Discrimination deprives individuals of their human dignity and opportunities to realise their full potential. Over the past five years, the EU has intensified efforts to fight discrimination, particularly against people from racialised communities, LGBTI people, persons with disabilities and people from religious minorities. However, measuring the effectiveness of anti‑discrimination initiatives can be difficult due to data collection and methodological challenges. In the absence of comparable, comprehensive official data sources, this report primarily analyses self-reported discrimination survey data to study the extent and effects of discrimination. After assessing the state and effects of discrimination, this report analyses OECD EU countries’ laws, policies and mainstreaming responses, using data from the OECD Anti-Discrimination Questionnaire. It highlights good practices such as comprehensive anti-discrimination laws and policies that are tailored to the needs of groups at risk of discrimination.
Combatting Discrimination in the European Union
Abstract
While discrimination can be defined in straightforward terms – as any distinction, exclusion, restriction or other differential treatment of a person because of their characteristics or beliefs, which constrains their human rights1 – its simplicity belies the complex, and sometimes subtle, ways in which discrimination operates within societies. Discrimination can stem from the actions of individuals, organisations, institutions or via systemic practices. It may involve explicit, intentional acts – such as verbal antagonism, avoidance, segregation and physical attack – or unconscious behaviours based on prejudicial attitudes. Moreover, discrimination can have a statistical or profiling basis, whereby a decision is made about a person based on their group belonging rather than their individual characteristics. Finally, the rules and norms of organisations and systems may function in ways that lead to differential treatment and exclusion, resulting in structural discrimination (National Research Council, 2004[1]).
Over the past five years, the European Union (EU) and its Member States have stepped up efforts to grapple with the complexity of discrimination – reforming policies and legislation to target the attitudes, stereotypes and structural factors that undergird and perpetuate discrimination and deprive people of the ability to enjoy their human rights and equal opportunities (Chapters 4 and 5). Key among these developments at the EU level are the Union of Equality, and its attendant Anti‑Racism Action Plan 2020-2025, the Roma Strategic Framework for Equality, Inclusion and Participation for 2020-2030, the LGBTIQ Equality Strategy 2020‑2025, the Gender Equality Strategy 2020‑2025 and the Strategy for the Rights of Persons with Disability 2021-2030, along with the 2024 legislative amendments to enhance the power of equality bodies to support victims of discrimination and increase public understanding of discrimination. These initiatives augment the EU Treaties, Equality Directives, the Charter of Fundamental Rights and the European Pillar of Social Rights, which emphasise the foundational importance of non‑discrimination and equality (European Parliament, 2012[2]).
Despite these commitments to equality, discrimination remains a persistent reality and challenge in the EU. Every day, discrimination limits people’s access to jobs, services and opportunities, and exposes them to violence and hate. Unfortunately, however, the ways in which discrimination affects people’s lives in the EU remain difficult to quantify – particularly in terms of how the well-being of people at risk of discrimination compares to the rest of the population – because data availability varies across Member States and protected grounds. Disaggregated data covering racialised2 communities and lesbian, gay, bisexual, transgender and intersex (LGBTI) people are especially lacking (Subgroup on Equality Data of the High Level Group on Non-Discrimination, Equality and Diversity, 2023[3]; 2021[4]).
To support the development of new evidence, the European Commission has co-funded this report to examine the state and effects of discrimination and identify policy and legal gaps, good practices and opportunities to further mainstream non‑discrimination, equality and inclusion. This report focuses on discrimination against people from racialised communities, LGBTI people, persons with disabilities, people from religious minorities, young people and older people. An intersectional approach is taken, where data are available, to examine how people’s sex interacts with their experiences of discrimination on the basis of their ethnicity, race, sexual orientation, gender identity, disability, religion and/or age. Understanding how discrimination affects people’s lives is essential for ensuring that policies and laws are well‑functioning. An intersectional lens assists with identifying the risks people face, which in turn, helps in designing policies that meet their needs and are attuned to their circumstances (Chapter 4).
This introductory chapter sets out the approach to examining discrimination in the EU by first emphasising why it is important to understand and combat discrimination (Section 1.1). It then explores measurement difficulties (Section 1.2) and describes the empirical strategies used throughout this report to assess the state of discrimination and ways governments are responding (Section 1.3). In the absence of official data sources, this report draws on a variety of surveys and opinion polls, as well as the results of the OECD Anti‑Discrimination Questionnaire, to highlight the ways in which data can be used to understand the magnitude and effects of discrimination, as well as policy responses – thereby demonstrating the potential analysis that could be undertaken in the event that official data collections become richer. The OECD Anti‑Discrimination Questionnaire collected information on national anti‑discrimination laws, policies and mainstreaming initiatives to promote the equality and inclusion of groups at risk of discrimination (see Chapters 3, 4 and 5 for more information on the questionnaire). Responses are current as of April 2024 and sourced from 21 out of 22 OECD EU countries.
Discrimination violates the principles of respect for the inherent dignity and the equal rights of every person, which include economic, social, political and civil rights.3 EU Member States are bound to uphold equality and non-discrimination, as per the EU legislative framework,4 the European Convention on Human Rights and the Universal Declaration of Human Rights and other United Nation declarations.5
In concrete terms, the EU Charter of Fundamental Rights prohibits discrimination based on any grounds such as sex, race, colour, ethnic or social origin, genetic features, language, religion or belief, political or any other opinion, membership of a national minority, property, birth, disability, age or sexual orientation. Further, secondary legislation, including the Racial Equality Directive (2000/43/EC) and Employment Equality Directive (2000/78/EC), articulate the acts that constitute discrimination and unfair treatment (Box 1.1).
The Racial Equality Directive (2000/43/EC) and Employment Equality Directive (2000/78/EC) expressly prohibit the following forms of discrimination and unfair treatment.
Direct discrimination – where one person is treated less favourably than another is, has been or would be treated in a comparable situation.
Indirect discrimination – where an apparently neutral provision, criterion or practice would put persons at a particular disadvantage compared with other persons, unless that provision, criterion or practice is objectively justified by a legitimate aim and the means of achieving that aim are appropriate and necessary.
Harassment – unwanted conduct with the purpose or effect of violating the dignity of a person and of creating an intimidating, hostile, degrading, humiliating or offensive environment.
Discrimination by instruction – whereby a person is ordered or induced to discriminate against another.
Victimisation – any adverse treatment or adverse consequence as a reaction to a complaint or to proceedings aimed at enforcing compliance with the principle of equal treatment.
Many countries’ laws go beyond the minimum EU protections, such as by prohibiting discrimination by association (where a person is discriminated against because of their relationship to another person who has experienced discrimination), or discrimination by assumption (which is based on the incorrect assessment of the characteristics or beliefs of the person facing discrimination).
All EU Member States have incorporated the EU laws into national legislation, although there are differences across countries. Many countries have more expansive anti-discrimination laws than required by the EU standards, for example, in terms of the areas of life covered, the number of protected grounds and the forms of discrimination that are prohibited (Chapter 4).
The legal prohibitions on discrimination are augmented by policies in many EU countries to promote equality and inclusion. Under the auspices of the Union of Equality, many EU countries have created national equality strategies to mainstream inclusion throughout their policymaking processes and ensure that policies are tailored to the needs and circumstances of groups at risk of discrimination (Chapters 4 and 5). This report showcases policy and mainstreaming initiatives in a number of areas: from awareness campaigns, to employment, education, health and justice programmes, as well as stakeholder engagement processes and equality data collection practices.
Taken together, the legal and policy responses to discrimination emphasise the importance that the EU and its Member States place on creating more equal and inclusive societies. The prominence given to equality and inclusion in the past five years is not just a recognition of the moral imperative of upholding human rights, but an indication of the perniciousness of the consequences of discrimination. This report shows that discrimination is widespread in the EU, with 56% of people who identify as part of an at-risk group based on their ethnicity or skin colour, disability status, religion, sexual orientation or gender identity stating that they experienced discrimination in the 12 months to May 2023 – up from 46% in 2019 (Chapter 2). This rise reflects a combination of demographic shifts, recent major events like COVID-19, and a growing public awareness of rights and discrimination. When compared to those who have not experienced discrimination, people with self‑reported experiences of discrimination have limited income-earning opportunities, live with housing and financial stress, are often subjected to violence and fear, and are more likely to be lonely and have poor mental and physical health outcomes. These consequences come at a very significant cost to the individuals directly affected, as well as to society as a whole, in terms of lost economic potential and threats to social cohesion (Chapter 3).
Despite the recognition of the importance of combatting discrimination, there are a number of challenges to understanding its extent, nature and effects, as well as the effectiveness of policy responses. In Europe, the paucity of data that can be used to assess discrimination is often noted by the European Commission (2023[3]; 2021[4]) and advocacy groups such as the European Network Against Racism (n.d.[5]; 2015[6]; 2014[7]). In addition, there are methodological issues to measuring discrimination empirically (National Research Council, 2004[1]). This section discusses these challenges.
Patchy data collection
Discrimination analysis in the EU is hampered by the lack of valid and comparable data covering all at‑risk groups. While all EU Member States publish at least some official data on dimensions of well-being6 disaggregated by sex, age and physical limitations on daily activities (as a proxy for disability status), data are not systematically collected for other groups at risk of discrimination. Disaggregated data covering racialised communities and LGBTI people are especially lacking (Subgroup on Equality Data of the High Level Group on Non-Discrimination, Equality and Diversity, 2023[3]; 2021[4]).
Among some EU Member States, the lack of well-being data on at-risk groups (especially people from racialised communities) stems from the sensitivity of the data, historical considerations or concerns that the data will be used to reinforce negative stereotypes (Balestra and Fleischer, 2018[8]; Farkas, 2017[9]; Subgroup on Equality Data of the High Level Group on Non-Discrimination, Equality and Diversity, 2021[4]; 2023[3]). Results from the OECD Anti-Discrimination Questionnaire indicate that more than two‑thirds of the 21 OECD EU respondent countries raised concerns about the effects on at-risk groups of collecting and using data, and over half mentioning privacy and legal concerns (Chapter 5).
These concerns are not, however, felt universally. As discussed in Chapter 5, there is overwhelming public support and repeated calls to expand the collection of data on all groups at risk of discrimination by civil society organisations such as the European Network Against Racism (n.d.[5]), and the European Commission (as exemplified by the Subgroup on Equality Data of the EU High Level Group on Non‑Discrimination, Equality and Diversity). This support is predicated on data being used for the benefit of at-risk groups, and being collected in ways that avoid stigmatising at-risk groups, including by responding seriously to at-risk groups’ concerns, providing at-risk groups with the tools to use the data to understand and advocate for their communities, and collecting personal data only where essential and on a voluntary basis (permitting a non-response option for those who do not wish to provide this information) (United Nations Office of the High Commissioner for Human Rights, 2018[10]; Balestra and Fleischer, 2018[8]). In the EU, the collection and use of data need to comply with Article 9 of the General Data Protection Regulation (Regulation (EU) 2016/679), which stipulates the conditions in which personal data7 can be collected and processed. These criteria include requiring an individual to give explicit consent to the processing of their data for specified purposes or that the data must be necessary for protecting a person’s vital interests, or that processing the data has a substantial public interest.
As calls to expand data collections grow and protocols for the processing and use of data are developed, some EU countries are investing in data that can be used to elucidate the experiences of groups at risk of discrimination. For example, Malta’s Census of Population and Housing (2021[11]) includes information8 on characteristics or beliefs that affect people’s risk of discrimination – such as their sexual orientation, gender identity, racial origin, religion, languages they spoke when growing up, and long-lasting physical and mental limitations. Bulgaria, Czechia, Ireland, Hungary, Lithuania, Poland, Romania and the Slovak Republic collect information on people’s ethnic, racial, national or cultural origin/background and their religion in their censuses (Czech Statistical Office, 2021[12]; Central Statistics Office Ireland, 2023[13]; Hungarian Central Statistical Office, 2022[14]; Ministry of Foreign Affairs Republic of Latvia, 2024[15]; Ministry of Justice Republic of Latvia, 2022[16]; Statistics Lithuania, 2021[17]; Central Statistical Office Poland, 2021[18]; National Statistics Institute Portugal, 2023[19]; Statistical Office of the Slovak Republic, 2021[20]; Republic of Bulgaria National Statistical Institute, 2021[21]) (National Institute of Statistics Romania, 2021[22]; National Institute of Statistics Romania, 2013[23]). Religious beliefs and affiliations (but not ethnic or racial origin) are also collected in the national censuses of Austria, Croatia, Estonia, Finland, and the Netherlands (Statistics Austria, 2023[24]; Croatian Bureau of Statistics, 2021[25]; Statistics Estonia, 2021[26]; Statistics Finland, 2023[27]; Statistics Netherlands, 2023[28]).
Outside of national censuses, some countries collect information on people’s ethnic origins and religions in official surveys (e.g. Portugal, Box 1.2) or registers (e.g. Latvia) (National Statistics Institute Portugal, 2023[19]; Ministry of Foreign Affairs Republic of Latvia, 2024[15]; Ministry of Justice Republic of Latvia, 2022[16]; Balestra and Fleischer, 2018[8]). In addition, population-based surveys have been used to collect information on people’s sexual orientation and gender identity, including in Belgium, Denmark, Finland, France, Germany, Ireland, Italy, the Netherlands and Sweden, although in many cases these surveys are not regularly conducted (Subgroup on Equality Data of the High Level Group on Non-Discrimination, Equality and Diversity, 2023[3]; OECD, 2019[29]). (See Chapter 5 for more information on data collections on groups at-risk of discrimination and emerging practices in OECD EU countries.)
While these developments in national data collections enable better understanding of the experiences and outcomes of groups at risk of discrimination, differences in how national survey questions are formulated make it difficult to compare results across European countries, which is important for studies, like this one, that cover the EU. Questions that capture people’s identities and characteristics reflect the cultures, histories and demographic makeup of each EU Member State. For instance, in Ireland people can choose if they identify as White Traveller, but such a category is not included in the Maltese census. Moreover, nationally representative surveys are rarely sufficiently granular to analyse gaps in health, education, employment, income and social protection between at-risk groups and the general population.
In the absence of official data that cover all at-risk groups in ways that enable comparisons across countries, the European Agency for Fundamental Rights (FRA) conducts targeted surveys that capture information on the experiences of LGBTI people, Roma and immigrants and their descendants (2020[30]; 2023[31]; 2022[32]); and European and other international general social surveys and opinion polls include questions on people’s characteristics, identities and experiences of discrimination – examples include the Discrimination in the EU Eurobarometer and the Opportunities Module of the 2022 OECD Risks that Matter Survey (Section 1.3). Finally, cross-national statistical offices and networks are developing guidance to harmonise the collection of data on groups at risk of discrimination. For example, Eurostat is chairing an Equality and Non‑Discrimination Statistics Taskforce to improve official data on at-risk groups, including by enhancing statistical coverage and comparability. The taskforce is expected to publish data collection recommendations in 2026 (European Commission, 2024[33]). Further, the United Nations Praia Group on Governance Statistics is developing guidance on using survey and administrative data sources to measure discrimination and disadvantage consistently across countries, which is anticipated to be released in 2026 (United Nations Economic and Social Council, 2024[34]).
All OECD EU countries collect information that can be used as proxies for ethnicity – such as country of birth, parents’ country of birth or year of arrival or citizenship – in official statistical surveys, registers or censuses (Balestra and Fleischer, 2018[8]; Valfort, forthcoming[35]). While these data closely map to ethnicity for some people, they are not perfect proxies for everyone, particularly in societies that are becoming more ethnically and racially diverse (Gill, Bhopal and Kai, 2005[36]). Indeed, the United Nations (2017[37]) recommends that country of birth or citizenship should not be used to derive a person’s ethnicity. In the case of Roma people, the largest ethnic non-territorial minority of Europe, questions of country of birth or citizenship would leave them without the option to identify as Roma. Similarly, a question related only to race, but not ethnicity, would leave a Roma person without an option to self-identify.
The drawbacks of using citizenship as a proxy for ethnicity are clearly visible in analysis of the 2023 Survey on Living Conditions, Origins and Trajectories of the Resident Population in Portugal. Almost 97% of people who identify as Gypsy are Portuguese citizens, compared to 70% of people who identify as Black and 40% of people who identify as Asian are Portuguese citizens (Figure 1.1). These ethnic breakdowns are masked when only citizenship questions are asked in official surveys, which in turn, renders invisible the diversity of the population, and the trends, outcomes and experiences of different population groups (United Nations, 2017[37]).
Malta has also recently started to include self-identified ethnicity in official statistical publications. In 2021, Malta included racial origin in the Census of Population and Housing for the first time (Sansone, 2021[38]). Like Portugal, results from Malta show that large proportions of ethnic minority groups have Maltese citizenship, namely 45% of people who identify as having more than one racial origin and 20% of people identifying as Arab or Hispanic or Latino are Maltese citizens (National Statistical Office Malta, 2023[39]). Recent developments in Portugal and Malta follow longstanding efforts in Ireland and Eastern European countries to collect information on people’s self-identified ethnicity in national censuses or registers. Where countries collect information on self-identified ethnic, racial or national origin, they also typically ask questions related to country of birth and spoken languages.
Overlap between ethnic origin and citizenship in Portugal, 2023
Note: The share of the population that self-identifies as Asian and acquired their Portuguese citizenship later in life has a high coefficient of variation, and should be interpreted with caution.
Source: OECD calculations based on National Statistics Institute Portugal (2023[19]), Survey on Living Conditions, Origins and Trajectories of the Resident Population in Portugal, https://www.ine.pt/xportal/xmain?xpid=INE&xpgid=ine_destaques&DESTAQUESdest_boui=625453580&DESTAQUESmodo=2.
Methodological issues
Discrimination is rarely captured directly in datasets and, as such, researchers need to infer its occurrence based on what would have happened to an individual if they did not have characteristics that put them at risk of discrimination. It is not sufficient to measure gaps in outcomes between groups at risk of discrimination and those not at risk. While gaps in outcomes may be indicative of discrimination or the legacy of a history of exclusion, observed gaps in income, for example, may be driven by other factors, including differences in educational attainment or health status (National Research Council, 2004[1]). Conversely, not observing gaps in outcomes between groups is not necessarily a sign of the absence of discrimination. People affected by discrimination may need to work harder to achieve the same outcomes as people not at risk of discrimination (National Research Council, 2004[1]). Just examining final outcomes will not reveal this extra effort necessary to overcome discrimination and hardship.
Economists, sociologists, health researchers and social psychologists have used various methods to identify the presence and effects of discrimination (Table 1.1). No method is perfect, but each is useful in revealing facets of discrimination, and when combined can provide a comprehensive picture of the extent of discrimination, how it affects people’s lives and how prejudicial attitudes lead to discriminatory acts. For example, experimental approaches (e.g. laboratory experiments, correspondence studies and audits) attempt to measure the effect of discrimination causally. Field studies are commonly used in European settings to examine discrimination in housing and employment applications. These approaches develop fictional people who are identical in all ways except for some characteristic that signals their risk of discrimination (e.g. their name or skin colour). These fictional people are then put forward as applicants for jobs or rental properties, and researchers examine whether there are differences in success rates based on their names or skin colour or on other characteristics. While these approaches can attribute differences in outcomes to discrimination, they can only feasibly be conducted in a few settings, such as job recruitment or rental applications (OECD, 2020[40]), and it can be difficult to scale them up for population-level analysis.
Statistical approaches are also commonly used to assess discrimination. These approaches typically use census and social survey data with demographic and economic variables to compute the gap in outcomes between the minority group of interest and the general population, where the outcomes of interest may be related to employment, income, health-relate or housing. After controlling for a range of additional factors that contribute to people’s outcomes (such as age, sex, education and location), the remaining gap could indicate discrimination, although discrimination is not directly observed and cannot be measured with certainty (OECD/European Union, 2015[41]; OECD, 2020[40]). Given the state of data collection in OECD EU countries, this type of analysis focuses on the outcomes of persons with disabilities, migrants, women and people at risk of age discrimination – although it may be possible to expand the analysis to racialised communities and LGBTI individuals, as countries continue to invest in the collection of equality data (discussed above and Chapter 5).
Alternatively, self-reported discrimination survey data can be used to examine the effects of discrimination on individuals, such as in relation to people’s employment and income outcomes, and their housing situation (Chapter 3). Some EU countries conduct national discrimination surveys. For instance, in France, the Trajectoires et Origines survey asks people about their experiences of racial discrimination and hate crimes and speech, and unfair or unequal treatment in various life domains (Institut National de la Statistique et des Études Économiques, 2019[42]; Beauchemin, Ichou and Simon, 2023[43]); Poland included questions on discrimination in the 2018 Social Cohesion Survey (Central Statistical Office Poland, 2020[44]); Luxembourg published a report on ethnic and racial discrimination in 2022 (Ministère de la Famille, de l'Intégration et à la Grande Région, 2022[45]); while Spain is set to release the results of survey on racism and racial discrimination in 2026 (Mahoney et al., 2024[46]).
Discrimination surveys have the benefit of asking people about their identities and experiences of discrimination, and some include a wide range of information on outcomes of interest such as income, housing, safety and health, along with demographic variables. This information enables the effect of discrimination to be observed, after controlling for other explanatory factors (similar to using statistical approaches, but with the addition of including self-reported discrimination as a factor).
Nevertheless, this approach is not without limitations. Self-reported discrimination data can be influenced by the framing of survey questions and the mode of delivery (Hou and Schimmele, 2022[47]). People’s perceptions of their experiences may also differ from what can be established through objective facts and which rely on people being aware of, and able to correctly identify, discriminatory action. For example, some people may not be aware they experienced discrimination (Kaiser and Major, 2006[48]). Some may even not report discrimination because they feel the need to conform to norms that they live in a meritocratic, egalitarian society, they have internalised stigma, they fear the repercussions of disclosing their experiences or they are exposed to discrimination so frequently that they perceive it as “normal” (Williams, 2016[49]). In addition, policy and legislative changes may foster environments in which people do not feel safe revealing their identities or reporting discrimination (Adams and McPhail, 2008[50]).9
Source: Adapted from Blank, Dabady and Citro (2004[1]), Measuring Racial Discrimination, National Academies Press, Washington DC; and Smith (2002[51]), Measuring Racial and Ethnic Discrimination, GSS Methodological Report No. 96, Chicago, https://gss.norc.org/Documents/reports/methodological-reports/MR096.pdf.
Conversely, some people may infer they have been discriminated against even where this is not the case (OECD/European Union, 2015[41]), which would indicate that perceptions could overstate the extent of discrimination. People may be more vigilant about labelling an event discriminatory if they have a history of experiencing discrimination, are heavily involved in advocacy, or as a way of protecting their self-worth when they experience a negative outcome like a poor job review or a termination (Kaiser and Major, 2006[48]; National Research Council, 2004[1]). In these cases, people may call an incident discrimination even if it does not constitute unequal behaviour on the basis of their characteristics.10
While there is some empirical evidence that people from at-risk groups are over-vigilant and overreport discrimination (Kaiser and Major, 2006[48]), the majority of studies find that there is either no systemic bias in detecting discrimination between at-risk groups and the general population (Major et al., 2002[52]),11 or alternatively, that there is a general tendency among people who experience discrimination to minimise what happened to them, contributing to under-reporting (Major et al., 2002[52]; Habtegiorgis and Paradies, 2013[53]; Williams, 2016[49]; Majors and Schmader, 2001[54]; Major, Quinton and McCoy, 2002[55]; Thurber et al., 2021[56]).12 For example, in Europe, some respondents to the Second European Union Minorities and Discrimination Survey were reluctant to disclose their experiences of discrimination, given the tense political climate and the refugee crisis unfolding during the survey period (European Union Agency for Fundamental Rights, 2017[57]).
Finally, cognitive testing of discrimination survey questions reveals that people provide reliable responses about their experiences when they are given clear definitions of discrimination and harassment (United Nations Praia Group on Governance Statistics, 2021[58]). Definitions help people interpret discrimination and harassment consistently, and enable them to confirm whether their experiences meet legal, normative or objective standards of discrimination and harassment. Indeed, in its guidance on conducting surveys to collect data to monitor the Sustainable Development Goals, the Office of the United Nations High Commissioner for Human Rights notes the importance of including definitional text before questions on experiences of discrimination and harassment to enhance the clarity of the terms and develop a “sufficiently common understanding … and prompt respondents’ recall of relevant incidents to contribute to the validity and reliability of the indicator” (United Nations Office of the High Commissioner for Human Rights, n.d., p. 8[59]). This highlights the importance of good survey design in mitigating some of the limitations of self‑reported discrimination data.
Given the data and methodological challenges inherent in studying discrimination in the EU, this report draws on a variety of survey data sources to explore the state and effects of discrimination and OECD EU countries’ policy and legal responses. The report primarily analyses non-official surveys and opinion polls, such as the 2019 and 2023 Discrimination in the EU Eurobarometers13 (covering all EU Member States), the Opportunities Module of the 2022 OECD Risks that Matter Survey14 of 17 OECD EU countries (accounting for 85% of the EU population), and the 2022 and 2023 AXA (2023[60]) Mind Health15 surveys of six OECD EU countries. These surveys are representative of the general population and include information on people’s self‑reported experiences of discrimination, their identities and personal characteristics and various aspects of, and outcomes in, people’s lives. With this information, this report analyses the extent, nature and drivers of discrimination across the EU (Chapter 2), and the similarities and differences in the employment, health, safety, civic and social outcomes of various at‑risk groups compared to the general population (Chapter 3). Further, the surveys used in this report enable an examination of the extent to which self‑reported experiences of discrimination are associated with differences in outcomes between the general population and people at risk of discrimination – particularly people from racialised communities, LGBTI people, people from religious minorities, persons with disabilities (Chapter 3). This type of analysis is not possible with targeted surveys, as they do not include the general population.
Since the data used are cross-sectional surveys, this report does not measure the causal effects of discrimination, but rather the outcomes that are associated with people’s perceived experiences of discrimination. There are both risks of over- and under-estimation using self‑reported experiences of discrimination, as discussed in Section 1.2, but experimental and empirical studies on the validity of self‑reported discrimination surveys, suggest, on balance, that underreporting of discrimination is more common than overreporting. As such, the estimates produced in this report are likely to be conservative; indicating the lower bounds of the extent and effects of discrimination in the EU.
Moreover, since these surveys are of the general population, they have smaller sample sizes of the groups of interest than targeted surveys. For example, the main surveys used in this report have samples of people who identify as LGBTI (or as belonging to a minority group on the basis of their sexual orientation or gender identity) that range from about 500 to 1 200.16 In comparison, FRA’s survey of LGBTI people had a sample of almost 140 000 (European Union Agency for Fundamental Rights, 2020[30]). Small sample sizes of groups at risk of discrimination in general population surveys make it difficult to conduct a granular analysis of every group (and their intersections), and often leads to results being presented at the EU-level rather than for each country. Where possible, the report draws on FRA survey results to cross-check the estimates derived from the general population surveys. In most cases, the results from the general population surveys are similar to the results presented in FRA reports (Chapters 2 and 3).
This report also draws on experimental studies conducted in Europe to augment the analysis using general population surveys, which are more likely to reflect objective incidents of discrimination (albeit in a narrower set of domains such as housing or job recruitment). Experimental studies consistently show that discrimination is present in recruitment and housing, with applications with identifiably ethnic or religious minority names receiving far fewer callbacks for interviews and invitations to view apartments compared to applications from the general population in Ireland, Belgium, France and Sweden, for example (Chapter 3).
After considering the state and effects of discrimination, this report turns to policy considerations. The report assesses the coverage of OECD EU countries’ policies and laws for fighting discrimination and promoting equality and inclusion, based on the OECD Anti-Discrimination Questionnaire, sent to OECD Members in April 2024 (and detailed in Chapter 4). The questionnaire was completed by 21 out of 22 OECD EU countries. Chapters 4 and 5 present results from this questionnaire, which collected information on the types of discrimination prohibited and the areas of life that covered by anti-discrimination laws, such as employment, education and training, health, social services, public safety and justice, online and in the media, and the provision of, and access to, products and housing. The questionnaire also asked about policies that promote equality of opportunity and social equality for people at risk of discrimination, as well as national governments’ equality mainstreaming efforts. The questionnaire focused on policies and programmes designed around the specific needs and circumstances of those at risk of discrimination.
The legal, policy and mainstreaming analysis indicates that, while some OECD EU countries adopt comprehensive approaches to support groups at risk of discrimination, many focus on specific populations – particularly people at risk of racial and ethnic discrimination and disabilities discrimination. The focus on these groups reflects the scope of EU anti‑discrimination laws, national political priorities, demographic trends, societal attitudes and levels of support for policy reform. To provide a more effective basis for supporting equality and inclusion for all, EU anti-discrimination laws should be harmonised across grounds, and countries should continue their efforts to design policies that meet the needs of at-risk groups, including by expanding mainstreaming efforts across grounds.
References
[50] Adams, C. and K. McPhail (2008), “Reporting and the politics of difference: (Non) disclosure on ethnic minorities”, Abacus, Vol. 40/3, pp. 405-435, https://doi.org/10.1111/j.1467-6281.2004.00164.x.
[63] Antman, F. and B. Duncan (2024), Ethnic Identity and Anti-immigrant Sentiment: Evidence from Proposition 187, Institute of Labor Economics, IZA DP No. 17195.
[65] AXA Group (2024), Mind Health Index, https://www-axa-com.cdn.axa-contento-118412.eu/www-axa-com/907da497-48f4-4add-acd5-bfc223efb79a_axa_mind-health_index_2024b.pdf.
[60] AXA Group (2023), Towards a New Understanding: How We Strengthen Mind Health and Wellbeing at Home, at Work and Online, https://www.axa.com/en/about-us/mind-health-report.
[8] Balestra, C. and L. Fleischer (2018), “Diversity statistics in the OECD: How do OECD countries collect data on ethnic, racial and indigenous identity?”, OECD Statistics Working Papers, No. 2018/09, OECD Publishing, Paris, https://doi.org/10.1787/89bae654-en.
[43] Beauchemin, C., M. Ichou and P. Simon (2023), “Trajectories and Origins 2 (2019-2020): A Survey on Population Diversity in France”, Population, Vol. 78/1, pp. 11-28, https://doi.org/10.3917/popu.2301.001.
[18] Central Statistical Office Poland (2021), National Population and Housing Census: National and Ethnic Affliation, Language Spoken at Home and Religion Affiliation, https://stat.gov.pl/spisy-powszechne/nsp-2021/nsp-2021-wyniki-ostateczne/tablice-z-ostatecznymi-danymi-w-zakresie-przynaleznosci-narodowo-etnicznej-jezyka-uzywanego-w-domu-oraz-przynaleznosci-do-wyznania-religijnego,10,1.html.
[44] Central Statistical Office Poland (2020), Quality of life and Social Capital in Poland: Results of the Social Cohesion Survey 2018, https://stat.gov.pl/en/topics/living-conditions/living-conditions/quality-of-life-and-social-capital-in-poland-results-of-the-social-cohesion-survey-2018,13,3.html.
[13] Central Statistics Office Ireland (2023), Census of Population 2022 - Summary Results, https://www.cso.ie/en/releasesandpublications/ep/p-cpsr/censusofpopulation2022-summaryresults/migrationanddiversity/.
[66] Cohen, G., C. Steele and L. Ross (1999), “The mentor’s dilemma: Providing critical feedback across the racial divide”, Personality and Social Psychology Bulletin, Vol. 25/10, pp. 1302-1318, https://doi.org/10.1177/0146167299258011.
[25] Croatian Bureau of Statistics (2021), Census Questionnaire, https://dzs.gov.hr/UserDocsImages/Popis%202021/PDF/Census%20Questionnaire%202021.pdf.
[12] Czech Statistical Office (2021), Population Census 2021, https://scitani.gov.cz/population.
[33] European Commission (2024), Report on the Implementation of EU Anti-Racism Action Plan 2020-2025 and on National Action Plans Against Racism and Discrimination, https://commission.europa.eu/document/download/4968fa88-5350-48d9-bf36-abd3c0142aa8_en?filename=Report%20Antiracism.pdf.
[61] European Commission (2023), Special Eurobarometer 535: Discrimination in the European Union, https://doi.org/10.2838/936462.
[62] European Commission (2019), Special Eurobarometer 493: Discrimination in the European Union, https://doi.org/10.2838/5155.
[64] European Commission Against Racism and Intolerance (2021), ECRI’s Opinion on the Concept of “Racialisation” - Adopted at ECRI’s 87th Plenary Meeting on 8 December 2021, https://rm.coe.int/ecri-opinion-on-the-concept-of-racialisation/1680a4dcc2.
[6] European Network Against Racism (2015), Equality Data Collection: Facts and Principles, https://www.enar-eu.org/wp-content/uploads/edc-general_factsheet_final.pdf.
[7] European Network Against Racism (2014), Equality Data Collection, https://www.enar-eu.org/equality-data-collection-151/ (accessed on 8 April 2025).
[5] European Network Against Racism (n.d.), Equality Data, https://www.enar-eu.org/about/equality-data/ (accessed on 20 January 2025).
[2] European Parliament (2012), Charter of Fundamental Rights of the European Union.
[31] European Union Agency for Fundamental Rights (2023), Being Black in the EU: Experiences of People of African Descent, https://fra.europa.eu/sites/default/files/fra_uploads/fra-2023-being-black_in_the_eu_en.pdf.
[32] European Union Agency for Fundamental Rights (2022), Roma in 10 European Countries: Main Results of the 2021 Roma Survey, https://fra.europa.eu/sites/default/files/fra_uploads/fra-2022-roma-survey-2021-main-results2_en.pdf.
[30] European Union Agency for Fundamental Rights (2020), A Long Way to Go for LGBTI Equality - Technical Report, https://fra.europa.eu/en/publication/2020/long-way-go-lgbti-equality-technical-report.
[57] European Union Agency for Fundamental Rights (2017), Second European Union Minorities and Discrimination Survey: Technical Report, https://fra.europa.eu/sites/default/files/fra_uploads/fra-2017-eu-midis-ii-technical-report_en.pdf.
[9] Farkas, L. (2017), Data Collection in the Field of Ethnicity: Analysis and Comparative Review of Equality Data Collection Practices in the European Union, https://commission.europa.eu/system/files/2021-09/data_collection_in_the_field_of_ethnicity.pdf.
[36] Gill, P., R. Bhopal and J. Kai (2005), “Limitations and potential of country of birth as proxy for ethnic group”, British Medical Journal, Vol. 330, p. 196, https://doi.org/10.1136/bmj.330.7484.196-a.
[53] Habtegiorgis, A. and Y. Paradies (2013), “Utilising self-report data to measure racial discrimination in the labour market”, Australian Journal or Labour Economics, Vol. 16/1, pp. 5-41, https://ajle.org/index.php/ajle_home/article/view/131.
[47] Hou, F. and C. Schimmele (2022), How Survey Mode and Survey Context Affect the Measurement of Self-Perceived Racial Discrimination across Cycles of the General Social Survey, Statistics Canada, https://www150.statcan.gc.ca/n1/pub/11-633-x/11-633-x2022006-eng.htm.
[14] Hungarian Central Statistical Office (2022), Census 2022: National Affiliation and Religion, https://nepszamlalas2022.ksh.hu/eredmenyek/vegleges-adatok/kiadvany/.
[42] Institut National de la Statistique et des Études Économiques (2019), Trajectoires et Origines 2, https://www.insee.fr/en/statistiques/7342918?sommaire=7344042.
[48] Kaiser, C. and B. Major (2006), “A social psychological perspective on perceiving and reporting discrimination”, Journal of Social Issues, Vol. 31/4, pp. 801-830.
[46] Mahoney, J. et al. (2024), “Measuring social connectedness in OECD countries - A scoping review”, OECD Papers on Well-Being and Inequalities, https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/09/measuring-social-connectedness-in-oecd-countries_02a04f4b/f758bd20-en.pdf.
[52] Major, B. et al. (2002), “Perceiving personal discrimination: The role of group status and legitimizing ideology”, Journal of Personality and Social Psychology, Vol. 82/3, pp. 269-282, https://doi.org/10.1037/0022-3514.82.3.269.
[55] Major, B., W. Quinton and S. McCoy (2002), “Antecedents and consequences of attributions to discrimination: Theoretical and empirical advances”, Advances in Experimental Social Psychology, Vol. 34, pp. 251-330, https://doi.org/10.1016/S0065-2601(02)80007-7.
[54] Majors, B. and T. Schmader (2001), “From social devaluation to self-esteem”, The Psychology of Legitimacy: Emerging Perspectives on Ideology, Justice and Intergroup Relations, Majors, J T & B Jost (eds.), Cambridge University Press, New York.
[45] Ministère de la Famille, de l’Intégration et à la Grande Région (2022), Le Racisme et les Discriminations Ethno‐Raciales au Luxembourg: Rapport d’Etude Quantitative et Qualitative, https://mfsva.gouvernement.lu/dam-assets/publications/rapport-etude-analyse/racisme/Rapport-d-etude-Enquete-Racisme.pdf.
[15] Ministry of Foreign Affairs Republic of Latvia (2024), Society Integration in Latvia, https://www.mfa.gov.lv/en/society-integration-latvia?utm_source=https%3A%2F%2Fwww.google.com%2F.
[16] Ministry of Justice Republic of Latvia (2022), Activities of Religious Organization for the Year 2021, https://www.tm.gov.lv/lv/2021-gada-publiskie-parskati?utm_source=https%3A%2F%2Fwww.google.com%2F.
[22] National Institute of Statistics Romania (2021), Population and Housing Census 2021 - Provisional Results, https://insse.ro/cms/sites/default/files/com_presa/com_pdf/cp-date-provizorii-rpl2021_0.pdf.
[23] National Institute of Statistics Romania (2013), What Does the 2011 Census Tell Us About Religion, https://insse.ro/cms/files/publicatii/pliante%20statistice/11_Pliant%20religii%20eng.pdf.
[1] National Research Council (2004), Measuring Racial Discrimination, National Academies Press, Washington, DC, https://doi.org/10.17226/10887.
[39] National Statistical Office Malta (2023), Census of Population and Housing 2021: Final Report: Population, Migration and Other Social Characteristics (Volume 1), https://nso.gov.mt/themes_publications/census-of-population-and-housing-2021-final-report-population-migration-and-other-social-characteristics-volume-1/#:~:text=For%20the%20first%20time%20ever,924%20females%20per%201%2C000%20males.
[11] National Statistical Office Malta (2021), Malta Census of Population and Housing, https://nso.gov.mt/wp-content/uploads/NSO_Census-Questionnaire_2021.pdf.
[19] National Statistics Institute Portugal (2023), Survey on Living Conditions, Origins and Trajectories of the Resident Population in Portugal, https://www.ine.pt/xportal/xmain?xpid=INE&xpgid=ine_destaques&DESTAQUESdest_boui=625453580&DESTAQUESmodo=2.
[40] OECD (2020), Over the Rainbow? The Road to LGBTI Inclusion, OECD Publishing, Paris, https://doi.org/10.1787/8d2fd1a8-en.
[29] OECD (2019), Society at a Glance 2019: OECD Social Indicators, OECD Publishing, Paris, https://doi.org/10.1787/19991290.
[41] OECD/European Union (2015), Indicators of Immigrant Integration 2015: Settling In, OECD Publishing, Paris/European Union, Brussels, https://doi.org/10.1787/9789264234024-en.
[21] Republic of Bulgaria National Statistical Institute (2021), Population and Housing Census: Ethno-cultural Characteristics of the Population, https://infostat.nsi.bg/infostat/pages/module.jsf?x_2=344.
[38] Sansone, K. (2021), “Census to collect data on race, sexual orientation and religion for the first time”, Malta Today, https://www.maltatoday.com.mt/news/national/109603/census_to_collect_data_on_race_sexual_orientation_and_religion_for_first_time.
[51] Smith, T. (2002), “Measuring racial and ethnic discrimination”, GSS Methodological Report, No. 96, National Opinion Research Center University of Chicago, https://gss.norc.org/Documents/reports/methodological-reports/MR096.pdf.
[20] Statistical Office of the Slovak Republic (2021), 2021 Population and Housing Census, https://www.scitanie.sk/en/population/basic-results/.
[24] Statistics Austria (2023), Religious Denomination, https://www.statistik.at/en/statistics/population-and-society/population/further-population-statistics/religious-denomination.
[26] Statistics Estonia (2021), Population and Housing Census 2021: Religion, https://andmed.stat.ee/en/stat/rahvaloendus__rel2021__rahvastiku-demograafilised-ja-etno-kultuurilised-naitajad__usk/RL21451.
[27] Statistics Finland (2023), Population and Society, https://stat.fi/tup/suoluk/suoluk_vaesto_en.html#Population%20by%20origin%20and%20language,%202019.
[17] Statistics Lithuania (2021), Statistical Survey on Ethnicity, Mother Language and Religion, https://osp.stat.gov.lt/paieska?p_p_id=101&p_p_lifecycle=0&p_p_state=maximized&p_p_mode=view&p_p_col_id=column-1&p_p_col_count=1&_101_struts_action=%2Fasset_publisher%2Fview_content&_101_assetEntryId=8412186&_101_type=content&_101_urlTitle=2021-gyventoju-.
[28] Statistics Netherlands (2023), Religious Involvement in the Netherlands, https://www.cbs.nl/nl-nl/longread/statistische-trends/2023/religieuze-betrokkenheid-in-nederland/2-methode.
[3] Subgroup on Equality Data of the High Level Group on Non-Discrimination, Equality and Diversity (2023), Guidance Note on the Collection and Use of Data For LGBTIQ Equality, https://commission.europa.eu/system/files/2023-07/JUST_Guidance%20note%20on%20the%20collection%20and%20use%20of%20data%20for%20LGBTIQ%20equality%20%E2%80%93%202023.pdf.pdf.
[4] Subgroup on Equality Data of the High Level Group on Non-Discrimination, Equality and Diversity (2021), Guidance Note on the Colletion and Use of Equality Data Based on Racial or Ethnic Origin, https://commission.europa.eu/system/files/2022-02/guidance_note_on_the_collection_and_use_of_equality_data_based_on_racial_or_ethnic_origin_final.pdf.
[56] Thurber, K. et al. (2021), “Developing and validating measures of self-reported everyday and healthcare discrimination for Aboriginal and Torres Strait Islander adults”, International Journal for Equity in Health, Vol. 20, pp. 1-10, https://doi.org/10.1186/s12939-020-01351-9.
[37] United Nations (2017), Principles and Recommendations for Population and Housing Censuses: Revision 3, New York, https://unstats.un.org/unsd/demographic-social/Standards-and-Methods/files/Principles_and_Recommendations/Population-and-Housing-Censuses/Series_M67rev3-E.pdf.
[34] United Nations Economic and Social Council (2024), Report of the Praia Group on Governance Statistics: Note by the Secretary-General, https://unstats.un.org/UNSDWebsite/statcom/session_55/documents/2024-7-PraiaGroup-E.pdf.
[10] United Nations Office of the High Commissioner for Human Rights (2018), A Human Rights-Based Approach to Data: Leaving No One Behind in the 2030 Agenda for Sustainable Development, https://www.ohchr.org/sites/default/files/Documents/Issues/HRIndicators/GuidanceNoteonApproachtoData.pdf.
[59] United Nations Office of the High Commissioner for Human Rights (n.d.), Guidance Note for Implementation of Survey Module on SDG Indicator 16.b.1 & 10.3.1 (Discrimination), https://www.ohchr.org/sites/default/files/Documents/Issues/HRIndicators/SDG_Indicator_16b1_10_3_1_Guidance_Note_.pdf.
[58] United Nations Praia Group on Governance Statistics (2021), Handbook on Governance Statistics, https://paris21.org/sites/default/files/2021-12/PRAIA%20Handbook%20final%20WEB-REVISED2021.pdf.
[35] Valfort, M. (forthcoming), Strengthening the Evidence Base for Racial/Ethnic Disadvantage in OECD Countries.
[49] Williams, D. (2016), “Improving the measurement of self-reported racial discrimination: Challenges and opportunities”, The Cost of Racism for People of Color: Contextualizing Experiences of Discrimination, Alvarez, A N; C T H Liang, H A Neville (eds.), American Psychological Association, https://doi.org/10.1037/14852-004.
← 1. The United Nations Human Rights Committee issued General Comment 18: Non-Discrimination (1989), which notes that discrimination “should be understood to imply any distinction, exclusion, restriction or preference which is based on any ground such as race, colour, sex, language, religion, political or other opinion, national or social origin, property, birth or other status, and which has the purpose or effect of nullifying or impairing the recognition, enjoyment or exercise by all persons, on an equal footing, of all rights and freedoms”.
← 2. In this report, the term “racialised communities” is based on the European Commission Against Racism and Intolerance’s (2021[64]) definition, in which racialisation is “the process of ascribing characteristics and attributes that are presented as innate to a group of concern to it and of constructing false social hierarchies in racial terms and associated exclusion and hostility. Regardless of where one is from and of personal circumstances, once identified or perceived as a member of a group, one is deemed as embodying characteristics based on, for instance, skin colour, ethnic or national origin inherent to all members of that group.”
← 3. The EU Charter of Fundamental Rights protects the right to liberty and security; freedom of thought and expression; protection from slavery and the freedom to choose an occupation and right to engage in work; right to education; right to property; right to asylum; rights to fair working conditions; the legal, economic and social protection of the family; the right to social security and social assistance; the right of access to health care; voting rights; right to good administration; right to an effective remedy and a fair trial; to name a few.
← 4. Article 2 of the Treaty on European Union (2012/C 326/01) states “The Union is founded on the values of respect for human dignity, freedom, democracy, equality, the rule of law and respect for human rights, including the rights of persons belonging to minorities. These values are common to the Member States in a society in which pluralism, non‑discrimination, tolerance, justice, solidarity and equality between women and men prevail”.
← 5. These include the International Covenant on Civil and Political Rights, International Covenant on Economic, Social and Cultural Rights, Convention on the Elimination of All Forms of Racial Discrimination, Convention on the Elimination of All Forms of Discrimination Against Women, and Convention on the Rights of Persons with Disabilities.
← 6. Well-being data include labour force status, educational attainment and health outcomes. See Chapter 5 for more information on the official data collections of most OECD EU countries related to groups at risk of discrimination.
← 7. As per Article 9 of the General Data Protection Regulation (Regulation (EU) 2016/679), personal data include a person’s racial or ethnic origin, religious or philosophical beliefs, or trade union membership, genetic and biometric data, health and sexual orientation.
← 8. In Malta, the 2021 Census of Population asked the following questions about individuals’ characteristics and identity. “Is the gender you identify with the same as your registered sex? (Specify)”, “Which of the following best describes your sexual orientation? (Straight/heterosexual, Gay or Lesbian, Bisexual, Other (Specify))”, “What religion, religious denomination or body that people belong to/identify with regardless of their level of practice (Roman Catholic, Orthodox, Church of England, Protestant, Islam, Judaism, Buddhism, Hinduism, Other (Specify) or No religious affiliation”, and “Do you have any of the following difficulties? (Difficulty to see even if wearing glasses, Difficulty to hear even if using a hearing aid, Difficulty to walk or going up stairs, Difficulty to remember or concentrate, Difficulty with self-care such as washing all over or dressing, Difficulty to communicate (to understand or to be understood) when using usual (customary) language”.
← 9. For example, after the passage of an anti-immigrant law in California in the 1990s, people with Mexican ancestry were less likely to identify as part of an ethnic minority (Antman and Duncan, 2024[63]).
← 10. In one experiment, African American college students who received critical feedback were more likely to say the evaluator was biased than European American students, even though all students received the same type of feedback and the evaluator was not aware of the students’ races (hence objectively bias did not exist) (Cohen, Steele and Ross, 1999[66]). However, African American students’ views of bias disappeared when evaluators’ critical feedback included an affirmation that the students were capable of meeting high standards, which suggests that perceptions of discrimination are influenced by the social context in which the threat of stigmatisation is managed (Kaiser and Major, 2006[48]; Cohen, Steele and Ross, 1999[66]).
← 11. In experiments testing the interplay of perceptions of discrimination, views on individual mobility and salience of belonging to at-risk groups, African Americans and Latin Americans were not found to generally under- or over-report discrimination relative European Americans (Major et al., 2002[52]). Similar patterns were found for women in comparison to men (Major et al., 2002[52]).
← 12. For example, Aboriginal and/or Torres Strait Islander Australians were asked if they were treated with less respect than other people, received worse service, had people act like they are not smart, were called names, were followed around stores, were watched more closely than others or had police unfairly bother them because of their Indigenous status. While many research participants revealed they had had these experiences, few labelled it discrimination because it was so common, it was considered ‘normal’ (Thurber et al., 2021[56]).
← 13. The Discrimination in the EU Eurobarometers cover people aged 15 years and over in all EU Member States. About 1 000 people are surveyed in each country (except for Cyprus, Luxembourg and Malta, where close to 500 people are surveyed). Samples are stratified by local region, sex and age (European Commission, 2023[61]; European Commission, 2019[62]). The surveys ask people about their perceptions of discrimination, attitudes towards groups at risk of discrimination, how they self-identify and well-being outcomes (such as occupation, experience of financial difficulty and household structure).
← 14. The Opportunities Module of the 2022 OECD Risks that Matter Survey, surveyed 1 000 18-64-year-olds in 27 OECD countries, including Austria, Belgium, Denmark, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Latvia, Lithuania, the Netherlands, Poland, Portugal, Slovenia and Spain. The survey asks people about their perceptions of risks and near-term and long-term concerns, preferences over government policies, experiences of discrimination, views on inequality, self-identification and well-being outcomes (e.g. employment, housing, safety and civic engagement). Sample stratification occurs by sex, age group, education level, income level, and employment status. The survey questions related to discrimination were designed in line with guidance from the United Nations Praia Group on Governance Statistics (2021[58]).
← 15. The AXA Mind Health Survey covers people aged 18 to 74 years in 16 countries. Including Belgium, France, Italy, Ireland, Italy and Spain. Two thousand people are surveyed per country in 2022 and 1 000 people per country in 2023, and data are weighted post-hoc to be representative of the general population in terms of sex, age, region and occupation. The survey asks people about their mental health, physical health, health activities, social supports, experiences of discrimination and self-identification (AXA Group, 2023[60]; AXA Group, 2024[65]).
← 16. The 2023 Discrimination in the EU Eurobarometer reveals that about 500 survey respondents (out of 26 000) identified as LGBTI. This compares to 900 people (out of 12 000) in the 2022 AXA Mind Health Survey and 500 people (out of 6 000) in the 2023 AXA Mind Health Survey who identify as lesbian, gay, bisexual, transgender, queer and/or questioning, and 1 200 people (out of 17 000) in the Opportunities Module of the 2022 OECD Risks that Matter Survey who identify as part of a sexual orientation or gender identity minority.