10.1371/journal.pclm.0000227.

[43] Nelson, T., Z. Oxley and R. Clawson (1997), “Toward a Psychology of Framing Effects”, Political Behaviour, Vol. 19, pp. 221-246, https://doi.org/10.1023/A:1024834831093.

[37] OECD (2025), “Mind Shift, Green Lift: Six behavioural science trends for environmental policy”, OECD Public Governance Policy Paper Series No. 69, https://doi.org/10.1787/162c5a27-en.

[61] OECD (2024), LOGIC: Good Practice Principles for Mainstreaming Behavioural Public Policy, OECD Publishing, Paris, https://doi.org/10.1787/6cb52de2-en.

[19] OECD (2024), OECD Economic Outlook, Volume 2024 Issue 2, OECD Publishing, Paris, https://doi.org/10.1787/d8814e8b-en.

[23] OECD (2024), OECD Survey on Drivers of Trust in Public Institutions – 2024 Results: Building Trust in a Complex Policy Environment, OECD Publishing, https://doi.org/10.1787/9a20554b-en.

[6] OECD (2023), Economic Policy Reforms 2023: Going for Growth, OECD Publishing, Paris, https://doi.org/10.1787/9953de23-en.

[64] OECD (2023), “Good practice principles for public communication responses to mis- and disinformation”, OECD Public Governance Policy Papers, No. 30, OECD Publishing, Paris, https://doi.org/10.1787/6d141b44-en.

[46] OECD (2023), Main Findings from the 2022 OECD Risks that Matter Survey, OECD Publishing, Paris, https://doi.org/10.1787/70aea928-en.

[59] OECD (2023), On Shaky Ground? Income Instability and Economic Insecurity in Europe, OECD Publishing, Paris, https://doi.org/10.1787/9bffeba6-en.

[74] OECD (2023), Pensions at a Glance 2023: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/678055dd-en.

[67] OECD (2022), “Good practice principles for ethical behavioural science in public policy”, OECD Public Governance Policy Papers, No. 20, OECD Publishing, Paris, https://doi.org/10.1787/e19a9be9-en.

[65] OECD (2022), OECD Guidelines for Citizen Participation Processes, OECD Public Governance Reviews, OECD Publishing, Paris, https://doi.org/10.1787/f765caf6-en.

[89] OECD (2022), OECD Reviews of Pension Systems: Slovenia, OECD Reviews of Pension Systems, OECD Publishing, Paris, https://doi.org/10.1787/f629a09a-en.

[66] OECD (2021), Administrative Capacity Building Self-assessment Instrument for Managing Authorities of EU Funds under Cohesion Policy, OECD Publishing, https://www.oecd.org/regional/regionaldevelopment/ACB_Self_assessment_Instrument.pdf.

[8] OECD (2021), Does Inequality Matter? How People Perceive Economic Disparities and Social Mobility, OECD Publishing, https://doi.org/10.1787/3023ed40-en.

[13] OECD (2021), Inheritance Taxation in OECD Countries, OECD Publishing, https://doi.org/10.1787/e2879a7d-en.

[38] OECD (2021), Main Findings from the 2020 Risks that Matter Survey, OECD Publishing, Paris, https://doi.org/10.1787/b9e85cf5-en.

[1] OECD (2021), Making Better Policies for Food Systems, OECD Publishing, Paris, https://doi.org/10.1787/ddfba4de-en.

[22] OECD (2021), OECD Report on Public Communication: The Global Context and the Way Forward, OECD Publishing, Paris, https://doi.org/10.1787/22f8031c-en.

[11] OECD (2021), Pensions at a Glance 2021: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/ca401ebd-en.

[21] OECD (2020), Innovative Citizen Participation and New Democratic Institutions: Catching the Deliberative Wave, OECD Publishing, Paris, https://doi.org/10.1787/339306da-en.

[20] OECD (2019), Economic Policy Reforms 2019: Going for Growth, OECD Publishing, Paris, https://doi.org/10.1787/aec5b059-en.

[60] OECD (2019), Tools and Ethics for Applied Behavioural Insights: The BASIC Toolkit, OECD Publishing, Paris, https://doi.org/10.1787/9ea76a8f-en.

[16] OECD (2019), “When Do Countries Reform? Insights from Going for Growth and proposals for future work.”, Economics Department ECO/WP1 Discussion Paper, Vol. (2019)18.

[10] OECD (2018), Good Jobs for All in a Changing World of Work: The OECD Jobs Strategy, OECD Publishing, Paris, https://doi.org/10.1787/9789264308817-en.

[9] OECD (2017), Multi-level Governance Reforms: Overview of OECD Country Experiences, OECD Multi-level Governance Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264272866-en.

[3] OECD (2017), The Political Economy of Biodiversity Policy Reform, OECD Publishing, Paris, https://doi.org/10.1787/9789264269545-en.

[18] OECD (2010), Making Reform Happen: Lessons from OECD Countries, OECD Publishing, Paris, https://doi.org/10.1787/9789264086296-en.

[12] OECD (2010), Tax Policy Reform and Economic Growth, OECD Tax Policy Studies, No. 20, OECD Publishing, Paris, https://doi.org/10.1787/9789264091085-en.

[32] Ostry, J., A. Berg and S. Kothari (2018), “Growth-Equity Trade-offs in Structural Reforms”, IMF Working Paper Series, Vol. No. 2018/005, https://www.imf.org/en/Publications/WP/Issues/2018/01/05/Growth-Equity-Trade-offs-in-Structural-Reforms-45540.

[88] Paatero, P. and U. Tapper (1994), “Positive Matrix Factorization: A non-negative factor model with optimal utilization of error estimates of data values”, Environmetrics, Vol. 5/2, pp. 111-126, https://doi.org/10.1002/env.3170050203.

[72] Parazzoli, S. et al. (forthcoming), “Taking the Pulse of Reform: What can a multi-channel analysis of media and parliamentary speeches tell us about public acceptability?”.

[41] Persson, T. and G. Tabellini (2000), Political Economics: Explaining economic policy, MIT Press.

[55] Potter, A. (2018), “What Drives Opposition to Reforms? The role of ex-ante uncertainty”, Political Research Quarterly, Vol. 71/3, pp. 560-572, https://doi.org/10.1177/1065912917749892.

[36] Reynolds, J. et al. (2019), “Public Acceptability of Nudging and Taxing to Reduce Consumption of Alcohol, Tobacco, and Food: A population-based survey experiment”, Social Science & Medicine, Vol. 236/112395, https://doi.org/10.1016/j.socscimed.2019.112395.

[71] Spitale, G., N. Biller-Andorno and F. Germani (2022), “Concerns Around Opposition to the Green Pass in Italy: Social listening analysis by using a mixed methods approach”, Journal of Medical Internet Research, Vol. 24/2, https://doi.org/10.2196/34385.

[34] Stantcheva, S. (2021), “Understanding Tax Policy: How do people reason?”, The Quarterly Journal of Economics, Vol. 136/4, pp. 2309-2369, https://doi.org/10.1093/qje/qjab033.

[77] Stokes, S. (1996), “Public Opinion and Market Reforms: The limits of economic voting”, Comparative Political Studies, Vol. 29/5, pp. 499-519, https://doi.org/10.1177/0010414096029005001.

[17] Tompson, W. (2009), The Political Economy of Reform: Lessons from Pensions, Product Markets and Labour Markets in Ten OECD Countries, OECD Publishing, https://doi.org/10.1787/9789264073111-en.

[54] van Wijnbergen, S. and T. Willems (2016), “Learning Dynamics and Support for Economic Reforms: Why good news can be bad”, The World Bank Economic Review, Vol. 30/1, pp. 1-23, https://doi.org/10.1093/wber/lhu005.

[2] Whaites, A. et al. (eds.) (2015), Putting Political Economy to Use in Aid Policies, OECD Publishing, https://www.oecd.org/dac/accountable-effective-institutions/Governance%20Notebook%201.4%20Hout.pdf.

[39] Williamson, J. (1994), The Political Economy of Policy Reform, Institute for International Economics.

Annex 2.A. Prototype for a Reformers’ Checklist

Examples of relevant questions to include in a general checklist

1st step: Assessing outcome legitimacy – To what extent does the case for reform clearly articulate the equity and efficiency implications of reform for the public?

Purpose of reform

• What is the intended purpose of reform? (Economic grounds; Fairness grounds; Political grounds)

  • Is that purpose clearly established and understood by policymakers?

  • Is there a shared consensus across government on the purpose and objective of reform?

• Is it in line with the way in which the public tends to frame and assess this type of reform?

  • Has analysis been conducted to understand how the public perceives the purpose of reform (through desk research, surveys or expert opinion/interviews)?

• How well is the purpose of reform communicated to the public?

  • Is the intended purpose of reform coherently reflected in communication campaigns?

  • Are good practices and OECD guidelines in terms of public communication being followed?

  • Has the public’s reception of the case for reform been explored through surveys, focus groups or other methods (social listening…)?

Prospects for reform

• Do policymakers clearly understand the economic and distributional impacts of reform?

  • What models and tools are in place to assess the economic and distributional impacts of a proposed reform (micro-simulation models, Distributional Impact Assessments…)? Where relevant, are they aligned with good practices?

  • Has the performance of the existing policy setting been properly assessed or benchmarked? Has the economic impact of reform been properly assessed?

  • Has the distributional impact of the existing policy setting and proposed reform been assessed?

  • Does the assessment indicate that:

    • A majority of the population is expected to benefit from reform (Economic dimension)?

    • The reform is expected to reduce inequality overall or for the key groups affected (Fairness dimension)?

• Does the case for reform take account of the public’s views on the need for reform?

  • How well aligned is the case for reform with the public’s understanding of:

    • The need for reform?

    • The likely impacts of reform on both grounds (i.e. how difficult is reform likely to be)?

  • What relevant perception gaps have been identified?

  • Have communication campaigns sought to address the key perception gaps identified? How effective have they been in shifting the relevant perceptions and attitudes?

Key challenges for reform and means for addressing them

• Does the reform strategy seek to improve the prospects for reform?

  • Have the economic impacts of reform been effectively communicated to the public? Have the results of Distributional Impact Assessments been effectively communicated to the public?

  • Have communication campaigns sought to address key perception gaps? How effective have they been in shifting the relevant attitudes and perceptions?

  • What more can be done to demonstrate the need for reform and/or the desirability of the expected outcomes?

• Does the case for reform take account of the main sector-specific challenges relating to equity and efficiency?

  • Have the main challenges been identified, based on analysis of the literature, consultation with key interest groups and expert advice?

  • Have trade-offs between equity and efficiency been clearly presented as part of the case for reform?

2nd step: Taking account of the behavioural dimensions of social change – What behavioural barriers may undermine support for reform?

State of Knowledge

• Can the public’s level of knowledge and openness to information be assessed? Have the implications been taken into account when designing communication around the reform?

• How well does the public understand the underlying policy issue? / How well do people feel they understand the policy issue?

• How likely is the public to engage with quantitative evidence? Is this reflected in the type of information and modes of communication used to build public support for reform?

• Do salient analogies exist with successful reforms in other areas?

Uncertainty

• What sources of uncertainty may affect public support for reform?

• Who do people perceive to be the main “winners” and “losers” of reform?

• How well do these perceptions align with the expected “winners” and “losers”?

Margins to adapt

• Have the broader behavioural costs been identified and considered when designing and implementing reforms?

• What behavioural adjustments are demanded by reform? Have groups with limited margins to adapt been identified?

• Have appropriate support, compensation or transition measures been put in place? Is the public sufficiently aware of these accompanying measures?

Area-specific challenges

• What additional behavioural traits need to be considered in this specific reform area?

  • Example of pension reform: Forms of time inconsistency (hyperbolic discounting, myopia…), information avoidance, attitudes towards risk and endowment effects...

• Have the implications of these behavioural traits for the design and communication of reform been taken into account?

  • Example of pension reform: Use of non-standard direct cash transfers (one-off lump-sum payments...) in the case of myopia; use of alternative modes of communication or choice architecture in the case of limited public knowledge, low openness to information or high complexity of the issue…

• Are behavioural barriers addressed through ethical means?

  • Have OECD guidelines and good practice principles been applied to ensure the ethical nature of interventions throughout the reform process?

3rd step: What elements of process are needed to support the case for reform?

Preliminary structural question

• How prominently is process legitimacy likely to feature in the case for reform?

  • In light of potential difficulties in building public support on grounds of content alone (prospects for reform; behavioural barriers)?

  • In light of past experience in this specific reform area?

Administrative and technical Capacity

• Is there appropriate administrative and technical capacity in place to deliver on reform objectives?

  • Have capacity needs been assessed and integrated at all stages of the reform process?

  • Are institutions in place that can manage the reform effectively or is there a need to create or strengthen such institutions?

• Has the role of administrative capacity in building public support been taken into account (i.e. its role in addressing key challenges identified under previous steps of the checklist)?

  • Does limited capacity create sources of uncertainty that may undermine support for reform?

  • Is additional capacity needed to help people adapt to reform and facilitate social change?

Evidence base

• How reliable is the evidence base supporting the case for reform?

  • Are there reliable indicators or estimates for the key outcomes impacted by reform? Are there estimates or counterfactuals for these outcomes in the absence of reform?

  • What relevant perceptual and behavioural data are available to understand public views of the policy issue? Where are the main gaps?

• How well is the evidence used?

  • To what extent has the design of reform been informed by the evidence?

  • How well has the evidence been communicated to the public? Has focus been put on ensuring that the messaging is coherent and effective, in line with existing good practices?

  • Are there trusted information providers that can communicate the evidence to the public?

  • Can limited public knowledge, uncertainty and gaps in trust limit the effectiveness of the evidence in making the case for reform? What type of information can help overcome these behavioural barriers?

Political Conditions

• Have the key political conditions for reform been satisfied?

  • Is there an electoral mandate for reform and is it clearly perceived by the public?

  • Does the reform have clearly identifiable “owners” and how are they perceived by the public?

  • How united and cohesive is government in support of reform?

• Is the expected timeframe for design, adoption and implementation of reform realistic?

  • Does the timeframe for reform take account of electoral horizons?

  • Does the timeframe for reform allow sufficient time for effective communication and consultation, in light of the challenges identified?

Stakeholder engagement

• How effective and inclusive has consultation with key stakeholders been?

  • Have OECD guidelines and good practices on stakeholder engagement been applied?

  • Is there a strategy for engaging with the constituencies threatened by reform?

  • Have efforts been made to engage with and mobilise constituencies in favour of reform?

• Has the role of stakeholder engagement in building public support been taken into account?

  • Can consultation help address challenges related to uncertainty about reform or identify ways to facilitate adaptation when margins are limited?

Notes

← 1.  This is notably done through the Economic Policy Reforms: Going for Growth series of flagship reports, which focus on structural reforms.

← 2.  For a discussion on the complementarity between quantitative and qualitative analysis of the political economy of reform and the benefits of combining them, see Tompson (2009, pp. 15-23[17]).

← 3.  These recommendations are in line with similar analyses by the European Commission (Aphecetche et al., 2022[76]) and the International Monetary Fund (Ciminelli et al., 2019[68]).

← 4.  This conclusion was notably underlined at the 2019 OECD conference on The Changing Political Economy of Reform which took stock of new challenges and areas for research on the political economy of reform ten years after Making Reform Happen. The proposal to create the OECD Expert Group on New Measures of the Public Acceptability of Reforms to explore possible responses to these new challenges was a major outcome of this conference.

← 5.  The multidimensional framework developed in Stokes (1996[77]) provides an interesting example of this type of approach. This framework extends traditional “economic voting” models by including a broader range of motivations that can explain public responses to reforms, beyond individual self-interest and expected payoffs. In order to do so, Stokes (1996[77]) identifies a series of key grounds through which citizens assess the economic effects of reforms, based on a review of the literature on economic voting and on the study of selected cases of structural reforms in Latin America and Eastern Europe in the 1980s-1990s. In addition to the economic and distributional impacts of reform, this broader approach highlights two additional elements: (i) the impact of reform on public expectations of future economic performance (i.e. how citizens perceive the outcomes of reform and whether a reform is “working” or not); and (ii) the impact of reform on public attitudes towards government (i.e. how and to what extent citizens attribute responsibility for the outcomes of reform). On this basis, the framework developed in (Stokes, 1996[77]) distinguishes 4 main dimensions that contribute to shape the public acceptability of reforms: (i) economic impacts; (ii) distributional impacts; (iii) intertemporal arbitrage; and (iv) government responsibility.

← 6.  See for example Alesina and Drazen (1991[78]), Drazen (2000[40]), Persson and Tabellini (2000[41]), Alesina, Ardagna and Trebbi (2006[79]) for classical formulations of this challenge.

← 7.  This conclusion is also echoed in the sectoral guidance provided by Making Reform Happen in different policy areas, including for example labour market reform (OECD, 2010, pp. 56-57[18]) and tax reform (OECD, 2010, p. 103[18]).

← 8.  A broad definition of “behavioural dimensions” is needed here as barriers to behavioural change may often be informational or cognitive in nature. When assessing the public acceptability of a reform, understanding how citizens view the underlying policy issue can help determine how receptive they will be to arguments on efficiency and equity grounds and how effective the case for reform is likely to be. Similarly, citizens’ capacity to make choices in a new environment or adapt effectively to new incentives and rules will depend in part on levels of public understanding and on whether or not they have access to tractable and actionable information as well as effective support.

← 9.  Tompson (2009[17]) for example provides both quantitative evidence (through cross-country econometric analysis) and qualitative evidence (through case studies) of the importance of institutional and political factors linked to process in the successful implementation of reforms.

← 10.  Under the basic assumption of economic voting, this constitutes the only dimension that matters for public acceptability. Even here, however, policy preferences and the relevant economic variables through which economic performance is assessed are likely to vary across different groups. See for example Bojar and Vlandas (2021[80]) for evidence on group heterogeneity in terms of policy preferences and key variables used to assess perceived economic performance.

← 11.  This point is in line with the Expert Group’s recommendation to focus on People’s understanding and evaluation of the “state of society” as a key factor to measure (see Section 1.3.1 of the previous chapter for further detail).

← 12.  This point is in line with the Expert Group’s recommendation to focus on Policymakers’ perceptions of public attitudes as a key factor to measure (see Section 1.3.1 of the previous chapter for further detail).

← 13.  This point is also in line with the Expert Group’s recommendation to focus on Policymakers’ perceptions of public attitudes as a key factor to measure (see Section 1.3.1 of the previous chapter for further detail).

← 14.  This may include, for example, reforms for which the stated objective is to maintain social stability. Conversely, reforms that were originally intended to address economic or fairness goals may instead be perceived as being motivated by electoral concerns and political signalling, notably if the case for reform is not clearly and convincingly made to the public or is effectively undermined by opponents. This latter case further underlines the relevance of assessing whether public perceptions of the purpose of reform are aligned with the purpose around which policymakers are framing the case for reform.

← 15.  Mus, Mercier and Chevallier (2023[81]) tests the implications of mental accounting for public policy and provides evidence that “matched earmarking” can help increase public support for carbon taxation by clearly linking carbon tax revenue to spending on green projects. These results are replicated for health-related earmarking of tobacco tax revenue (another consumer tax) and social-related earmarking of inheritance tax revenue (a non-consumer tax).

← 16.  As exemplified for instance by the use of matching donations in philanthropy.

← 17.  On the issue of information avoidance, see Golman, Hagmann and Loewenstein (2017[82]). For evidence on knowledge gaps and misperceptions relating to the tax system, see notably Stantcheva (2021[34]) which uses large-scale surveys to study the public’s understanding of income tax and inheritance tax in the US. Results show that US respondents significantly over-estimate the share of households that pay the top income tax rate (an estimated 20% on average against an actual share of less than 1%) and underestimate the share of households that pay no income tax (an estimated 25% against an actual share of 44%). They also overestimate the share of households that pay the estate tax by several orders of magnitude (an estimated 36% on average against an actual share of less than 0.1%).

← 18.  On good practices in the area of public communication, see (OECD, 2021[22]; 2023[64]). On good practices in the communication of pension information, see for example Atkinson et al. (2012[90]). See also the case of Slovenia, which uses a staged approach involving (i) communicating the problem; (ii) communicating solutions to the problem; and (iii) communicating entitlements (OECD, 2022[89]).

← 19.  See Stantcheva (2021[34]) and Dechezleprêtre et al. (2022[5]) on the impact of messaging on support for tax and environmental policies respectively. See also Duch and Torres (2023[58]) on the provision of audit information and its impact on beliefs about corruption and integrity at the municipal level in Chile.

← 20.  For example, populations with limited alternative mobility or consumption options in the case of environmental taxes; or workers close to retirement age in the case of pension reform. Dechezleprêtre et al. (2022, p. 8[5]) captures this idea well: “Opposition to climate policies is strongly correlated with lower availability of public transportation, more reliance on cars, and, to a lesser extent, higher gas expenses”.

← 21.  See for example Flipo (2021[83]) and Martin and Islar (2021[84]) on the case of the Gilets Jaunes in France. In public debates on environmental taxes, this idea of unfair differentiated treatment based on the capacity to adapt is typically expressed by the argument that “the poor change their behaviour and the rich just pay the tax”.

← 22.  Focus on administrative and technical capacity should not be limited to the implementation stage. Before implementation, the assessment of capacity has an important role to play in informing the design of reforms, notably by identifying the potential “winners” and “losers” from reform, ensuring reform objectives are aligned with available means of implementation, and determining whether compensation measures can be delivered effectively. Following implementation, evaluation of the impacts of reform (including public satisfaction) and communication of the results of evaluation can help ensure that capacity remains adapted and public support for the reform is maintained.

← 23.  The European Commission’s 2022 Communication on Better Assessing the Distributional Impact of Member-states’ Policies offers an interesting example through the guidance and support it provides on the use of ex-ante Distributional Impact Assessments to assist in the design of reforms.

← 24.  Independent fiscal, productivity and regulatory institutions can help address challenges linked to uncertainty and limited public understanding of policy issues, for instance by providing trusted and authoritative evaluations of the impact of tax reforms (see p.19). The Netherlands Bureau for Economic Policy Analysis (CPB) and the Australian Productivity Commission (APC) were highlighted as good examples of such trusted institutions in (OECD, 2010, pp. 32-33[18]).

← 25.  For example, better indicators on the quality of public spending were highlighted as one possible way to help overcome limited willingness to pay in the context of tax and benefits reforms (see Box 2.5).

← 26.  Some evidence exists linking trust in government and institutions with support for reform (OECD, 2024[23]), including in the area of social policy (Garritzmann, Neimanns and Busemeyer, 2023[85]).

← 27.  Including those of constituencies that are impacted by reform, but whose interests may not be well represented in the public debate – such as youth and future workers in the case of pension reform.

← 28.  It should be noted that the prototype described in Annex 2.A represents a general and comprehensive checklist designed to fully cover the different dimensions of the PAT across reform areas. Implementing this ideal checklist may require access to an extensive range of both objective and perceptual/behavioural data, some of which may be lacking or costly to collect. In practice, however, shorter and more targeted versions of this general checklist can be drawn up for specific reform areas and contexts, based notably on the identification of key challenges and prioritisation of the different dimensions provided by the sectoral roadmaps (see following section). These efforts would be further supported by the establishment of an OECD Perceptions Database, as suggested in Section 1.4, to compile key perceptual and behavioural data and help address the main data gaps.

← 29.  This standardised checklist is designed to help policymakers ensure that ethical considerations are integrated into the application of behavioural science at all stages of the policy process and includes a series of prompting questions to accompany and guide the use of the checklist.

← 30.  For example, the outsized importance of process dimensions in the case of pension reform highlighted by the Expert Group can be explained by a common failure to establish a sufficiently strong and popular case on equity and efficiency grounds due to the difficulties associated with the informational/behavioural aspects of the issue (complexity of pension systems; limited technical understanding; information avoidance…). As a result, pension reform may end up relying entirely on process legitimacy unless these challenges are better addressed. This would also imply limited opportunities for building outcome legitimacy and creating buy-in and support for the actual economic and distributive outcomes of pension reform.

← 31.  These “reform-making” and “reform-breaking” data would be used to complement simpler traditional measures of public acceptability, such as polling measures of support for reform or focus group results from key constituencies. They present the advantage of providing a finer picture of the elements that contribute to the public acceptability of a reform and may give a clearer and earlier indication of why the case for a particular reform is likely to succeed or fail and on what grounds.

← 32.  See for example Duval et al. (2018[87]) and Aumond, Di Tommaso and Rünstler (2022[86]).

← 33.  The ISI Foundation (Istituto per l’Interscambio Scientifico) is a research institution that is based in Turin, Italy. It specialises in the application of data science and computational modelling to issues of environmental sustainability, social inclusion and public health. For more information on the ISI Foundation and its work, see: https://www.isi.it/.

← 34.  Non-negative matrix factorisation (NMF) was used for this task (Paatero and Tapper, 1994[88]). To avoid issues with low topic coherence and poor usability of information, the analysis focused on extracting a smaller number of topics – specifically, 10 topics from each corpus.

← 35.  Coupled matrix factorisation with non-negativity constraints was applied to compare news media articles and parliamentary speeches on the same dimension of the PAT. This technique enables the analysis of two distinct bodies of text or “corpora” by searching for latent topics in two collections of documents simultaneously. Through this analysis, a set of common topics is identified across the two sets of documents. Mismatches between them in terms of coverage can then be detected by assessing the presence of these common topics in each corpus.