Content or Discontent? Perceptions of Social Protection in France, Germany and the United Kingdom

Report

12 July 2024

4. Socio-economic profiles and satisfaction with social protection

Laurenz Baertsch

Valerie Frey

Abstract

This chapter explores how perceptions of risks, social protection services and benefits, and frictions in administrative procedures vary within France, Germany and the United Kingdom depending on respondents’ socio-economic characteristics. In all three countries, worries about social and economic risks are greater among lower-income respondents and parents of dependent children. Stronger differences emerge when looking at satisfaction with social programmes: in all three countries, satisfaction with social programmes is lower among older respondents, and with few exceptions, parents are more satisfied than non-parents. In France and Germany, political partisanship is statistically associated with satisfaction with social services. Political partisanship appears to function as a proxy measure for perceptions of representativeness, inclusion and fairness in governance. When considering frictions in application processes for social benefits, perceptions of the social protection system in France roughly align with how means-tested benefits work in practice: low-income respondents and those who have received (multiple) benefits over the past year are much more positive about the eligibility for benefits, the fairness of the application process, and the time, knowledge and effort required to access benefits.

Introduction

This chapter analyses how perceptions of risks, social protection services and benefits, and frictions in administrative procedures vary within France, Germany and the United Kingdom depending on the respondents’ socio-economic characteristics. In other words, how do views of social protection vary across different population subgroups? How do “perception profiles” differ across countries? These perception profiles are also used to shed light on potential drivers of cross-country differences in perceptions, which are particularly large in the case of satisfaction with social protection services and benefits (Chapter 2).

This chapter finds similar short-term risk profiles in France, Germany and the United Kingdom, although some differences do exist. In all three countries, worries about social and economic risks are greater among lower-income respondents and parents of dependent children. In France, for example, households living in the lowest income tercile are 7 percentage points more worried about short-term risks on average than those in the highest income tercile. Similarly, French parents are 7 percentage points more likely to worry about short-term risks on average than respondents without dependent children, which stems from higher child and family-related worries such as not finding adequate childcare or education. Women are 4‑5 percentage points more likely to worry about short-term risks, on average, in Germany and the United Kingdom than men, while this relationship is weak in France.

Stronger differences emerge across socio-economic subgroups when looking at satisfaction with social protection services. In all three countries, satisfaction with social services is lower among respondents who are older: 50‑64 year‑olds, for example, are 21 percentage point less likely to be satisfied than 18‑29 year‑olds. Additionally, with few exceptions, parents are more satisfied with social services across policy areas in Germany (+13 percentage points on average across policy areas compared to respondents without children) and the United Kingdom (+9 percentage points), while French parents show higher satisfaction with social services only in some areas, such as family benefits (6 percentage points).

Importantly, in France and Germany, political partisanship is statistically associated with satisfaction with social services. Supporters of establishment parties are substantially more satisfied with social services than supporters of anti‑establishment parties (8 percentage points compared to radical left-wing and 15 percentage points relative to radical right-wing in France) or non-voters (19 percentage points in France). Combined with a higher share of radical right-wing supporters and non-voters in France (31%) than in Germany (16%), the substantially lower satisfaction with social services among these groups contributes to the high average level of dissatisfaction with social services in France relative to Germany.1

Socio-economic differences in satisfaction with income replacement benefits are similar to the results found for social services: satisfaction with social benefits decreases with the respondents’ age, while satisfaction is higher for respondents with dependent children (relative to those respondents without) and establishment voters (relative to non-establishment voters and non-voters). For both services and benefits, past benefit receipt has a positive but not consistently significant relationship with satisfaction with social protection in Germany and the United Kingdom.

Political partisanship is likely a proxy measure for perceptions of representativeness, inclusion and fairness in governance. Negative views of representativeness and fairness seem to underlie dissatisfaction with social protection among supporters of non-establishment parties and non-voters. Perceptions of representation in policy making and fairness considerations – for example, agreement with the statement “I receive a fair share in public benefits given my contributions”) – are strongly positively related with satisfaction with public services in France, Germany and the United Kingdom. In fact, once perceptions of representation in political processes and fairness considerations are taken into account, the relationship between political partisanship and satisfaction with social protection becomes substantially less important and – in the case of France – statistically insignificant in many policy areas.

When it comes to frictions in application processes for social benefits, the perception profiles differ across the three countries. In contrast to Germany, perceptions of the social protection system in France roughly align with how means-tested benefits work in practice: low-income respondents and those who have received (multiple) benefits over the past year are much more positive about the eligibility for benefits, the fairness of the application process, and the time, knowledge and effort required to access benefits. Supporters of non-establishment parties and non-voters to report higher levels of frictions in all three countries. In contrast, the time tax varies little across subgroups within France, Germany and the United Kingdom.

Perceptions of social and economic risks vary by parental status, income and gender

Short-term risk perceptions at the country-level are relatively similar in France, Germany and the United Kingdom (Chapter 2). On average across all risk categories, 60% of respondents in France, 59% in Germany and 62% in the United Kingdom worry about short-term risks (on average across policy areas). How do these worries vary by individual-level characteristics? Figure 4.1 shows the statistical associations between socio-economic subgroup and average short-term risk perceptions, while Annex Table 4.A.1 – Annex Table 4.A.3 these associations by policy area, e.g. worries about making ends meeting, accessing good-quality childcare, etc.

Subgroups in France, Germany and the United Kingdom show similar risk profiles, however, some differences do exist (Figure 4.1). In general, being a parent, living in a low-income household and being a woman is most strongly (and significantly) associated with greater average worries about social and economic risks. In all three countries, parents of dependent children are 7‑9 percentage points more likely than non-parents (of dependent children) to worry about all short-term risks on average, which stems from higher child and family-related worries such as not finding adequate childcare or education. Respondents in high-income households are 3‑10 percentage points less likely to worry about risks than those in low-income households in France and Germany, whereas this relationship is somewhat weaker in the United Kingdom. Additionally, women are 4‑5 percentage points more likely to worry about short-term risks in Germany and the United Kingdom; the result that women are more worried in France is suggestive but not statistically significant on average across policy areas. Somewhat surprisingly, only in the United Kingdom are older people (30 years and older) 7‑9 percentage points more worried about short-term risks across policy areas.

Women tend to be more worried about short term risks than men, although on average across risk categories this difference is only statistically significant in Germany and the United Kingdom, where women worry 4‑5 percentage points more. In all three countries, women worry 8‑11 percentage points more about paying all expenses. There are fewer similarities in terms of gender across the three countries in other risk areas. In France, women are also more concerned about the risk of losing employment (7 percentage points compared to men) and violence/crime (8 percentage points). In Germany, women are more concerned than men about health-related issues, namely becoming ill or disabled (11 percentage point), accessing healthcare (7 percentage points), and giving up the job to care for a family member (8 percentage points).

Although perceived short-term risks do not vary by age in France and Germany, on average across policy areas, there is age‑related variation in short-term risks at the level of policy areas. Fewer older (aged 50‑64) than young respondents (aged 18‑29) worry about finding and maintaining adequate housing in France (‑20 percentage points) and Germany (‑18 percentage points), while older respondents are 11 percentage point more worried about accessing good-quality healthcare in both countries. Additionally, older respondents in France are less worried than their younger counterparts about crime (‑16 percentage points) and giving up their job to care for a family member (‑11 percentage point). Age differences in health-related worries are more pronounced in Germany than in France, as German respondents older than 30 are also 23 percentage points more concerned about becoming ill or disabled than their younger counterparts. There is no such difference in France.

Figure 4.1. Parents, women and low-income households tend to be more worried than counterparts about short-term social and economic risks

Statistical relationship between average short-term concerns across risk categories (e.g. paying all bills, accessing good-quality healthcare) and individual-level characteristics (vertical axis) as estimated in country-specific linear regression models, by country

Perhaps unsurprisingly, higher income households are 3‑10 percentage points less likely to worry about short-term risks on average across all risk categories than low-income households, although this association is limited to high-income respondents (i.e. highest tercile in the country-level income distribution) in the United Kingdom. In all three countries, the higher average level of worries for low-income (compared to high-income) households largely stems from greater worries about paying all expenses/making ends meet: there is a roughly 20 percentage points difference across low- and high-income households. Finding and maintaining adequate housing is another significant driver of worries across income groups, with a difference of between ‑12 and ‑21 percentage point depending on the country. While other policy categories contribute to higher perceived risks among low-income households on average in Germany and the United Kingdom (e.g. becoming ill or disabled), no other risk categories are statistically significant in France.

Parents are 7‑9 percentage points more worried about short term risks on average, representing the strongest and most consistent risk heterogeneity across countries. The positive relationship between parental status and perceived risks is due to parents worrying more about child- and family-related risks, namely not finding adequate childcare or education (29‑31 percentage point) and having to give up employment to take care of a family member (12‑15 percentage points), than those without children. In both Germany and the United Kingdom, parents tend to be more worried about crime (9 percentage points), while UK parents are also more worried about employment loss (10 percentage points) and healthcare (8 percentage points) than respondents without children.

Respondents in (rural) villages in the United Kingdom tend to be more worried than those in small or big cities on average across risk categories. While this is true in some areas in Germany as well, there is no such difference in France. Somewhat surprisingly, respondents in (rural) villages are more worried about finding and maintaining adequate housing both in Germany (10 percentage points) and the United Kingdom (14 percentage points) than those in bigger cities. However, in Germany this association disappears once homeownership status, i.e. whether a respondent owns or rents his/her home, is taken into account. In the United Kingdom, there is evidence that house prices are indeed less affordable in predominantly rural than in predominantly urban areas (excluding London) (Department for Environment, Food & Rural Affairs, 2022[1]). Furthermore, British respondents in rural areas are also more worried about employment-related risks (13 percentage points for employment loss and 11 percentage point for giving up their job to take care of a family member), crime (12 percentage points) and about access to healthcare (8 percentage points). The latter could be indicative of a lower density of healthcare provision in rural areas.

There is little relationship between political affiliation, i.e. the political party’s orientation for whom a respondent would vote if elections were held tomorrow, and risk perceptions in the three countries. Only in the United Kingdom are respondents who identify with radical right-wing parties 8 percentage points more worried about short-term risks than respondents who would vote for a centrist party (averaged across all risk categories). This is driven by higher worries about accessing long-term care for the young, finding good-quality childcare and education for their children and crime. In contrast to satisfaction with social protection (see the next section Satisfaction with social protection services shows significant variation within countries below), there is no systematic difference in terms of short-term risk perceptions between respondents who would vote for establishment parties and those who would vote for non-establishment parties or non-voters.

Some characteristics have surprisingly little (or no) relationship with risk perceptions. Respondents who have a tertiary education degree are not less worried about short-term risks (averaged across all risk categories) than those who have lower educational qualifications. With a few exceptions, such as Germans with tertiary education worrying 12 percentage points less about paying all expenses, this absence of a relationship is consistent when considering the risk categories separately.

Somewhat surprisingly, employment status also shows little relation with perceived risks conditional on all other characteristics. When considering average risk perceptions across all categories, the point estimates of being unemployed and of having a stable employment contract are negative, yet statistically insignificant in all three countries. A similar picture emerges when considering risk categories separately. Unexpectedly, only in France are respondents with a stable employment contract less concerned about losing their job (‑10 percentage points) than those with temporary contracts.

The absence of a statistical association between these characteristics (i.e. education and employment status) is most likely due to the fact that many channels through which such characteristics affect risk perceptions are already taken into account by other explanatory variables in the linear regression model. For example, having a tertiary education degree likely affects risk perceptions through higher income, which is taken into account explicitly in the model.

Box 4.1. Individual-level characteristics in linear regression models

This chapter uses linear regression models to estimate the statistical associations between the outcomes of interest (e.g. social and economic worries or satisfaction with social protection) and individual-level characteristics. These statistical associations are interpreted as the (percentage point) difference in the outcome between a characteristic (e.g. being a women) and the corresponding reference group (e.g. men). Unless otherwise stated, differences are only reported if they are statistically significant at the 5% significance level.

The subgroups (or explanatory variables) and the corresponding reference group are defined as follows:

Women: women compared to men.
Age middle/young: respondents aged 30‑49/50‑64 compared to 18‑29.
Education >= tertiary: respondents with at least tertiary education compared to those with lower or no educational qualification.
Income medium/high: respondents in households in the middle/highest income decile compared to those in the lowest income decile.
Parent: individuals with a child younger than 18 and living at home compared to the rest of the sample.
Unemployed: respondents without a job compared to those employed.
Employment stable: Respondents with a stable job (i.e. employed on a permanent contract) compared to those who are employed on a temporary contract, without contract or those who are unemployed.
Benefit received: respondents who public benefit(s) in at least one policy area in the last 12 months compared to those without any public benefit over the same period.
# benefits received: number of policy areas in which a respondent received public benefits in the past 12 months. The coefficient estimates the change in the statistical relationship with the outcome variable for each additional area in which the respondent received a public benefit.
Politics: respondents who report that they would vote a for radical left-wing, radical right-wing, other party or not vote if elections were held tomorrow, compared to those who would vote for an establishment party. See Box 4.2 for details on the classification of parties.

In the figures in this chapter the dots (i.e. regression coefficients) illustrate the difference in the statistical relation between an outcome of interest and a subgroup (e.g. women) relative to outcome´s relation with the corresponding reference group (men). Thus, a coefficient on the vertical line (i.e. difference = 0) means that both the subgroup (e.g. women) and its reference category (men) have the same statistical relation with the outcome. Coefficients to the left (right) of the vertical line indicate that the outcome is lower (higher) among the subgroup than among the reference group, with ranges that illustrate the confidence interval. Differences between a subgroup and its reference group are only statistically significant if the confidence interval does not cross the vertical line.

Satisfaction with social protection services shows significant variation within countries

Cross-nationally, satisfaction with public social services varies significantly across France, Germany and the United Kingdom (reference Chapter 2). On average across all policy areas, fewer respondents in France (28%) are satisfied with social services than in Germany (37%) and the United Kingdom (38%). How these aggregate satisfaction levels vary by individual-level characteristics is shown in Figure 4.2 for the average across policy areas and in Annex Table 4.A.4 – Annex Table 4.A.6 for each policy area separately.

The within-country variation is higher for satisfaction with social services than in the case of risk perceptions, yet the satisfaction profile is similar in all three countries: satisfaction with social services is higher among younger respondents and those who would vote for centrist/establishment parties, as opposed to anti‑establishment parties or respondents who would not vote. Additionally, in Germany – but not in France or the United Kingdom – higher income and tertiary education are associated with higher levels of satisfaction with social protection services. Women are less satisfied with social services than men in France and the United Kingdom, but not in Germany.

In all three countries, the variation in satisfaction with social services is strongly negatively associated with the respondents’ age. In fact, age is among the strongest within-country heterogeneities for social service satisfaction, as older respondents (50‑64 years old) are 13‑23 percentage points and respondents aged 30‑49 are 9‑17 percentage points less satisfied than young respondents (18‑29 years old) on average across policy areas. This age gradient is particularly strong when looking at satisfaction outcomes in the areas of family support (14‑30 percentage points) and education (13‑31 percentage point). Yet, with a few exceptions, such as finding and maintaining adequate housing in Germany, it is present and statistically significant in all policy areas in the three countries.

Parents, defined as having a child younger than 18 living in the same household, are on average more satisfied with social protection services than individuals without children in Germany (+13 percentage points) and the United Kingdom (+10 percentage points). In both countries, the higher level of satisfaction among parents is particularly strong in child- and family related policy areas, such as education (15 percentage points in the United Kingdom and 13 percentage points in Germany) and family support (11 percentage point in the United Kingdom and 24 percentage points in Germany). High satisfaction with family support among parents coincides with a strong policy commitment in this area in Germany according to available policy indicators (Chapter 2). In contrast, differences in satisfaction by parental status are less pronounced in France: although the point estimates suggest higher satisfaction among parents in some areas, such as family support (6 percentage points), housing (6 percentage points) and long-term care in old age (7 percentage points), they are only marginally significant (i.e. at the 10% significance level).

Women are 5‑7 percentage points less satisfied with social services on average than men in France and the United Kingdom. This negative association stems from different policy areas in the two countries. Satisfaction is lower among French women than men in education (‑7 percentage points), public safety (‑8 percentage points) and disability (‑14 percentage points). In the United Kingdom women are particularly dissatisfied in family support (6 percentage points less satisfied than men), employment services (10 percentage points), and housing (6 percentage points). Although there is no difference between women and men in terms of satisfaction with social services on average across policy areas in Germany, German women are 9 percentage points less satisfied with disability services than their male counterparts.

There is mixed evidence for the hypothesis that public benefit receipt (in the past 12 months) affects satisfaction with public services. In Germany, respondents who received public benefits in multiple policy areas, i.e. those who rely relatively strongly on public benefits, are more satisfied with public services (4 percentage points for each additional policy area in which public benefits were received). Considering policy areas individually, respondents who received public employment benefits are 18 percentage points and those who received public disability benefits are 35 percentage points more likely to be satisfied with public services, however, other policy areas do not show such an association. In France, a similar positive relationship is found in the case of employment benefits (14 percentage points) and housing benefits (16 percentage points). However, on average across policy areas there is no statistical association between public benefit reception and satisfaction with social protection services in France or in the United Kingdom.

Figure 4.2. Older people, women, non-voters, and those voting for non-mainstream political parties are least satisfied with social services

Statistical relationship between average satisfaction with social protection services across policy areas (e.g. family support, housing,…) as dependent variable and individual-level characteristics (vertical axis) as estimated in country-specific linear regression models, by country

In France and the United Kingdom, respondents in rural areas and villages are 6–7 percentage points more satisfied with public services than those in big cities. In some policy areas, such as housing (9‑10 percentage points), this is intuitive since in rural areas and villages housing markets are typically less tight, i.e. costs are lower and housing supply is higher. In other areas, such as education and employment services in France, the positive association with satisfaction might reflect lower demand for public services (relative to their supply), less exposure to public services or differences in preferences and expectations regarding the delivery of public services compared to respondents living in urban areas.

Respondents who report being part of a minority group (defined as minority by ethnicity or skin colour, language, disability, sexual orientation or gender identification, religion or belief, migrant status, political opinion or other), are 6 percentage points more satisfied with social services than non-minority respondents in France. This positive association is equally strong for cultural minorities, defined as minorities by ethnicity, language, religion or belief and migrant status, and other minorities (i.e. the remaining minority categories). In Germany, only cultural minorities are somewhat more satisfied with social services than non-minorities (8 percentage points but statistically insignificant). Additionally, both French and German minority respondents are also more satisfied with social benefits than non-minority respondents (4 percentage points in France and 8 percentage points in Germany). The higher satisfaction with social protection among minority groups in France might come as a surprise, as minority respondents also report greater perceived short-term risks (6‑9 percentage points in all three countries). This seeming contradiction might be due to minorities having a different reference point (e.g. their home country in the case of immigrants) when it comes to reporting their satisfaction with social protection. Similarly, cultural differences in reporting satisfaction levels could also explain this difference (Senik, 2014[2]) (Chapter 3).

A second major determinant of dissatisfaction with the welfare state is partisanship and (seemingly) disillusionment with mainstream political parties, particularly in France and Germany. When asked about their vote intention if elections were held tomorrow, French and German non-establishment voters (i.e. voters of radical right- or left-wing parties) and non-voters are substantially less satisfied with public services across all policy areas than establishment voters (centre-left or centre-right parties, the reference category). This association is particularly strong for non-voters (‑19 percentage points in France and ‑23 percentage points in Germany on average across policy area) and for those who would vote for radical right-wing parties (‑15 percentage points in France and ‑16 percentage points in Germany on average across policy areas). With few exceptions, this also holds in each policy area considered separately. Public safety emerges as a major area of concern in both countries: far-right voters are 19 and 22 percentage points more worried about crime and violence in Germany and France, respectively. In Germany and France, far-left voters also tend to be less satisfied with social services (‑8 percentage points in France and ‑10 percentage points in Germany across policy areas), however, the estimates are statistically insignificant in many policy areas in France.

The share of respondents who are right-wing voters and non-voters is substantially larger in France compared to Germany (Figure 4.3), with reported vote preference in parentheses next to each party descriptor. In combination with the fact that politically alienated respondents in both countries are far less satisfied with social protection, on average across policy areas, this partially explains the substantially lower satisfaction with public services in France relative to Germany. Figure 4.3 shows that right-wing voters and non-voters are the most dissatisfied respondents in both Germany and France. However, in France these two groups account for 31% (see numbers in parentheses) of all respondents, whereas in Germany they only make up 16% of all respondents. Similarly, in both countries the most satisfied group of respondents are centrist voters, yet their fraction is 11 percentage point higher in Germany (29%) compared to France (18%). Nevertheless, for each political preference (except for non-voters) satisfaction with social services is lower in France than in Germany. Thus, the apparently stronger political alienation (i.e. the higher share of non-establishment voters) in France can only partially explain the higher aggregate dissatisfaction in France compared to Germany.

Figure 4.3 also shows that, in the United Kingdom, both the variation in satisfaction according to political preference, i.e. the difference in satisfaction between establishment and non-establishment voters and non-voters, and the political alienation, i.e. the vote share of non-establishment voters and non-voters, is small compared to France and Germany. Since establishment voters in the United Kingdom are similarly satisfied as those in Germany, these two facts partially explain the high aggregate levels of satisfaction in the United Kingdom.

The vertical axis corresponds to the question: “Please indicate the degree to which you agree or disagree with the following statement: “I think that my household and I have/would have access to good quality and affordable public services in the area of […], if needed.” Family support (e.g. childcare, parenting support services, etc.)/Education (e.g. schools, universities, professional/vocational training, adult education, etc.)/Employment (e.g. job search supports, skills training supports, self-employment supports, etc.)/Housing (e.g. social housing, housing benefit, etc.)/Health (e.g. public medical care, subsidised health insurance, mental health support, etc.)/Disability/incapacity-related needs (e.g. disability benefits and services, long-term care services for persons with disability, community living resources, etc.)/Long-term care for older people (e.g. home, community-based and/or institutional care)/Public safety (e.g. policing)/Public transportation”. Respondents could choose between: “Strongly disagree”; “Disagree”; “Neither agree nor disagree”; “Agree”; “Strongly agree”; “Can’t choose”. The share of respondents who report “strongly disagree” or “disagree” averaged across policy areas is shown. The size of each dot and the fraction reported in parentheses represents the share of RTM respondents that reported that they would vote for a party belonging to the corresponding grouping of political parties.

Source: RTM 2022.

Some characteristics show less (or a less consistent) association with satisfaction with social protection services than one might expect. The statistical relationships between social service satisfaction and tertiary education as well as with income are only statistically significant in Germany, but not in France and the United Kingdom. In Germany, the 5 percentage points higher average satisfaction with social services (across policy areas) among the tertiary educated primarily stems from higher satisfaction with education (9 percentage points) and health services (8 percentage points). Additionally, German respondents in middle- and high-income households (second and third highest income tercile) are 9–10 percentage points more satisfied with social services on average across policy areas compared to households in the lowest income tercile.

Similarly, employment stability, defined as having a permanent employment contract, is only significantly related with social service satisfaction in the United Kingdom, but not in France or Germany. In the United Kingdom satisfaction is higher among stably employed respondents (relative to those without an employment contract or those with a temporary contract) in the areas of employment services (9 percentage points), housing (8 percentage points), disability (10 percentage points) and public safety (12 percentage points). Less stringent dismissal regulations in the United Kingdom, e.g. for temporary contract holders, might contribute to differences in satisfaction among stably and unstably employed the United Kingdom respondents. No differences in satisfaction with social services between employed and unemployed respondents are found in France, Germany or the United Kingdom.

Box 4.2. Classification of political parties in RTM responses

This chapter uses an academic classification of political parties (Weisstanner, de Romémont and Bargu, 2021[3]), which groups political parties into five groups according to harmonised criteria in a large number of countries: major left and major right parties are left (or centre‑left) and right (or centre‑right) parties on the economic left-right dimensions. These parties are considered “major” because they were dominant actors in the political landscape “in terms of mobilising the class cleavage, including a realistic possibility to form governments and shape government policies.” Radical left parties are those that “reject the structure and principles of capitalism and that promote opposition to capitalist elites and institutions”. Parties that promote a nativist political platform combined with a populist discourse dividing the society into “the corrupt elite” and “the pure people” are classified as radical right parties. Parties that do not fall into any of these categories, e.g. because they cannot be clearly identified as either left or right on the economic spectrum, are classified as other parties.

To focus the analysis on differences in perceptions between respondents with moderate political views and those who oppose the current political system (radical left, radical right and) or do not vote, this chapter further combines major left and major right parties into one group, referred to as establishment parties.

The following classification results from this procedure:

Establishment (major left/right and centre):

France: La République En Marche!, Mouvement Démocrate/Modem, Les Républicains.

Germany: Sozialdemokratische Partei Deutschlands (SPD), Christlich-Demokratische Union (CDU)/Christlich-Soziale Union (CSU).

United Kingdom: Labour Party, Conservative Party.

Radical left-wing :

France: La France Insoumise, Parti Communiste Français.

Germany: Die Linke.

United Kingdom: none.

Radical right-wing :
- France: Rassemblement National, Reconquête!

Germany: Alternative für Deutschland (AfD).

United Kingdom: United Kingdom Independence Party.

Other parties are grouped in the category “Other parties”.

Note: Respondents that did not express any voting preference or declared that they would not vote are classified as “no response” (3% of respondents in Germany and the United Kingdom, and 6% in France). This group consists of respondents that cannot choose, i.e. they do not know which party to vote for, (63% of “no response” group in France, Germany and the United Kingdom on average), that they do not want to answer (31%) and that they are not eligible to vote (6%).

Source: (Weisstanner, de Romémont and Bargu, 2021[3]), “Trends in preferences over redistribution: A new harmonised dataset”.

Satisfaction with income support shows a similar perception profile as in the case of social services

Asked about the available public income support (i.e. social benefits) in case of income loss due to specific circumstances, such as having another child or unemployment, respondents on average in France (20% satisfied) report slightly lower levels of satisfaction than Germany (23%) and the United Kingdom (23%) (Chapter 2). Yet this difference is statistically significant only in few policy areas, such as income support in case of retirement, with which the French (17%) are less satisfied than respondents in Germany (22%) than those in the United Kingdom (26%).

The within-country satisfaction profile for social benefits is similar to the one observed in the case of social services (Figure 4.4; Annex Table 4.A.7 – Annex Table 4.A.9): satisfaction with social benefits is higher among younger people (relative to older), parents of dependent children (relative to respondents without children), and establishment voters (compared to non-establishment voters and non-voters). Higher education (in the case of France and Germany) and higher number of experiences with social benefits (in the case of Germany and the United Kingdom) are significantly and positively associated with satisfaction with income support and seem to play a somewhat larger explanatory role than they do in the regressions around satisfaction with services.

Satisfaction with social benefits decreases with age in all three countries. Respondents aged 30‑49 and 50‑64 are less likely to be satisfied with social benefits than those aged 18‑29 in France (a reduction of 12‑16 percentage points), Germany (7‑12 percentage points) and the United Kingdom (11‑20 percentage points). This negative association is consistent and statistically significant among all categories and tends to be particularly pronounced when respondents are asked to consider the specific scenarios of becoming ill or disabled and the death of a spouse or partner (for respondents aged 50‑64). Interestingly, while older people are less satisfied with income replacement during retirement than younger people, the size of this relationship is not larger than in the other policy categories.

Parents are 6‑12 percentage points more satisfied with public benefits than respondents without children in all three countries (Figure 4.4). This positive association is strongest in Germany, in particular for benefits in the case of having another) child (16 percentage points higher), coinciding with a particularly generous German paid parental leave scheme (Chapter 2). The positive relationship between parenthood and satisfaction with public benefits is weaker in France and only significant at the 10%-level in the United Kingdom. However, in all countries parents are more satisfied with benefits that allow them to leave their job to take care of a family member.

Tertiary educated respondents report higher levels of satisfaction with social benefits in Germany (8 percentage points), while this relationship is weaker and statistically insignificant in France (3 percentage points) and in the United Kingdom (2 percentage points). In Germany, the relationship is strongest for income replacement in the case of unemployment (10 percentage points) and having another child (11 percentage point), but statistically significant in all categories. French tertiary educated respondents are more satisfied with social benefits in the case of death of their spouse or partner (3 percentage points) and unemployment (5 percentage points; significant at 10% significance level) than respondents with lower levels of education. In the United Kingdom there do not exist any differences in satisfaction with social benefits between tertiary and non-tertiary educated respondents.

Past benefit reception is positively related with satisfaction with social benefits in Germany and the United Kingdom, however, not in France. In Germany and the United Kingdom, respondents who received social benefits in more categories (e.g. unemployment or family benefits) in the past 12 months are more satisfied with social benefits on average across categories (4‑5 percentage points per additional category). This aligns with results in the satisfaction with services question (Figure 4.2). Considering specific circumstances separately, past benefit reception is positively associated with satisfaction in the case of unemployment in both Germany (13 percentage points) and the United Kingdom (9 percentage points) and with retirement only in the United Kingdom (20 percentage points).

Figure 4.4. Age and political partisanship are also strong predictors of satisfaction with social benefits

Statistical relationship between average satisfaction with public income support across circumstances (e.g. retirement, death of a spouse,..) as dependent variable and individual-level characteristics (vertical axis) as estimated in country-specific linear regression models, by country.

Box 4.3. Voting intentions reported in RTM align with those from external sources

The validity of this report’s findings for the countries’ population as a whole hinges on the assumption that the underlying data are representative. Since political partisanship is one of the strongest determinants for satisfaction with social protection, it is important to ensure that voting intentions reported in RTM align with those found in external sources. To do so, this box compares voting intentions reported in RTM, for which respondents were surveyed between October and November 2022, with the first-round presidential election in the case of France, held in April 2022, and a nationally representative, bi-weekly survey of voting intentions in Germany, administered by Forschungsgruppe Wahlen.

Figure 4.5. Voting patterns for establishment parties and radical right- and left-wing parties are similar in RTM and external sources

Share of respondents who would vote for different political party groups if the elections were held tomorrow, as indicated in RTM (vertical axes) and in external sources (horizontal axes), by country, 2022

Voting intentions reported by French RTM respondents – aggregated according to the classification of political parties used in this section (see Box 4.2) – broadly align with the observed party voting patterns in the first round of the French presidential elections. In particular, the RTM share of votes for radical right-wing (28%), radical left-wing (15%) and establishment parties (i.e. centre-left/right, 21%) are similar to the ones in the official election results (22%, 18% and 24% for radical right, radical left and establishment parties, respectively). The proportion of RTM respondents supporting other parties (26% in RTM and 9% in the presidential election) and of non-voters (10% in RTM and 27% in the presidential election) is less aligned with official election results.

In Germany voting intentions reported in RTM align even more closely with those observed in the external source, namely the bi-weekly opinion poll from Forschungsgruppe Wahlen (FW). RTM and FW show similar proportions of respondents supporting radical right-wing (11% in RTM and 12% in FW) and radical left-wing (7% in RTM and 4% in FW), and those who would not vote (5% in RTM and 7% in FW). Somewhat fewer respondents in RTM than in FW report voting for establishment parties (29% in RTM and 38% in FW), while the opposite is true for the category “other parties” (37% in RTM and27% in FW). However, these discrepancies are unlikely to significantly affect the results in this report since both supporters of establishment parties and of “other parties” hold similar opinions about the social protection system in Germany (e.g. see Figure 4.3).

There are several potential reasons for the closer alignment between the RTM survey and external voting data in Germany compared to France. First, the opinion poll used for Germany is more similar in nature to the RTM survey, i.e. it is asking about voting intentions (including a “cannot choose” option like RTM), than the first-round presidential election. Second, the timing of the German bi-weekly opinion poll, whose survey rounds from October to November were averaged for this analysis, matches the period when RTM was fielded. In contrast, the French RTM respondents’ choice of a political party might have changed between the first-round presidential election and RTM, which was held six to seven months later. Lastly, it is not uncommon for respondents to surveys to overreport their reporting behaviour, which could partly explain the gap between actual non-voting and reported non-voting in France. Overreporting one’s voting history has long been a problem in survey research and is commonly explained by memory failure or social desirability (i.e. a respondent recalls that they did not vote, but claims to have voted to align with some perceived social good) (Belli et al., 1999[4]; McAllister and Quinlan, 2021[5]); the social desirability mechanism is likely at play in estimates of future voting as well.

Women tend to be 3‑6 percentage points less satisfied with income replacement benefits than men in all three countries. Although this relationship is not statistically significant in Germany on average across categories, women are substantially less satisfied (‑8 to ‑9 percentage points than men) with public benefits in the case of retirement in all three countries. This is perhaps unsurprising, given the persistent gender-gap in pensions and women’s higher risk of old-age poverty across countries (OECD, 2021[6]). Additionally, women in France are less satisfied with income support in case of illness or disability (‑7 percentage points) and leaving their job to care for a family member (‑4 percentage points), while there is no gender difference in satisfaction with social benefits in case of having a(nother) child.

Satisfaction with public benefits tends to be higher for respondents in higher income households in Germany and the United Kingdom. In both countries this mostly stems from respondents in households in the highest income tercile being 7‑12 percentage points more satisfied with pensions than respondents in households in the bottom income tercile. This positive association is limited to the upper income tercile in the United Kingdom, German respondents in medium-income households (i.e. in the second income tercile) are also more satisfied with income support in case of unemployment (9 percentage points). There are no differences in satisfaction with public social benefits by income in France. There is some evidence that respondents in smaller cities are more satisfied with public benefits than those in big cities and their suburbs (4‑9 percentage points). In Germany and France, respondents in small cities and villages are more likely to believe that income support in case of unemployment (10‑12 percentage points) and becoming ill or disabled (7–14 percentage points) is adequate. In the United Kingdom the same is true for social benefits in case of leaving the job to care for a family member (10 percentage points).

Non-establishment voters and non-voters are substantially less satisfied with social benefits than those who support establishment parties in France and Germany. This association is strongest for radical right-wing parties (‑16 percentage points in France and ‑12 percentage points in Germany, relative to mainstream party voters) and non-voters (‑21 percentage point in France and ‑12 percentage points in Germany, relative to mainstream party voters) and consistent across all categories. For these groups of respondents, dissatisfaction with unemployment benefits appears to be particularly pronounced in both countries (‑16 to ‑21 percentage point in Germany and ‑21 to – 33 percentage points in France, relative to mainstream voters). Radical left-wing voters are 13 percentage points less satisfied with social benefits in France, whereas this relationship is negative but not statistically significant in Germany. Political polarisation plays a smaller role for satisfaction with social benefits in the United Kingdom, as only the United Kingdom non-voters are 8 percentage points less satisfied than supporters of establishment parties (at 10% significance level).

Figure 4.6. Satisfaction with social benefits among establishment voters is higher in France than in Germany and the United Kingdom

Average satisfaction with public income support across areas (vertical axis), by vote share (marker size and number in parentheses), by country (horizontal axis).

In absolute terms, more French establishment voters (identified as “centre‑left/right” on the plot) are satisfied (32%) with social benefits than in Germany (28%) and the United Kingdom (26%) (Figure 4.6, vertical axis). This contrasts with lower satisfaction with social services among this group in France compared to Germany and the United Kingdom (see above). In line with the regression results (Figure 4.3), Figure 4.6 also shows a larger gap between establishment and non-establishment as well as non-voters in France compared to Germany and the United Kingdom. Again, given the relative size of centre‑left and centre‑right voters relative to the poles of the political spectrum, this comparatively satisfied “centre” is not large enough to shift dramatically the average satisfaction results for France.

Similar to satisfaction with social services, employment status-related variables show little relation with satisfaction with social benefits.

Perceptions of representation in policy making and fairness views as an underlying determinant of (dis)satisfaction with social protection?

The previous sections found large heterogeneities in satisfaction with social protection by political partisanship, suggesting some voters (and non-voters) feel disassociated from mainstream politics. Motivated by this fact, this section analyses the relationship between satisfaction with the social protection system and perceptions of representation and fairness in the political system and the welfare state. To do so, perceptions of representation in policy making and fairness considerations regarding the social protection system are included in the linear regression model along the previously analysed explanatory variables.

More specifically, the RTM questions used for this purpose ask respondents whether they feel that their government incorporates their views when designing or reforming public benefits and services; if they feel that they receive a fair share of public benefit, given the taxes and social contributions they pay or have paid in the past; and whether they believe that many people receive public benefits without deserving them.

French respondents hold significantly more negative perceptions about (their own) representation in policy making and about the fairness of the social security system (Annex Figure 4.B.1). Only 11% of respondents in France believe that the government incorporates their views when designing or reforming public benefits and services, compared to 18% in Germany and 20% in the United Kingdom. Additionally, the share of respondents who believe that they receive a fair share of public benefits, given the taxes and social contributions they pay is substantially lower in France (18%) than in Germany (25%) and the United Kingdom (26%). Similarly, more French respondents believe that many people receive public benefits without deserving them (66%), than respondents in Germany (56%) and the United Kingdom (58%). The differences between Germany and the United Kingdom are not statistically significant.

Perceptions of representativeness in social policy design and fairness of the social protection system are strongly related with satisfaction with public services (Figure 4.7; Annex Table 4.A.16, Annex Table 4.A.18). In all three countries average satisfaction with social services across policy areas is higher among respondents who feel represented in social protection design and reform (an increase of 26‑31 percentage point in satisfaction among respondents who feel their views are incorporated in France, Germany and the United Kingdom) and those who think that they receive their fair share in social benefits given their contribution (15‑21 percentage point) compared to those who do not hold these beliefs. These estimates are remarkably consistent across policy areas: the significant result holds for each policy area considered individually.

However, the degree to which respondents believe that many benefit recipients are undeserving is only weakly related with satisfaction with social services in the United Kingdom (3 percentage points), and there is no significant relationship in France or Germany.

These perceptions of representativeness and fairness are strongly associated with political partisanship. Indeed, once perceptions of representation in political processes and fairness are included, the relationship between political partisanship and satisfaction with social protection services and benefits becomes less important and – in some cases – insignificant. For example, whereas average satisfaction with social services is 15 percentage points lower among supporters of radical right-wing parties in France if perceptions of representation and fairness are disregarded, this relationship becomes weaker (‑6 percentage points and only statistically significant at the 10% level) once these beliefs are included in the regression model. This suggests that negative views on representation and fairness are an important source of dissatisfaction among supporters of non-establishment parties. Furthermore, the fact that general perceptions of representation and fairness are strong predictors of satisfaction with social protection helps to illustrate why it is challenging to explain cross-country variation in satisfaction with area-specific policy indicators.

Figure 4.7. Perceptions of representation in policy making and fairness are strongly associated with satisfaction with social protection

Statistical relationship between average satisfaction with social protection services across policy areas (e.g. family support, housing,…) as dependent variable and individual-level characteristics (vertical axis) as estimated in country-specific linear regression models, by country.

Three groups of respondents emerge when clustering respondents based on perceptions of social protection, risks, and socio‑economic characteristics

While the previous analysis considered each policy area (and its relationship with socio‑economic characteristics) as an independent outcome, this section analyses patterns in satisfaction with social protection and short-term risk perceptions jointly in all policy areas in a cluster analysis. This type of analysis can, for example, shed light on whether there exist groups of respondents that are satisfied with social protection in some areas but dissatisfied in others.

In the cluster analysis – specifically, this report uses Hierarchical Agglomerative Clustering – respondents are grouped into clusters according to their levels of satisfaction with social protection (each area separately), short-term risk perceptions (each area separately) and all characteristics used in the regression analysis, which are described in Box 4.1.

In all three countries, three groups of respondents (or clusters) are identified (Figure 4.7,Figure 4.8.). Although the naming of these clusters is arbitrary, this analysis refers to the three clusters according to their level of satisfaction with social protection, namely as high-, medium- and low-satisfaction clusters. Interestingly, respondents in the high-satisfaction cluster worry more about short-term risks on average across areas, than respondents in the medium- and low-satisfaction clusters. Furthermore, the ranking of these clusters in terms of the respondents’ satisfaction with social services and perceptions of short-term risks is remarkably stable across policy areas. Thus, a respondent who is satisfied (dissatisfied) with social services in one policy area, is likely also satisfied (dissatisfied) with social services in other areas.

In France, the high-satisfaction cluster (70% of respondents satisfied with social services on average across policy areas) is the smallest one (18% of all French respondents), followed by the medium-satisfaction cluster (58% satisfied; 24% of all respondents) and the low-satisfaction cluster (6% satisfied; 58% of all respondents). In Germany, respondents in all clusters are slightly more satisfied with social services (71%, 61% and 12% satisfied in the high-, medium- and low-satisfaction clusters, respectively), while the size of each cluster is comparable to the ones in France.

Interestingly, both in France and in Germany, respondents in the high-satisfaction cluster worry substantially more about all risk categories on average (78% worried in France), than respondents in the low- and medium-satisfaction clusters (62% and 44% worried, respectively, in France). In other words, there exists a group of respondents that worries strongly about short-term risks, yet still reports high levels of satisfaction with social protection. This aligns with the pattern of parents of dependent children, who report both higher perceived risks and higher satisfaction with social protection in the regression analysis (see above). In contrast, the medium- and low-satisfaction clusters show negative relationships between satisfaction with social protection and perceived risks. Put differently, the low-satisfaction group is more worried about short-term risks than the medium-satisfaction group.

There is no variation in the ranking of these clusters in terms of satisfaction with social protection or short-term risks at the policy-area level (Annex Figure 4.B.2, Annex Figure 4.B.2). Thus, a respondent who is satisfied (dissatisfied) with social protection in one policy area, is most likely also satisfied (dissatisfied) with other policy areas. This aligns with the finding that general perceptions of representativeness and fairness of the social protection system explain differences in satisfaction within countries in all policy areas (see above).

Similar to the regression analysis (see above), strong links between satisfaction with social services and political partisanship, perceptions of representativeness (in policy design and reform), and financial fairness views (regarding the social protection system) also emerge in this cluster analysis for France and Germany. For example, in France, the high-satisfaction cluster shows a higher share of establishment voters (+14 percentage points) and respondents with positive representativeness (+36 percentage points) and fairness views (+32 percentage points) than the medium- and low-satisfaction clusters.

Note: The figure above reports the characteristics of respondents in each cluster in France. For example, 70% of respondents in the high-satisfaction cluster, 18% of respondents in the medium-satisfaction cluster, and 6% of respondents in the low-satisfaction cluster are satisfied with social services on average across policy areas (first row). The characteristics age, income and city size are recorded on an integer scale from 1‑3. The corresponding rows show the average of this scale within cluster, normalised to a scale from 0‑1 (“standardised score”).

Source: RTM 2022.

Administrative procedures and time cost play a role

This chapter has so far established that satisfaction with social protection services and benefits varies significantly by age, parental status and political partisanship. This section analyses to what extent perceptions of frictions in administrative procedures and associated time costs vary by the respondents’ characteristics in France, Germany and the United Kingdom.

Perceptions of administrative frictions partially differ between Germany and France

While half (51%) of French respondents say they do not feel they could easily receive public benefits if they needed them – a rate higher than the United Kingdom (48%) and Germany (43%) – Chapter 3 showed that perceptions of frictions in the application process for social benefits are relatively similar across the three countries at most stages of the application process.

In an effort to understand why respondents viewed benefits as difficult to access, Risks that Matter 2022 asked respondents about their perceived eligibility for benefits, perceived fairness of the application process for services/benefits, knowledge required to apply, difficulty of applying for benefits/services, and time costs for various hypothetical administrative procedures. Across countries, respondents overall are similarly skeptical about how easily, fairly and quickly they could access benefits: an exception is the respondents’ perception of their own eligibility for benefits, as 42% of Germans but only 29% of the UK respondents and 24% of French believe that they would qualify for social benefits. This contrasts with a relatively high share of French respondents (49%) reporting that they know how to apply compared to Germans (47%) and respondents in the United Kingdom (43%).

When looking at subgroups’ perceptions of frictions in social protection, a few significant results emerge (Annex Table 4.A.10 – Annex Table 4.1.12).

In France, there is some evidence that means-testing is viewed accurately by the population: low-income respondents in RTM are significantly more likely than higher-income respondents to say they could easily receive benefits if they needed them, that they would qualify for benefits, and that they would know how to apply – in knowledge of how to apply, the difference is 13 percentage points Respondents who received benefits in more policy areas in the past year are also more likely than those who received benefits in fewer areas to say they would qualify and are more likely to report knowing how to apply, relative to those who received fewer benefits.

At the same time, supporters of far-right parties, other (not left) non-mainstream parties, and non-voters are much more likely to perceive frictions in accessing social protection. Radical right party supporters are 15 percentage points less likely to say they could easily receive benefits if they needed them, are 14 percentage points less likely to say the application process would be simple and quick and are a remarkable 17 percentage points less likely to say that they would be treated fairly by government officials, relative to supporters of mainstream centre‑right/centre‑left parties. Similar relationships hold for other non-mainstream party voters and non-voters.

In Germany, subgroup results do not perfectly align with those in France, and there are some counterintuitive results by income on perceived benefit eligibility. Yet for two subgroups the results for Germany are similar to France: German respondents who have received more benefits in the past are much more positive about the application process, eligibility, and fairness of the system, and German far-right voters and non-voters anticipate frictions across every measure of the question (perceived eligibility for benefits, perceived fairness of the application process for services/benefits, knowledge required to apply, difficulty of applying for benefits/services).

Time costs show little variation within countries

Respondents reported annual time tax on social protection, i.e. the time spent on administrative procedures, tends to be lower in France than in Germany, in particular in the areas of filing taxes (4.3 hours per year in France and 6.8 hours in Germany) and organising healthcare (4.9 hours annually in France and 7.3 hours in Germany). UK respondents spend a similar amount of hours on administrative procedures as Germans, with the exception that they spend less time spent on filing taxes (5.3 hours annually).

Within countries, the time tax varies little by the respondents’ characteristics. This can be observed in all areas, namely filing taxes, applying for a government benefit, Applying to and enrolling my children in school (or day care) and organising healthcare, as well as in all countries (Annex Table 4.A.13 to Annex Table 4.A.15).

References

[4] Belli, R. et al. (1999), “Reducing vote overreporting in surveys: Social desirability, memory failure and source monitoring”, Public Opinion Quarterly, Vol. 63, pp. 90-108, https://doi.org/10.1086/297704 (accessed on 1 March 2022).

[1] Department for Environment, Food & Rural Affairs (2022), Housing availability and affordability.

[5] McAllister, I. and S. Quinlan (2021), “Vote overreporting in national election surveys: a 55-nation exploratory study”, Acta Politica 2021, pp. 1-19, https://doi.org/10.1057/S41269-021-00207-6.

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

[2] Senik, C. (2014), “The French unhappiness puzzle: The cultural dimension of happiness”, Journal of Economic Behavior & Organization, Vol. 106, pp. 379-401, https://doi.org/10.1016/j.jebo.2014.05.010.

[3] Weisstanner, D., J. de Romémont and A. Bargu (2021), Trends in preferences over redistribution: A new harmonised dataset.

Annex 4.A. Tables

	Illness/ disabled	Employment loss	Find/maintain housing	Pay expenses	Childcare/ education	Elderly care	Young care	Crime	Giving up job for care	Health care	Average
Female	3.7	6.7*	1.5	8.0**	7.9	-0.8	0.1	8.2*	3.1	2.1	2.5
	(1.32)	(2.50)	(0.39)	(3.12)	(1.83)	(-0.23)	(0.01)	(2.76)	(1.51)	(0.65)	(1.33)
Age: middle	1.6	1.1	-10.1	2.2	1.1	6.8	4.5	-9.0	-5.6	12.7	0.1
	(0.21)	(0.20)	(-2.01)	(0.42)	(0.18)	(1.22)	(0.71)	(-1.28)	(-1.05)	(2.10)	(0.03)
Age: older	-0.1	-9.6	-20.1*	-1.9	-9.5	0.4	-1.2	-15.7*	-11.4*	11.7*	-2.0
	(-0.01)	(-1.58)	(-3.04)	(-0.40)	(-1.94)	(0.07)	(-0.20)	(-2.93)	(-2.52)	(2.33)	(-0.95)
Education: >= tertiary	-6.9	-1.7	1.6	-4.8	-5.4	-0.1	-6.6	1.4	-0.9	3.4	-1.5
	(-2.12)	(-0.49)	(0.56)	(-1.84)	(-1.46)	(-0.02)	(-1.82)	(0.55)	(-0.29)	(0.99)	(-1.34)
Income: medium	0.6	-6.6	-9.5	-10.9**	-3.1	-0.7	-0.9	-0.3	-3.5	0.0	-3.1
	(0.28)	(-1.73)	(-2.15)	(-3.97)	(-0.81)	(-0.16)	(-0.28)	(-0.07)	(-0.81)	(0.01)	(-2.10)
Income: high	-1.8	-2.5	-19.8*	-24.1***	2.2	3.5	-3.6	-5.4	-2.2	-0.9	-6.4***
	(-0.44)	(-0.52)	(-3.04)	(-9.96)	(0.67)	(1.08)	(-0.72)	(-1.26)	(-0.46)	(-0.25)	(-4.63)
Parent	1.2	2.7	1.0	5.1	31.1***	3.1	5.0	2.4	12.0**	-3.1	6.5**
	(0.27)	(0.84)	(0.29)	(1.50)	(8.45)	(0.73)	(1.76)	(0.70)	(3.77)	(-0.76)	(3.44)
Unemployed	2.0	-10.0	-7.2	1.8	-9.1*	-0.9	5.0	3.9	1.0	3.5	-2.1
	(0.27)	(-1.82)	(-1.17)	(0.33)	(-2.29)	(-0.22)	(1.46)	(0.82)	(0.22)	(0.75)	(-1.44)
Empl. stable	-2.3	-9.8*	-5.9	3.1	-2.5	1.9	2.3	1.8	0.1	0.0	-3.0
	(-0.40)	(-2.72)	(-1.74)	(0.70)	(-0.60)	(0.50)	(0.47)	(0.35)	(0.04)	(0.01)	(-1.46)
Small city	2.5	4.3	2.4	1.2	-0.8	6.6	8.6	5.8	1.9	3.6	1.3
	(0.70)	(1.16)	(0.65)	(0.33)	(-0.17)	(1.85)	(1.58)	(1.45)	(0.48)	(0.86)	(0.64)
(Rural) village	-2.9	-0.8	6.3	-4.7	-3.4	-6.7	-2.9	7.8	0.6	-2.9	-0.9
	(-0.92)	(-0.26)	(1.35)	(-1.28)	(-0.72)	(-1.84)	(-0.50)	(1.84)	(0.15)	(-0.72)	(-0.53)
Politics: radical left	-5.4	-8.4	-4.1	8.4*	-1.1	0.6	-7.0	-22.2**	-6.1	-3.7	-0.8
	(-0.83)	(-1.69)	(-0.78)	(2.33)	(-0.20)	(0.17)	(-1.28)	(-4.23)	(-1.37)	(-1.31)	(-0.43)
Politics: radical right	3.4	1.4	-5.9	10.2	1.8	4.2	1.2	-2.5	-10.3**	2.5	-0.5
	(0.38)	(0.26)	(-1.06)	(1.90)	(0.40)	(0.89)	(0.21)	(-0.51)	(-3.27)	(0.46)	(-0.21)
Politics: other party	-0.4	-0.2	-0.1	5.7	11.3*	5.9	0.9	-12.1*	-3.9	-1.6	2.7
	(-0.09)	(-0.05)	(-0.02)	(1.15)	(2.49)	(1.26)	(0.15)	(-2.85)	(-0.98)	(-0.37)	(1.59)
Politics: no vote	-0.3	-1.0	8.5	9.0	4.5	5.7	5.7	-2.3	-1.2	4.4	3.1
	(-0.03)	(-0.14)	(1.31)	(1.55)	(1.56)	(1.16)	(0.69)	(-0.60)	(-0.24)	(0.82)	(1.49)
Politics: other response	-1.8	2.7	-0.1	10.1	7.7	1.0	-1.0	-7.2	-7.9**	1.4	0.3
	(-0.21)	(0.55)	(-0.02)	(1.73)	(1.74)	(0.18)	(-0.12)	(-1.16)	(-3.33)	(0.30)	(0.13)
Constant	51.2***	48.2***	59.2***	66.1***	25.0**	46.0***	43.1**	52.2***	30.3***	52.3***	61.9***
	(7.79)	(7.28)	(7.85)	(9.24)	(3.29)	(6.56)	(3.59)	(6.67)	(6.40)	(7.05)	(14.69)
Observations	1019	1019	1019	1019	1019	1018	1019	1019	1019	1019	1019

Note: The table shows the coefficients of linear regression models with the individuals’ short-term concerns in different risk categories as dependent variable and the variables on the vertical axis as independent variables. The risk categories are: Becoming ill or disabled; Losing a job or self-employment income; Not being able to find/maintain adequate housing; Not being able to pay all expenses and make ends meet; Not being able to access good-quality childcare or education for your children (or young members of your family); Not being able to access good-quality long-term care for elderly family members; Not being able to access good-quality long-term care for young or working-age family members with an illness or disability; Being the victim of crime or violence; Having to give up my job to care for children, elderly relatives, or relatives with illness or disability; Accessing good-quality healthcare. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 2.5 percentage points more likely to worry about short-term risks across categories than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1 Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Illness/ disabled	Employment loss	Find/maintain housing	Pay expenses	Childcare /education	Elderly care	Young care	Crime	Giving up job for care	Health care	Average
Female	10.5*	-0.4	7.7	7.8*	7.2	3.0	1.8	5.4	7.6*	6.8*	5.0**
	(2.73)	(-0.11)	(2.01)	(2.30)	(2.09)	(0.67)	(0.47)	(1.42)	(2.38)	(2.47)	(3.48)
Age: middle	12.5*	4.2	-12.3*	-1.7	-1.2	-0.5	4.8	-9.7	4.6	14.1**	0.7
	(2.67)	(0.66)	(-2.48)	(-0.41)	(-0.28)	(-0.10)	(0.86)	(-1.74)	(0.87)	(3.47)	(0.40)
Age: older	22.4***	-2.1	-18.1***	-1.5	-13.9**	-2.6	2.9	-8.0	5.0	11.0*	0.6
	(4.19)	(-0.60)	(-4.44)	(-0.50)	(-2.96)	(-0.78)	(0.77)	(-1.54)	(1.17)	(2.64)	(0.36)
Education: >= tertiary	-4.4	0.2	-0.5	-11.5**	1.4	-1.6	-3.9	-6.7	0.9	-1.8	-1.8
	(-1.51)	(0.05)	(-0.18)	(-3.90)	(0.45)	(-0.30)	(-1.02)	(-1.71)	(0.30)	(-0.47)	(-0.98)
Income: medium	-11.9**	-6.1	-13.6*	-16.5**	-10.2*	-6.3	-7.3	-11.2*	-5.0	-9.7	-6.0**
	(-3.96)	(-1.79)	(-2.75)	(-3.23)	(-2.71)	(-1.64)	(-1.54)	(-2.64)	(-1.43)	(-2.12)	(-3.20)
Income: high	-13.7*	-11.9**	-21.3**	-21.6**	-11.7**	-7.3*	-8.0	-1.1	-0.3	-8.5	-10.1***
	(-2.72)	(-3.29)	(-3.60)	(-3.08)	(-3.17)	(-2.42)	(-1.60)	(-0.19)	(-0.12)	(-1.72)	(-4.71)
Parent	-5.1	-0.2	5.5	3.5	29.3***	0.4	9.4	8.6	13.6*	7.4	7.0**
	(-0.92)	(-0.04)	(1.40)	(0.99)	(6.12)	(0.09)	(2.05)	(2.13)	(2.92)	(1.79)	(3.06)
Unemployed	-5.3	-23.7**	-7.9	-10.5	-11.0	-3.8	-8.0	-9.2	-15.2	3.6	-6.3
	(-0.72)	(-3.19)	(-1.26)	(-1.51)	(-2.01)	(-0.47)	(-1.35)	(-1.48)	(-1.97)	(0.61)	(-1.72)
Empl. stable	-1.2	-11.2	-9.4	-1.1	-4.5	-0.7	-1.0	-0.4	-3.7	-2.4	-2.2
	(-0.17)	(-1.51)	(-1.67)	(-0.16)	(-0.82)	(-0.13)	(-0.15)	(-0.06)	(-0.51)	(-0.34)	(-0.54)
Small city	-7.5	-1.1	4.5	-3.6	1.2	-4.5	-1.0	3.0	-1.4	-6.9	-2.3
	(-1.97)	(-0.24)	(1.15)	(-0.98)	(0.33)	(-1.57)	(-0.31)	(0.86)	(-0.33)	(-1.80)	(-1.23)
(Rural) village	-6.6	0.6	10.3**	-1.8	-4.7	-9.3*	-4.7	9.9	-2.2	-6.1	-0.8
	(-1.57)	(0.15)	(3.24)	(-0.40)	(-1.15)	(-2.75)	(-1.72)	(2.13)	(-0.71)	(-1.31)	(-0.53)
Politics: radical left	-1.0	-8.5	5.7	13.0	7.1	-2.8	-12.0	-2.9	-3.1	12.6	-1.5
	(-0.13)	(-1.22)	(0.85)	(2.11)	(1.20)	(-0.36)	(-1.55)	(-0.35)	(-0.45)	(1.73)	(-0.42)
Politics: radical right	0.6	0.3	7.0	4.7	13.4*	-2.3	0.7	20.1**	-5.7	17.7*	-0.4
	(0.12)	(0.05)	(1.22)	(0.99)	(2.53)	(-0.74)	(0.14)	(3.74)	(-1.28)	(2.77)	(-0.19)
Politics: other party	-4.0	-7.3	0.5	1.0	-3.8	-18.7***	-13.6**	-9.3	-9.2	0.7	-3.4
	(-0.92)	(-1.74)	(0.10)	(0.31)	(-1.10)	(-5.64)	(-3.66)	(-2.01)	(-2.10)	(0.16)	(-1.62)
Politics: no vote	6.1	2.8	16.6*	5.7	3.1	-18.9*	-9.4	2.4	5.0	12.4	-3.9
	(0.92)	(0.33)	(2.35)	(0.95)	(0.46)	(-2.48)	(-1.36)	(0.36)	(0.64)	(1.50)	(-1.11)
Politics: other response	-5.0	-6.6	-6.8	5.4	-2.5	-11.8	-11.4	-2.0	-3.8	0.8	-6.8
	(-0.85)	(-0.93)	(-0.94)	(1.12)	(-0.51)	(-1.88)	(-1.83)	(-0.32)	(-0.65)	(0.15)	(-2.02)
Constant	56.1***	58.1***	66.6***	80.1***	39.4***	75.2***	55.4***	52.6***	39.4***	48.8***	66.4***
	(6.52)	(8.45)	(14.86)	(10.91)	(7.30)	(12.70)	(8.85)	(6.63)	(6.16)	(6.81)	(18.68)
Observations	1009	1009	1009	1008	1009	1009	1009	1009	1009	1009	1009

Note: The table shows the coefficients of linear regression models with the individuals’ short-term concerns in different risk categories as dependent variable and the variables on the vertical axis as independent variables. The risk categories are: Becoming ill or disabled; Losing a job or self-employment income; Not being able to find/maintain adequate housing; Not being able to pay all expenses and make ends meet; Not being able to access good-quality childcare or education for your children (or young members of your family); Not being able to access good-quality long-term care for elderly family members; Not being able to access good-quality long-term care for young or working-age family members with an illness or disability; Being the victim of crime or violence; Having to give up my job to care for children, elderly relatives, or relatives with illness or disability; Accessing good-quality healthcare. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 2.5 percentage points more likely to worry about short-term risks across categories than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Illness/ disabled	Employment loss	Find/maintain housing	Pay expenses	Childcare /education	Elderly care	Young care	Crime	Giving up job for care	Health care	Average
Female	3.7	0.7	2.0	10.5**	3.6	-0.5	-2.4	-1.6	1.7	1.3	4.1*
	(1.13)	(0.21)	(0.62)	(3.17)	(1.31)	(-0.13)	(-0.90)	(-0.48)	(0.45)	(0.63)	(2.49)
Age: middle	-3.3	-8.5*	-20.1***	-9.4*	-16.0***	-6.4	-10.4**	-11.2	-13.5***	-0.6	-6.7***
	(-0.96)	(-2.86)	(-5.94)	(-2.90)	(-5.21)	(-1.99)	(-3.68)	(-2.11)	(-4.53)	(-0.15)	(-4.48)
Age: older	5.6	-22.4***	-30.1***	-9.4**	-26.3***	-5.3	-18.3**	-9.2	-14.6*	3.5	-8.9***
	(2.12)	(-5.10)	(-7.43)	(-3.33)	(-6.29)	(-1.11)	(-4.30)	(-1.83)	(-3.06)	(0.75)	(-4.82)
Education: >= tertiary	3.7	1.7	-0.4	3.1	6.0	8.2*	1.6	-0.2	6.6	3.8	2.5
	(1.04)	(0.34)	(-0.10)	(1.24)	(2.00)	(2.27)	(0.58)	(-0.06)	(1.96)	(1.26)	(1.56)
Income: medium	-0.4	-0.3	0.5	-5.0	1.2	-0.4	-2.1	-2.7	-4.4	1.5	-1.2
	(-0.11)	(-0.15)	(0.15)	(-1.27)	(0.63)	(-0.11)	(-0.97)	(-0.78)	(-2.16)	(0.46)	(-0.86)
Income: high	-12.3*	-5.2	-12.7*	-20.2***	-2.5	-8.2*	-9.5*	-1.9	-11.2*	-2.5	-5.7**
	(-2.53)	(-1.11)	(-2.78)	(-5.52)	(-0.72)	(-2.23)	(-2.69)	(-0.42)	(-2.68)	(-0.80)	(-3.21)
Parent	5.3	10.1*	5.5	3.5	30.4***	9.3*	9.6*	9.3*	15.3***	7.6*	8.9***
	(1.71)	(3.07)	(1.46)	(1.03)	(12.80)	(2.66)	(3.07)	(2.51)	(4.51)	(2.40)	(5.95)
Unemployed	3.3	-19.1**	-6.8	-0.5	-5.7	2.1	-5.2	7.0	-12.4***	0.9	-2.5
	(0.56)	(-4.28)	(-1.52)	(-0.13)	(-1.70)	(0.44)	(-1.09)	(1.45)	(-4.87)	(0.17)	(-1.37)
Empl. stable	-4.6	-9.9	-4.3	0.1	-3.1	-1.5	-0.7	-0.8	-1.3	-5.6	-1.0
	(-1.42)	(-2.12)	(-0.97)	(0.02)	(-1.43)	(-0.43)	(-0.20)	(-0.19)	(-0.60)	(-1.35)	(-0.62)
Small city	-5.0	4.4	1.1	0.9	-1.1	-3.0	-4.6	0.4	0.9	5.1	1.6
	(-1.19)	(0.68)	(0.32)	(0.28)	(-0.31)	(-0.57)	(-1.05)	(0.08)	(0.22)	(1.61)	(1.03)
(Rural) village	3.3	13.0*	13.7**	5.2	5.7	7.6	4.9	12.3*	10.8*	8.1*	6.4**
	(0.57)	(2.23)	(4.41)	(1.93)	(1.70)	(1.88)	(1.51)	(2.69)	(2.42)	(2.47)	(4.38)
Politics: radical right	10.2	13.1*	20.6**	13.0	21.6**	11.6	23.2**	14.4**	17.6*	17.6*	7.6**
	(1.45)	(2.52)	(3.30)	(1.77)	(4.03)	(2.16)	(3.96)	(3.26)	(2.82)	(2.37)	(3.63)
Politics: other party	1.4	0.3	7.5*	-0.5	-1.0	-3.9	-4.3	-4.5	-1.7	2.6	-0.7
	(0.46)	(0.14)	(2.23)	(-0.10)	(-0.46)	(-1.23)	(-1.32)	(-1.09)	(-1.10)	(0.75)	(-0.46)
Politics: no vote	-5.1	-0.8	0.5	-0.7	13.1*	0.3	6.6	-7.4	9.7	-2.6	-4.3
	(-1.07)	(-0.14)	(0.12)	(-0.15)	(2.53)	(0.05)	(1.24)	(-1.63)	(1.65)	(-0.51)	(-2.08)
Politics: other response	-7.6	4.3	6.1	4.2	1.2	-0.7	4.5	-4.4	-3.1	0.2	0.5
	(-1.14)	(0.92)	(1.01)	(1.53)	(0.18)	(-0.14)	(0.78)	(-0.72)	(-0.68)	(0.03)	(0.26)
Constant	54.9***	60.7***	59.9***	72.0***	33.0***	49.3***	53.5***	52.9***	38.3***	60.9***	61.6***
	(7.49)	(6.91)	(7.95)	(13.98)	(5.00)	(6.56)	(7.03)	(5.92)	(6.96)	(9.50)	(30.35)
Observations	1037	1037	1037	1037	1037	1037	1037	1037	1037	1037	1037

Note: The table shows the coefficients of linear regression models with the individuals’ short-term concerns in different risk categories as dependent variable and the variables on the vertical axis as independent variables. The risk categories are: Becoming ill or disabled; Losing a job or self-employment income; Not being able to find/maintain adequate housing; Not being able to pay all expenses and make ends meet; Not being able to access good-quality childcare or education for your children (or young members of your family); Not being able to access good-quality long-term care for elderly family members; Not being able to access good-quality long-term care for young or working-age family members with an illness or disability; Being the victim of crime or violence; Having to give up my job to care for children, elderly relatives, or relatives with illness or disability; Accessing good-quality healthcare. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 2.5 percentage points more likely to worry about short-term risks across categories than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Family support	Education	Employment	Housing	Health	Disability	Long-term care	Public safety	Average
Female	-5.3*	-6.7*	-3.3	-6.2*	-3.7	-13.7**	-5.4	-8.3*	-6.7**
	(-2.23)	(-2.36)	(-1.28)	(-2.27)	(-1.41)	(-3.77)	(-1.40)	(-2.36)	(-3.07)
Age: middle	-15.8**	-23.3***	-13.8	-13.4**	-19.6***	-11.3*	-21.6***	-19.4***	-17.1***
	(-3.46)	(-5.01)	(-2.02)	(-3.41)	(-6.14)	(-2.66)	(-4.74)	(-8.13)	(-5.80)
Age: older	-18.9***	-31.2***	-21.7**	-18.3***	-24.2***	-14.8**	-22.9***	-18.4***	-21.2***
	(-5.09)	(-7.66)	(-3.20)	(-5.23)	(-6.07)	(-3.59)	(-4.60)	(-4.70)	(-7.82)
Education: >= tertiary	2.4	6.6	3.4	-1.4	1.6	-0.3	-0.2	2.8	1.7
	(0.69)	(1.48)	(1.27)	(-0.41)	(0.50)	(-0.08)	(-0.05)	(0.81)	(0.65)
Income: medium	-6.5	0.5	3.7	0.6	5.5	-2.8	2.3	-1.6	0.3
	(-1.71)	(0.15)	(0.86)	(0.16)	(1.76)	(-0.85)	(0.75)	(-0.40)	(0.11)
Income: high	-6.4	-3.1	-2.0	-5.3	-4.7	-9.9*	-1.6	-1.4	-4.3
	(-1.49)	(-0.78)	(-0.56)	(-1.10)	(-1.10)	(-2.34)	(-0.33)	(-0.45)	(-1.55)
Parent	6.4	3.8	0.4	5.7	-0.0	1.2	6.8	-0.4	2.4
	(2.08)	(1.09)	(0.10)	(1.89)	(-0.01)	(0.53)	(1.91)	(-0.15)	(1.04)
Unemployed	-4.3	-1.7	-1.5	-6.4	-4.7	-3.8	-5.5	-3.0	-3.8
	(-1.00)	(-0.42)	(-0.36)	(-1.42)	(-1.20)	(-0.86)	(-1.73)	(-0.62)	(-1.30)
Empl. stable	4.9	-3.5	4.0	-1.4	-0.8	-2.1	1.0	-0.2	0.2
	(1.09)	(-1.00)	(1.76)	(-0.34)	(-0.19)	(-0.58)	(0.35)	(-0.06)	(0.06)
Small city	0.8	5.1	4.6	5.3	-0.5	-1.6	2.0	-2.9	1.7
	(0.31)	(1.91)	(1.28)	(1.83)	(-0.18)	(-0.71)	(0.67)	(-0.96)	(1.06)
(Rural) village	4.3	9.5***	7.6***	8.8*	3.4	2.2	5.6	2.7	5.6**
	(1.29)	(4.35)	(5.22)	(2.38)	(1.58)	(0.72)	(2.13)	(1.32)	(3.64)
Benefit(s) received	1.7	1.5	14.3**	16.1*	1.5	6.2			-2.3
	(0.45)	(0.25)	(3.36)	(2.80)	(0.47)	(1.71)			(-0.77)
Politics: radical left	-7.4	-7.0	-6.2	-7.6	-3.2	-6.4	-11.4*	-10.1	-7.9*
	(-1.43)	(-1.25)	(-1.01)	(-1.63)	(-0.88)	(-1.58)	(-2.41)	(-2.02)	(-2.69)
Politics: radical right	-14.9**	-14.6***	-12.4	-16.7***	-17.2***	-10.5***	-14.6**	-18.5**	-15.1***
	(-3.56)	(-4.82)	(-2.13)	(-4.52)	(-9.06)	(-4.66)	(-3.44)	(-3.77)	(-5.65)
Politics: other party	-10.5*	-9.0*	-10.5*	-15.5***	-12.5**	-3.9	-14.5***	-14.2**	-11.4***
	(-2.94)	(-2.30)	(-2.36)	(-5.00)	(-3.18)	(-0.84)	(-5.12)	(-3.51)	(-4.86)
Politics: no vote	-15.2**	-16.0**	-20.3***	-20.5***	-22.1**	-14.0***	-17.6**	-23.1**	-18.8***
	(-3.42)	(-3.71)	(-5.95)	(-4.76)	(-3.50)	(-7.45)	(-4.16)	(-3.06)	(-6.44)
Politics: other response	-9.5	-3.8	-11.9	-15.2**	-9.2	-8.0*	-16.3*	-17.5*	-11.6*
	(-1.87)	(-0.64)	(-1.48)	(-3.80)	(-1.59)	(-2.54)	(-2.81)	(-2.46)	(-2.62)
# benefits received									1.7
									(1.43)
Constant	48.6***	58.9***	45.3***	48.7***	61.6***	51.9***	54.3***	63.6***	54.3***
	(6.84)	(10.21)	(5.42)	(6.96)	(12.16)	(13.13)	(8.78)	(16.78)	(12.27)
Observations	1019	1019	1019	1019	1019	1019	1019	1019	1019

Note: The table shows the coefficients of linear regression models with the individuals’ satisfaction with social protection services in different policy areas (namely Family support; Education; Employment; Housing; Health; Disability/incapacity-related needs; Long-term care for older people; Public safety) as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 6.7 percentage points less likely to be satisfied with social protection services than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Family support	Education	Employment	Housing	Health	Disability	Long-term care	Public safety	Average
Female	0.3	1.2	-2.1	-3.6	2.7	-9.1**	-3.4	-3.1	-2.1
	(0.10)	(0.36)	(-0.60)	(-1.38)	(0.85)	(-3.13)	(-1.14)	(-0.98)	(-1.43)
Age: middle	-13.9**	-13.0*	-0.8	-10.7	-8.1	-5.4	-8.2	-8.2*	-8.8*
	(-3.14)	(-2.73)	(-0.18)	(-2.03)	(-2.12)	(-1.21)	(-2.04)	(-2.27)	(-2.94)
Age: older	-20.5**	-21.1***	-10.5*	-10.1	-12.7**	-9.4*	-8.5	-8.5	-12.6**
	(-3.04)	(-4.22)	(-2.32)	(-2.09)	(-3.27)	(-2.56)	(-1.91)	(-1.61)	(-3.96)
Education: >= tertiary	4.0	8.8**	1.7	3.0	7.7**	6.1	5.8*	5.8	5.1***
	(1.56)	(4.05)	(0.83)	(1.00)	(3.07)	(1.44)	(2.65)	(1.46)	(5.36)
Income: medium	7.2*	11.7**	12.4**	5.7	7.8*	8.8*	8.0*	10.1	9.0***
	(2.59)	(3.01)	(3.58)	(1.59)	(2.15)	(2.59)	(2.87)	(2.13)	(4.18)
Income: high	7.0	18.0*	14.4**	6.9	6.6	3.8	10.2	8.1	9.8**
	(1.89)	(2.93)	(3.76)	(1.84)	(1.42)	(0.90)	(1.79)	(1.40)	(3.82)
Parent	24.1***	13.0*	10.2*	14.7***	7.8	13.2***	12.4**	12.1***	12.8***
	(5.15)	(2.88)	(2.81)	(4.96)	(1.65)	(4.65)	(3.25)	(4.82)	(6.49)
Unemployed	-1.1	-9.9	-6.9	0.2	-8.1	1.5	-2.1	-0.8	-4.0
	(-0.16)	(-1.82)	(-2.00)	(0.04)	(-1.52)	(0.34)	(-0.45)	(-0.17)	(-1.48)
Empl. stable	-1.2	-7.3	-0.0	-0.1	-6.7	-5.8	-5.5	-1.7	-3.0
	(-0.28)	(-1.75)	(-0.01)	(-0.03)	(-1.89)	(-1.52)	(-1.01)	(-0.48)	(-1.44)
Small city	5.0	3.5	6.0	4.9	0.7	-3.0	1.7	4.8	2.8
	(1.03)	(0.67)	(1.38)	(1.37)	(0.12)	(-0.81)	(0.42)	(1.09)	(0.80)
(Rural) village	6.2	2.4	5.7	3.5	-0.2	2.3	3.4	5.0	3.4
	(1.65)	(0.63)	(1.37)	(1.31)	(-0.05)	(0.72)	(1.04)	(1.33)	(1.27)
Benefit(s) received	5.0	6.9	18.1**	14.3	1.8	34.7***			-4.0
	(0.89)	(1.09)	(3.59)	(1.30)	(0.43)	(5.99)			(-1.50)
Politics: radical left	-16.1*	-5.6	-11.6	-5.8	-12.9	-11.7	-5.3	-9.2	-10.1*
	(-2.43)	(-0.77)	(-1.66)	(-0.98)	(-1.83)	(-1.80)	(-0.87)	(-1.24)	(-2.33)
Politics: radical right	-11.9	-15.1	-17.6**	-17.3**	-17.5***	-15.3*	-14.0*	-22.4**	-16.2**
	(-1.83)	(-1.83)	(-3.81)	(-3.26)	(-4.09)	(-2.61)	(-2.25)	(-3.65)	(-3.69)
Politics: other party	-4.1	0.9	-7.6*	-3.4	-8.1	-4.3	-4.7	-2.4	-4.4
	(-1.27)	(0.22)	(-2.28)	(-0.82)	(-2.00)	(-1.10)	(-0.89)	(-0.46)	(-1.32)
Politics: no vote	-18.3**	-31.7***	-22.9***	-17.3**	-26.1***	-21.8**	-12.6	-32.7***	-22.9***
	(-3.31)	(-6.80)	(-4.50)	(-3.47)	(-4.19)	(-3.86)	(-1.79)	(-5.17)	(-7.05)
Politics: other response	-8.2	-11.2**	-16.6**	-5.2	-18.3**	-10.8*	-10.8*	-8.3**	-10.6***
	(-1.97)	(-3.00)	(-3.13)	(-1.14)	(-3.38)	(-2.49)	(-2.17)	(-3.74)	(-5.01)
# benefits received									4.2**
									(3.86)
Constant	33.8***	52.0***	34.5***	30.1**	58.7***	37.3***	31.8**	51.0***	40.3***
	(4.32)	(9.67)	(4.54)	(3.98)	(7.78)	(4.54)	(3.71)	(6.63)	(8.04)
Observations	1009	1009	1009	1009	1009	1009	1009	1009	1009

Note: The table shows the coefficients of linear regression models with the individuals’ satisfaction with social protection services in different policy areas (namely Family support; Education; Employment; Housing; Health; Disability/incapacity-related needs; Long-term care for older people; Public safety) as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 6.3 percentage points less likely to be satisfied with social protection services than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Family support	Education	Employment	Housing	Health	Disability	Long-term care	Public safety	Average
Female	-5.9*	-0.9	-10.1**	-6.2*	-7.3	-4.4	-5.6	-2.2	-5.2*
	(-2.71)	(-0.35)	(-3.20)	(-2.34)	(-2.15)	(-1.27)	(-1.74)	(-0.61)	(-2.51)
Age: middle	-14.7**	-16.8*	-11.4**	-10.3	-10.5**	-7.8*	-9.6*	-8.7*	-11.4**
	(-3.43)	(-2.94)	(-3.55)	(-1.90)	(-3.68)	(-2.46)	(-2.43)	(-2.43)	(-3.44)
Age: older	-30.0***	-26.8***	-22.5***	-23.4**	-24.0***	-12.4*	-21.9***	-20.4***	-22.7***
	(-5.38)	(-5.16)	(-6.58)	(-3.90)	(-7.16)	(-2.48)	(-4.49)	(-5.05)	(-6.61)
Education: >= tertiary	-0.9	5.9	4.8	5.4	3.3	3.6	1.3	2.6	3.3
	(-0.26)	(1.88)	(1.27)	(1.42)	(1.13)	(1.02)	(0.41)	(0.69)	(1.16)
Income: medium	-2.7	-0.5	-1.8	-1.7	-2.4	-6.0*	-2.4	2.6	-1.8
	(-0.77)	(-0.11)	(-0.61)	(-0.54)	(-0.66)	(-2.22)	(-0.71)	(0.62)	(-0.67)
Income: high	7.5	4.2	8.4	5.8	5.2	4.0	8.9	11.7*	7.1
	(1.43)	(1.23)	(1.88)	(1.35)	(1.10)	(0.95)	(1.86)	(2.40)	(1.89)
Parent	11.2**	15.0**	10.4**	7.0*	5.5	8.8*	12.0**	6.0	9.5**
	(3.54)	(3.13)	(4.00)	(2.50)	(1.40)	(2.48)	(3.82)	(1.56)	(4.00)
Unemployed	0.3	5.0	0.8	0.5	0.3	2.0	-1.1	10.4	1.9
	(0.06)	(0.75)	(0.17)	(0.13)	(0.07)	(0.59)	(-0.20)	(1.83)	(0.51)
Empl. stable	7.4	9.2	9.3*	7.8*	8.3	9.8*	3.5	12.4*	8.6*
	(2.02)	(1.98)	(2.29)	(2.77)	(1.77)	(2.84)	(0.95)	(2.39)	(2.49)
Small city	0.7	4.4	3.3	6.4	2.4	4.0	5.1	3.2	3.9
	(0.10)	(1.40)	(0.64)	(1.35)	(0.47)	(1.62)	(1.27)	(0.66)	(1.12)
(Rural) village	3.8	3.1	4.6	10.3*	7.0	7.9*	10.3*	6.0	6.8*
	(0.78)	(0.91)	(0.99)	(2.70)	(1.43)	(2.55)	(2.67)	(1.35)	(2.23)
Benefit(s) received	9.1	6.7	1.8	33.3***	2.1	2.2			-2.5
	(1.54)	(0.94)	(0.23)	(6.86)	(0.56)	(0.45)			(-0.96)
Politics: radical right	-5.8	-12.7*	-10.5	-4.5	-5.5	-8.8	-8.6	-14.2*	-9.0
	(-0.76)	(-2.63)	(-1.47)	(-0.69)	(-0.99)	(-1.39)	(-1.32)	(-2.65)	(-1.75)
Politics: other party	-12.2***	-7.7	-7.0**	-4.4	-5.3	-0.8	-2.0	-5.4*	-5.6**
	(-6.23)	(-1.59)	(-3.42)	(-1.67)	(-1.51)	(-0.40)	(-1.05)	(-2.22)	(-3.11)
Politics: no vote	-11.7*	-17.7*	-11.6	-12.2*	-17.6**	-9.5	-8.7	-22.3**	-13.8**
	(-2.23)	(-2.37)	(-1.88)	(-2.76)	(-3.76)	(-2.05)	(-1.95)	(-3.98)	(-3.53)
Politics: other response	-15.1**	-16.1***	-10.3**	-15.5**	-2.4	-10.3**	-7.9*	-11.2*	-10.9***
	(-3.59)	(-4.47)	(-3.31)	(-3.78)	(-0.51)	(-3.11)	(-3.07)	(-2.65)	(-4.84)
# benefits received									2.4
									(1.90)
Constant	46.6***	51.6***	44.8***	35.4***	49.0***	27.3***	32.4***	39.6***	40.3***
	(4.50)	(7.56)	(5.54)	(6.96)	(6.14)	(4.77)	(5.91)	(6.39)	(7.01)
Observations	1037	1037	1037	1037	1037	1037	1037	1037	1037

Note: The table shows the coefficients of linear regression models with the individuals’ satisfaction with social protection services in different policy areas (namely Family support; Education; Employment; Housing; Health; Disability/incapacity-related needs; Long-term care for older people; Public safety) as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 6.3 percentage points less likely to be satisfied with social protection services than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Unemployment	Illness/Disability	Having an(other) child	Leave job for family member care	Retirement	Death of spouse or partner	Average
Female	1.8	-7.1**	-0.2	-4.0*	-8.8**	-4.3	-3.9*
	(0.88)	(-3.39)	(-0.08)	(-2.22)	(-3.33)	(-1.95)	(-2.51)
Age: middle	-13.1*	-15.9**	-13.0***	-8.3**	-10.8**	-12.4**	-12.1***
	(-2.54)	(-3.15)	(-4.91)	(-3.20)	(-3.30)	(-3.21)	(-4.90)
Age: older	-17.5**	-20.4**	-19.6***	-7.8**	-14.7***	-14.9**	-15.8***
	(-3.40)	(-4.08)	(-6.93)	(-4.09)	(-4.66)	(-3.60)	(-6.54)
Education: >= tertiary	4.3	3.8	1.9	3.5	1.8	3.2*	2.9
	(2.07)	(1.47)	(0.80)	(1.61)	(0.55)	(2.66)	(2.05)
Income: medium	3.5	2.5	2.7	1.5	1.6	2.5	2.6
	(0.91)	(0.56)	(0.82)	(0.49)	(0.71)	(0.85)	(1.09)
Income: high	2.6	-3.0	-4.4	-1.0	-2.6	-3.9	-1.9
	(0.56)	(-0.66)	(-1.07)	(-0.30)	(-0.84)	(-1.37)	(-0.66)
Parent	8.9*	2.9	4.7	6.3*	7.2**	6.4*	5.8*
	(2.98)	(1.30)	(1.36)	(2.78)	(3.16)	(2.25)	(2.76)
Unemployed	-8.8	-2.3	-9.9*	-4.8	-4.2	-8.8	-6.5
	(-1.43)	(-0.55)	(-2.20)	(-1.00)	(-1.31)	(-1.95)	(-1.86)
Empl. stable	-1.3	1.0	-1.8	-0.5	-1.4	-0.9	-0.7
	(-0.26)	(0.32)	(-0.48)	(-0.10)	(-0.35)	(-0.22)	(-0.21)
Small city	5.4	7.8**	4.5	-0.0	5.2*	2.7	4.4*
	(1.55)	(3.43)	(1.77)	(-0.00)	(2.69)	(1.18)	(2.61)
(Rural) village	12.4**	7.0*	1.1	-0.6	3.7	2.3	4.3
	(3.15)	(3.05)	(0.45)	(-0.17)	(1.06)	(0.94)	(1.83)
Benefit(s) received	5.2	-3.9	1.1		5.2		0.8
	(1.19)	(-1.20)	(0.35)		(1.48)		(0.25)
Politics: radical left	-19.5**	-12.2	-13.4**	-2.3	-16.2**	-14.6**	-13.3**
	(-3.48)	(-2.15)	(-3.83)	(-0.62)	(-3.56)	(-3.53)	(-4.22)
Politics: radical right	-20.9***	-11.5*	-17.4**	-11.9**	-17.3***	-19.6***	-16.4***
	(-6.35)	(-2.48)	(-3.20)	(-4.06)	(-5.89)	(-6.64)	(-6.35)
Politics: other party	-15.8**	-3.7	-12.0*	-7.0	-13.0**	-10.5*	-10.3**
	(-4.23)	(-1.07)	(-2.61)	(-2.06)	(-4.06)	(-2.60)	(-4.04)
Politics: no vote	-33.4***	-16.6***	-22.5*	-12.1**	-21.3***	-21.0**	-21.0***
	(-6.83)	(-4.79)	(-3.01)	(-3.65)	(-5.70)	(-3.92)	(-5.89)
Politics: other response	-21.9***	-9.6	-20.8***	-14.3***	-17.9***	-16.1**	-16.7***
	(-5.60)	(-2.00)	(-4.69)	(-4.37)	(-6.22)	(-3.84)	(-6.69)
# benefits received							0.3
							(0.33)
Constant	43.7***	40.4***	48.3***	29.5**	39.5***	40.8***	39.4***
	(4.65)	(4.92)	(9.64)	(4.12)	(5.69)	(5.96)	(6.25)
Observations	1019	1019	1019	1019	1019	1019	1019

Note: The table shows the coefficients of linear regression models with the individuals’ satisfaction with public income support in different areas (namely Unemployment; Illness/disability; Having a child/having more children; Leaving work to care for elderly family members or family members with disabilities; Retirement; Death of spouse or partner) as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 3.9 percentage points less likely to be satisfied with public income support than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Unemployment	Illness/Disability	Having an(other) child	Leave job for family member care	Retirement	Death of spouse or partner	Average
Female	-1.6	0.5	-1.5	-7.9**	-8.0**	-1.6	-3.4
	(-0.55)	(0.17)	(-0.43)	(-2.96)	(-3.47)	(-0.53)	(-1.80)
Age: middle	-3.7	-11.4**	-3.5	-10.8	-4.6	-8.8	-7.4*
	(-0.80)	(-3.07)	(-0.76)	(-1.90)	(-1.99)	(-1.61)	(-2.20)
Age: older	-9.9*	-12.3*	-12.8*	-12.8*	-8.2	-16.3*	-11.8*
	(-2.15)	(-2.64)	(-2.19)	(-2.16)	(-1.99)	(-2.71)	(-2.70)
Education: >= tertiary	10.2**	7.4*	10.7***	8.6**	7.6*	7.4*	8.4***
	(2.99)	(2.76)	(4.92)	(3.71)	(2.67)	(2.67)	(5.01)
Income: medium	8.2*	5.6	5.2	0.1	6.6	3.3	4.8
	(2.26)	(1.89)	(1.99)	(0.04)	(2.06)	(0.95)	(2.13)
Income: high	4.5	5.0	3.3	1.8	7.4*	1.2	4.4*
	(0.95)	(1.97)	(1.27)	(0.59)	(2.86)	(0.40)	(2.26)
Parent	9.7*	15.9***	16.2**	12.2**	11.6*	11.7**	12.4**
	(2.68)	(4.77)	(3.67)	(3.41)	(2.92)	(3.68)	(3.91)
Unemployed	5.6	4.0	-1.3	5.2	2.2	-5.5	1.4
	(0.98)	(0.75)	(-0.23)	(0.96)	(0.53)	(-1.21)	(0.37)
Empl. stable	-4.1	-5.1	-10.5*	-1.9	-7.6*	-7.2	-5.8
	(-1.27)	(-1.46)	(-2.57)	(-0.42)	(-2.21)	(-1.42)	(-1.96)
Small city	2.2	6.6	1.6	5.9	8.4	6.4	5.1
	(0.60)	(1.57)	(0.35)	(1.59)	(1.67)	(1.40)	(1.34)
(Rural) village	9.9*	13.7**	7.5	6.8**	9.7	5.7	8.8*
	(2.81)	(2.95)	(1.91)	(3.89)	(1.84)	(1.30)	(2.79)
Benefit(s) received	13.4*	3.5	4.7		-1.8		-6.5*
	(2.58)	(0.84)	(0.68)		(-0.49)		(-2.30)
Politics: radical left	-7.2	-11.3	-7.9	-10.9	-6.6	-8.4	-8.9
	(-1.13)	(-1.66)	(-1.72)	(-2.06)	(-0.94)	(-0.99)	(-1.60)
Politics: radical right	-16.4***	-13.3*	-8.2*	-12.6**	-10.6**	-10.7*	-11.6***
	(-5.03)	(-2.77)	(-2.94)	(-3.33)	(-3.21)	(-2.78)	(-4.48)
Politics: other party	-2.0	-6.5	-4.8	-6.4	-1.7	-5.6	-4.6
	(-0.34)	(-1.41)	(-1.01)	(-1.59)	(-0.40)	(-1.05)	(-1.13)
Politics: no vote	-20.7***	-8.0	-8.3	-16.7***	-7.8	-9.1	-11.7*
	(-5.71)	(-1.15)	(-1.24)	(-4.53)	(-1.64)	(-1.75)	(-2.81)
Politics: other response	-5.9	-6.0	-9.3	-8.1	-1.8	-13.2	-6.8
	(-0.97)	(-1.53)	(-1.63)	(-1.99)	(-0.46)	(-1.85)	(-1.81)
# benefits received							4.6**
							(3.79)
Constant	23.5***	21.5**	28.5**	26.9**	20.3**	30.3**	24.8***
	(4.41)	(3.48)	(3.33)	(3.92)	(3.23)	(3.50)	(4.25)

Note: The table shows the coefficients of linear regression models with the individuals’ satisfaction with public income support in different areas (namely Unemployment; Illness/disability; Having a child/having more children; Leaving work to care for elderly family members or family members with disabilities; Retirement; Death of spouse or partner) as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 3.7 percentage points less likely to be satisfied with public income support than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Unemployment	Illness/Disability	Having an(other) child	Leave job for family member care	Retirement	Death of spouse or partner	Average
Female	-5.1*	-5.1	-5.9	-5.8	-8.7**	-6.4	-6.0*
	(-2.67)	(-1.72)	(-1.95)	(-1.93)	(-3.44)	(-2.05)	(-2.83)
Age: middle	-13.0*	-13.4**	-11.0*	-6.0	-11.9*	-11.4**	-11.3**
	(-2.87)	(-3.28)	(-2.24)	(-2.03)	(-2.38)	(-3.49)	(-3.22)
Age: older	-20.4***	-24.5***	-22.9***	-15.1***	-17.0**	-22.6***	-20.4***
	(-4.72)	(-6.07)	(-5.35)	(-5.97)	(-3.40)	(-6.21)	(-6.10)
Education: >= tertiary	3.1	0.9	2.0	2.7	1.1	2.1	2.0
	(1.19)	(0.25)	(0.64)	(1.06)	(0.33)	(0.63)	(0.74)
Income: medium	-4.1	-2.7	-4.1	-3.8	0.6	-2.0	-2.3
	(-1.27)	(-0.75)	(-1.26)	(-1.02)	(0.19)	(-0.44)	(-0.75)
Income: high	4.1	3.6	2.5	2.7	11.9**	4.3	5.5*
	(1.34)	(0.99)	(0.99)	(0.96)	(3.24)	(1.13)	(2.43)
Parent	8.3	7.9	5.2	8.8*	10.3	9.2	7.8
	(2.03)	(1.59)	(1.76)	(2.40)	(1.83)	(1.84)	(2.11)
Unemployed	-1.7	-0.3	2.3	-2.2	1.6	6.1	-0.3
	(-0.38)	(-0.06)	(0.59)	(-0.54)	(0.32)	(1.46)	(-0.08)
Empl. stable	1.9	0.6	5.4	7.0	4.0	5.1	4.7
	(0.61)	(0.12)	(1.84)	(1.96)	(0.95)	(1.32)	(1.34)
Small city	1.0	-0.4	0.2	1.0	4.1	0.2	1.2
	(0.26)	(-0.10)	(0.05)	(0.32)	(0.74)	(0.05)	(0.35)
(Rural) village	5.0	5.8	6.3	10.1*	8.2	5.8	6.7*
	(1.34)	(1.86)	(2.05)	(3.03)	(1.91)	(1.59)	(2.42)
Benefit(s) received	8.9	4.6	9.4		20.3*		-0.3
	(2.11)	(0.83)	(1.92)		(3.10)		(-0.07)
Politics: radical right	0.4	-3.4	8.2	10.6	-2.3	0.6	1.9
	(0.09)	(-0.62)	(1.09)	(1.11)	(-0.31)	(0.12)	(0.42)
Politics: other party	-4.8	-1.5	-2.0	0.2	-3.8	-2.9	-2.4
	(-2.16)	(-0.43)	(-0.62)	(0.05)	(-1.45)	(-1.04)	(-0.98)
Politics: no vote	-2.6	-10.2	-10.5*	-9.2	-10.1*	-8.1	-8.1
	(-0.60)	(-1.78)	(-2.33)	(-2.06)	(-2.44)	(-1.59)	(-2.09)
Politics: other response	-14.2**	-14.5***	-8.6*	-7.8	-7.2	-11.9**	-9.9**
	(-4.27)	(-4.66)	(-2.29)	(-1.82)	(-1.88)	(-3.15)	(-3.17)
# benefits received							3.6*
							(3.06)
Constant	34.1**	40.1***	30.1**	19.2**	26.8***	29.1***	27.4***
	(4.42)	(6.01)	(4.31)	(4.40)	(5.44)	(4.58)	(4.71)
Observations	1037	1037	1037	1037	1037	1037	1037

Note: The table shows the coefficients of linear regression models with the individuals’ average satisfaction with public income support in different areas (namely Unemployment; Illness/disability; Having a child/having more children; Leaving work to care for elderly family members or family members with disabilities; Retirement; Death of spouse or partner) as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 3.7 percentage points less likely to be satisfied with public income support than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Could easily receive if needed	Would qualify	Know how to apply	Simple and quick application process	Fair treatment by government office	Average
Female	-4.0	-1.8	9.3	-5.0	-7.4	-0.5
	(-0.92)	(-0.52)	(1.57)	(-1.75)	(-1.99)	(-0.15)
Age: middle	-6.0	-7.0	3.8	-7.8*	-7.6	-5.4*
	(-1.89)	(-2.11)	(0.94)	(-2.21)	(-1.75)	(-2.59)
Age: older	-9.0**	-10.9*	6.8	-4.4	-1.5	-3.5
	(-3.33)	(-3.04)	(1.48)	(-1.48)	(-0.29)	(-1.59)
Education: >= tertiary	5.5	0.5	0.2	0.5	9.6*	3.9
	(1.38)	(0.12)	(0.05)	(0.15)	(2.69)	(1.26)
Income: medium	-4.2	-8.4	-3.8	-3.0	2.8	-4.8
	(-1.59)	(-1.83)	(-1.02)	(-0.91)	(0.67)	(-1.89)
Income: high	-10.6**	-15.2***	-13.4*	-8.4	-5.6	-12.4***
	(-4.00)	(-4.81)	(-2.94)	(-1.87)	(-1.45)	(-5.37)
Parent	-1.4	-1.0	0.4	1.7	3.4	3.1
	(-0.37)	(-0.30)	(0.11)	(0.60)	(0.79)	(1.07)
Unemployed	-6.3	0.1	-3.6	-3.5	-1.4	-1.1
	(-1.28)	(0.03)	(-0.77)	(-1.03)	(-0.33)	(-0.41)
Empl. stable	2.2	-1.0	1.6	-0.1	2.2	0.2
	(0.51)	(-0.32)	(0.37)	(-0.04)	(0.75)	(0.07)
Small city	3.4	-0.4	11.0**	4.6	2.3	3.4
	(0.75)	(-0.09)	(4.03)	(1.99)	(0.74)	(1.41)
(Rural) village	6.9	-0.8	12.2*	3.3	8.3*	5.9*
	(1.97)	(-0.17)	(2.98)	(1.06)	(2.61)	(2.19)
Benefit received	0.6	-0.9	6.0	1.0	1.3
	(0.14)	(-0.26)	(1.21)	(0.21)	(0.18)
# benefits received	3.1	4.0*	5.2*	2.0	5.1*
	(2.06)	(2.73)	(2.46)	(1.12)	(2.50)
Politics: radical left	-7.8	-5.4	-4.0	-7.0	-3.1	-3.7
	(-1.44)	(-1.00)	(-0.72)	(-1.08)	(-0.48)	(-0.86)
Politics: radical right	-14.8**	-4.9	-9.1	-14.2**	-17.4**	-12.1**
	(-3.55)	(-1.24)	(-1.67)	(-3.95)	(-4.20)	(-4.29)
Politics: other party	-15.1**	-3.1	-1.9	-10.9**	-10.9*	-9.0*
	(-4.22)	(-0.68)	(-0.31)	(-3.14)	(-2.95)	(-2.83)
Politics: no vote	-21.0***	-14.1*	-6.2	-18.8**	-21.1**	-16.8***
	(-6.05)	(-2.19)	(-1.30)	(-3.21)	(-3.86)	(-4.60)
Politics: other response	-9.7*	-11.8*	-7.9	-11.3**	-14.9**	-11.6**
	(-2.29)	(-2.28)	(-1.55)	(-3.99)	(-3.32)	(-4.08)
Constant	38.4***	41.6***	31.7**	36.9***	37.0***	43.0***
	(5.30)	(4.73)	(4.20)	(8.46)	(7.32)	(8.57)
Observations	1019	1019	1019	1019	1019	1019

Note: The table shows the coefficients of linear regression models with the individuals’ perceived frictions at different stages of the benefit application process as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (e.g. “would qualify for public benefits”) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Could easily receive if needed	Would qualify	Know how to apply	Simple and quick application process	Fair treatment by government office	Average
Female	-7.3**	-6.0*	-2.3	-5.1	-4.4	-4.9**
	(-2.98)	(-2.27)	(-0.88)	(-2.07)	(-1.70)	(-3.58)
Age: middle	2.6	2.0	11.2*	-1.3	0.4	3.6
	(0.71)	(0.47)	(2.87)	(-0.54)	(0.06)	(1.51)
Age: older	-6.1	-6.6	14.1**	-3.5	-8.6*	-2.0
	(-1.31)	(-1.57)	(3.18)	(-1.57)	(-2.26)	(-1.11)
Education: >= tertiary	9.7**	11.9**	-0.8	3.7	11.8**	7.6*
	(3.03)	(3.61)	(-0.21)	(1.12)	(3.89)	(2.93)
Income: medium	4.0	7.0*	5.1	5.9*	9.3*	6.0*
	(1.30)	(2.86)	(1.83)	(2.50)	(2.42)	(2.52)
Income: high	5.1	-0.6	8.9*	5.8	7.5*	4.4
	(1.68)	(-0.17)	(2.60)	(1.78)	(2.59)	(2.12)
Parent	8.6**	3.9	8.0	11.6*	1.0	7.9**
	(2.96)	(0.94)	(1.77)	(2.29)	(0.24)	(2.97)
Unemployed	2.7	2.8	5.0	2.4	6.2	5.1
	(0.67)	(0.63)	(0.97)	(0.62)	(0.90)	(1.36)
Empl. stable	0.1	-0.2	-2.8	-3.1	0.3	-2.6
	(0.02)	(-0.05)	(-0.63)	(-0.69)	(0.07)	(-0.76)
Small city	6.3	3.8	1.4	9.0*	1.8	4.8
	(1.65)	(0.73)	(0.18)	(2.22)	(0.65)	(1.42)
(Rural) village	11.5***	9.7*	5.2	9.9*	5.5	8.8**
	(4.08)	(2.75)	(1.00)	(2.62)	(2.05)	(3.37)
Benefit received	-4.7	-1.9	3.2	-8.5*	-5.9
	(-1.27)	(-0.37)	(0.59)	(-2.78)	(-1.32)
# benefits received	4.1**	7.9***	5.3*	3.8*	6.3**
	(3.09)	(4.61)	(2.34)	(2.58)	(3.13)
Politics: radical left	-12.5	-12.5	-4.0	-10.4	-27.9***	-13.0*
	(-2.00)	(-1.63)	(-0.57)	(-1.93)	(-4.38)	(-2.38)
Politics: radical right	-16.7***	-22.9***	-16.7**	-17.3***	-28.8***	-20.7***
	(-4.25)	(-6.23)	(-3.61)	(-4.81)	(-4.10)	(-5.65)
Politics: other party	-5.8	-2.6	-3.8	-7.9*	-8.4	-5.4
	(-1.62)	(-0.70)	(-0.88)	(-2.32)	(-1.57)	(-1.61)
Politics: no vote	-14.8*	-20.3**	-15.0	-15.4***	-26.2**	-18.4***
	(-2.67)	(-3.75)	(-1.78)	(-4.65)	(-3.61)	(-4.30)
Politics: other response	-11.6*	-3.0	-8.3	-10.7*	-14.7**	-10.9***
	(-2.46)	(-0.68)	(-2.09)	(-2.64)	(-3.48)	(-4.38)
Constant	21.0***	33.3***	28.6**	18.3**	35.4***	29.9***
	(5.09)	(5.89)	(3.60)	(3.10)	(4.79)	(5.82)
Observations	1009	1008	1009	1009	1009	1009

Note: The table shows the coefficients of linear regression models with the individuals’ perceived frictions at different stages of the benefit application process as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Could easily receive if needed	Would qualify	Know how to apply	Simple and quick application process	Fair treatment by government office	Average
Female	-10.5**	-11.4*	-8.7*	-9.3***	-9.4**	-10.4***
	(-3.65)	(-2.87)	(-2.94)	(-4.55)	(-3.19)	(-6.39)
Age: middle	-3.3	-3.6	1.7	-7.2*	-5.2	-3.0
	(-0.87)	(-1.14)	(0.31)	(-2.64)	(-1.37)	(-1.05)
Age: older	-12.4*	-12.8**	2.9	-13.2**	-11.8*	-9.7*
	(-3.08)	(-3.70)	(0.60)	(-3.60)	(-2.27)	(-2.69)
Education: >= tertiary	1.0	-3.9	4.8	2.6	3.2	1.5
	(0.40)	(-1.69)	(1.57)	(1.14)	(1.20)	(0.97)
Income: medium	-7.0*	-5.2	-5.5	-1.8	-3.5	-5.6*
	(-2.96)	(-1.41)	(-1.82)	(-0.85)	(-1.19)	(-2.57)
Income: high	-1.7	-6.8*	-5.5	4.5	5.3	-2.1
	(-0.49)	(-2.81)	(-1.20)	(1.09)	(1.47)	(-0.77)
Parent	7.9	7.3*	10.8*	13.2***	10.8***	10.9**
	(1.41)	(2.53)	(2.87)	(4.75)	(4.84)	(3.52)
Unemployed	-3.7	2.7	5.1	-0.8	0.7	3.4
	(-1.01)	(0.48)	(1.10)	(-0.15)	(0.11)	(1.01)
Empl. stable	2.0	5.2	5.2	4.2	5.3	3.3
	(0.52)	(0.98)	(0.93)	(1.00)	(0.69)	(0.76)
Small city	2.0	2.2	4.5	-1.8	9.6*	2.7
	(0.48)	(0.47)	(0.97)	(-0.52)	(2.61)	(0.82)
(Rural) village	8.2	10.1	11.8	11.6**	15.4***	11.6**
	(1.94)	(2.10)	(2.08)	(3.66)	(4.66)	(3.35)
Benefit received	-0.5	-3.1	-7.0*	1.0	-0.1
	(-0.16)	(-0.53)	(-2.24)	(0.20)	(-0.01)
# benefits received	4.4*	9.4***	11.0***	0.9	3.6
	(2.65)	(4.48)	(5.95)	(0.54)	(1.63)
Politics: radical right	-2.4	-12.5*	-17.8**	-4.3	-3.4	-6.9
	(-0.44)	(-3.04)	(-4.17)	(-0.80)	(-0.54)	(-1.94)
Politics: other party	-4.2	-5.5**	-2.5	-6.8	-3.3	-4.6
	(-1.27)	(-3.22)	(-0.66)	(-1.74)	(-0.97)	(-1.92)
Politics: no vote	-12.1*	-7.8	-5.5	-13.7*	-7.2	-9.9*
	(-2.83)	(-1.17)	(-1.01)	(-2.49)	(-1.45)	(-2.69)
Politics: other response	-13.7**	-14.3**	-8.3	-3.9	-4.3	-10.2**
	(-3.29)	(-3.86)	(-1.38)	(-0.87)	(-0.86)	(-3.66)
Constant	33.7**	33.7***	29.3***	23.4**	24.8**	33.4***
	(4.03)	(7.67)	(5.40)	(4.14)	(3.18)	(6.19)
Observations	1037	1037	1037	1037	1036	1037

Note: The table shows the coefficients of linear regression models with the individuals’ perceived frictions at different stages of the benefit application process as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). Detailed definitions of the explanatory variables are reported in Box 4.1.. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Filing taxes	Applying for government benefit (excl. health care)	Applying/enrolling children in school	Applying/enrolling children in daycare	Organising healthcare
Female	-1.2	-1.7**	-2.5***	-0.4	-0.5
	(-2.11)	(-3.34)	(-4.77)	(-0.44)	(-0.95)
Age: middle	-1.2	-1.3	-3.1	-3.6*	-0.8
	(-1.83)	(-1.35)	(-1.66)	(-2.22)	(-1.22)
Age: older	-2.3*	-2.8*	-5.7*	-5.3*	-3.0***
	(-2.97)	(-2.37)	(-2.91)	(-2.56)	(-4.34)
Education: >= tertiary	-0.4	-0.3	0.8	0.4	-0.8*
	(-1.25)	(-0.45)	(1.74)	(0.41)	(-2.69)
Income: medium	-1.4	-1.3	-0.7	0.2	-0.3
	(-2.10)	(-1.65)	(-1.20)	(0.23)	(-0.63)
Income: high	-0.5	-1.3	0.2	-0.1	1.6*
	(-0.52)	(-1.23)	(0.15)	(-0.06)	(2.19)
Parent	0.8	1.1	0.0	0.0	1.8*
	(1.44)	(1.43)	(.)	(.)	(2.46)
Unemployed	-0.1	2.2	0.5	0.1	0.8
	(-0.14)	(1.74)	(0.36)	(0.13)	(1.07)
Empl. stable	-0.8	-0.9	0.6	2.2	-0.3
	(-1.22)	(-1.30)	(0.62)	(1.58)	(-0.66)
Small city	-0.5	0.0	-0.5	0.1	-0.3
	(-0.80)	(0.04)	(-0.49)	(0.11)	(-0.32)
(Rural) village	-0.1	-0.5	-0.9	0.7	-0.1
	(-0.08)	(-0.47)	(-0.73)	(0.51)	(-0.32)
Politics: radical left	-0.6	-1.5	-4.2*	-3.0	-0.8
	(-0.53)	(-1.12)	(-2.68)	(-1.24)	(-0.89)
Politics: radical right	-1.8	-1.6	-3.4	-2.4	-1.2
	(-1.79)	(-1.42)	(-2.10)	(-0.94)	(-1.94)
Politics: other party	-2.7*	-1.9	-3.7*	-2.3	-0.9
	(-2.88)	(-1.48)	(-2.40)	(-0.98)	(-0.82)
Politics: no vote	-2.8**	-3.1	-4.1*	-4.0	-1.9*
	(-3.68)	(-1.99)	(-3.05)	(-1.74)	(-2.40)
Politics: other response	-1.8*	-2.9*	-2.4	-1.8	-0.8
	(-2.62)	(-2.99)	(-1.14)	(-1.34)	(-1.37)
Constant	9.1***	8.7***	10.8**	6.7***	6.8***
	(9.76)	(7.22)	(3.83)	(5.51)	(6.97)
Observations	978	806	333	253	986

Note: The table shows the coefficients of linear regression models with the individuals’ reported time spent annually on different administrative procedures as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (e.g. time spent on filing taxes) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Filing taxes	Applying for government benefit (excl. health care)	Applying/enrolling children in school	Applying/enrolling children in daycare	Organising healthcare
Female	-1.6	-0.2	-3.2*	-2.1	0.6
	(-2.02)	(-0.34)	(-2.45)	(-1.48)	(0.76)
Age: middle	0.8	0.5	-0.7	-1.2	2.2
	(1.23)	(0.89)	(-0.49)	(-0.52)	(1.92)
Age: older	0.8	0.8	-3.9**	-5.4*	2.2*
	(1.00)	(0.80)	(-3.29)	(-2.26)	(2.33)
Education: >= tertiary	2.9***	0.6	1.0	0.2	0.9
	(5.16)	(0.82)	(0.93)	(0.13)	(1.40)
Income: medium	0.5	-1.3	1.6	-0.9	-0.4
	(0.55)	(-1.51)	(1.24)	(-0.68)	(-0.54)
Income: high	1.3*	-0.9	-0.8	-0.8	0.9
	(2.15)	(-0.79)	(-0.39)	(-0.33)	(0.71)
Parent	1.4*	3.2**	0.0	0.0	1.9**
	(2.14)	(3.51)	(.)	(.)	(3.26)
Unemployed	-2.2	-0.3	-2.9	-0.7	0.2
	(-1.59)	(-0.21)	(-0.68)	(-0.18)	(0.10)
Empl. stable	-2.3	-3.7	-6.4*	-3.7	-2.2
	(-1.54)	(-2.09)	(-2.55)	(-1.72)	(-1.18)
Small city	-0.1	1.4	0.4	1.1	0.4
	(-0.16)	(1.95)	(0.18)	(0.59)	(0.36)
(Rural) village	-0.2	2.0*	2.6	3.4	0.8
	(-0.42)	(2.72)	(1.19)	(1.96)	(0.93)
Politics: radical left	-1.2	0.7	-2.3*	-0.6	1.0
	(-1.12)	(0.43)	(-2.16)	(-0.43)	(0.82)
Politics: radical right	-0.4	-0.1	0.5	1.4	-1.0
	(-0.39)	(-0.05)	(0.22)	(0.91)	(-0.79)
Politics: other party	-0.9	-1.3	0.3	1.0	-0.1
	(-1.74)	(-1.77)	(0.29)	(0.70)	(-0.19)
Politics: no vote	-2.0	-2.9*	-3.1	-0.4	-2.2*
	(-1.67)	(-2.24)	(-1.86)	(-0.13)	(-2.92)
Politics: other response	-0.8	-2.8*	-1.5	-3.0**	-2.2
	(-0.61)	(-2.26)	(-0.66)	(-3.13)	(-1.69)
Constant	7.6***	6.5**	11.2*	9.2	5.5**
	(6.12)	(3.02)	(2.90)	(1.89)	(3.84)
Observations	913	760	241	238	915

Source: RTM 2022.

	Filing taxes	Applying for government benefit (excl. health care)	Applying/enrolling children in school	Applying/enrolling children in daycare	Organising healthcare
Female	-0.6	-1.8*	-1.1	-0.5	1.0
	(-1.10)	(-2.70)	(-1.32)	(-0.28)	(1.14)
Age: middle	0.4	-0.6	-0.0	-1.3	0.2
	(0.54)	(-0.60)	(-0.01)	(-1.36)	(0.12)
Age: older	-0.9	-2.2	-0.7	-4.0	-0.2
	(-1.26)	(-2.03)	(-0.44)	(-2.13)	(-0.14)
Education: >= tertiary	2.2***	1.5*	1.9	1.3	2.3*
	(4.76)	(2.40)	(1.29)	(0.81)	(3.07)
Income: medium	-0.8	-1.7*	0.1	0.4	-0.5
	(-1.37)	(-2.60)	(0.04)	(0.35)	(-0.54)
Income: high	1.5	-0.9	-0.3	1.5	-0.5
	(1.45)	(-0.85)	(-0.16)	(0.76)	(-0.39)
Parent	2.3**	0.2	0.0	0.0	-0.4
	(3.62)	(0.21)	(.)	(.)	(-0.58)
Unemployed	-1.4	2.1	2.1	2.4	1.1
	(-1.32)	(1.53)	(1.08)	(0.86)	(0.90)
Empl. stable	-1.3	-0.4	0.5	0.4	-1.0
	(-1.70)	(-0.80)	(0.37)	(0.33)	(-0.88)
Small city	0.1	-0.9	-0.1	-1.1	0.2
	(0.12)	(-0.76)	(-0.04)	(-0.48)	(0.19)
(Rural) village	0.6	-0.7	1.2	-0.1	-0.6
	(0.71)	(-0.62)	(0.53)	(-0.06)	(-0.70)
Politics: radical right	2.0	3.6	-2.6**	-2.0	4.0*
	(0.74)	(2.18)	(-3.28)	(-2.03)	(2.32)
Politics: other party	0.0	0.3	1.1	0.5	1.1
	(0.05)	(0.61)	(0.84)	(0.39)	(1.32)
Politics: no vote	-1.7*	-1.2	-2.7*	-1.2	-2.1
	(-2.52)	(-1.38)	(-2.59)	(-0.91)	(-1.79)
Politics: other response	-0.6	-0.2	-2.3	-0.6	-0.5
	(-0.55)	(-0.12)	(-1.87)	(-0.28)	(-0.39)
Constant	4.3**	6.7**	3.6	4.5	6.1*
	(3.49)	(3.93)	(0.96)	(1.23)	(2.96)
Observations	818	782	316	285	962

Note: The table shows the coefficients of linear regression models with the individuals’ reported time spent annually on different administrative procedures as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Family support	Education	Employment	Housing	Health	Disability	Long-term care	Public safety	Average
Female	-2.6	-4.1	-0.8	-3.1	-1.2	-11.3**	-2.7	-6.2	-4.1
	(-1.11)	(-1.42)	(-0.28)	(-1.15)	(-0.46)	(-3.26)	(-0.72)	(-1.74)	(-1.94)
Age: middle	-12.7*	-21.3***	-11.1	-10.8*	-17.0***	-8.7*	-18.6**	-16.9***	-14.4***
	(-2.94)	(-4.96)	(-1.83)	(-2.94)	(-5.28)	(-2.19)	(-4.15)	(-5.71)	(-5.64)
Age: older	-13.7**	-27.2***	-16.9*	-13.4**	-19.6***	-10.4*	-17.7**	-14.1**	-16.4***
	(-4.10)	(-6.38)	(-2.71)	(-4.19)	(-4.66)	(-2.45)	(-3.53)	(-3.43)	(-6.36)
Education: >= tertiary	2.4	6.2	3.3	-1.7	1.2	-0.3	-0.4	2.7	1.6
	(0.71)	(1.43)	(1.16)	(-0.51)	(0.39)	(-0.10)	(-0.13)	(0.88)	(0.63)
Income: medium	-4.3	2.6	5.6	2.5	7.6*	-1.2	4.4	0.0	2.2
	(-1.40)	(1.17)	(1.36)	(0.79)	(2.80)	(-0.38)	(1.65)	(0.01)	(1.21)
Income: high	-2.8	0.2	1.3	-2.1	-1.3	-7.1	2.0	1.4	-1.1
	(-0.59)	(0.05)	(0.39)	(-0.60)	(-0.39)	(-2.05)	(0.57)	(0.44)	(-0.55)
Parent	4.2	2.6	-1.4	3.9	-1.9	-0.3	4.8	-2.1	0.7
	(1.45)	(0.82)	(-0.42)	(1.32)	(-0.52)	(-0.14)	(1.41)	(-0.83)	(0.30)
Unemployed	-1.3	0.9	1.3	-3.2	-2.0	-1.3	-2.5	-0.6	-0.9
	(-0.36)	(0.28)	(0.33)	(-0.83)	(-0.55)	(-0.31)	(-1.04)	(-0.12)	(-0.40)
Empl. stable	5.7	-2.1	4.8*	-0.2	0.3	-1.5	2.0	0.4	1.1
	(1.76)	(-0.76)	(2.19)	(-0.05)	(0.07)	(-0.45)	(0.96)	(0.12)	(0.64)
Small city	-1.8	2.6	2.0	2.5	-3.1	-3.8	-0.9	-5.2	-0.9
	(-0.71)	(1.30)	(0.58)	(1.00)	(-1.36)	(-2.04)	(-0.37)	(-1.63)	(-0.76)
(Rural) village	2.0	6.9**	5.5**	6.4	1.0	0.4	3.3	1.0	3.4*
	(0.65)	(4.26)	(4.30)	(1.70)	(0.54)	(0.12)	(1.51)	(0.46)	(2.70)
Benefit(s) received	3.2	-0.5	14.0**	11.7	2.6	4.9			-3.7
	(0.70)	(-0.10)	(3.36)	(2.09)	(0.81)	(1.33)			(-1.47)
Politics: radical left	-0.6	-2.7	-0.1	-1.7	2.5	-0.7	-4.7	-4.0	-1.9
	(-0.09)	(-0.52)	(-0.02)	(-0.42)	(0.68)	(-0.18)	(-0.95)	(-0.85)	(-0.70)
Politics: radical right	-5.2	-5.6	-3.4	-6.1	-7.9*	-2.4	-4.8	-10.9	-6.0
	(-1.26)	(-1.44)	(-0.61)	(-1.82)	(-2.84)	(-1.09)	(-1.09)	(-2.04)	(-2.10)
Politics: other party	-2.6	-3.2	-3.4	-8.1*	-5.5	2.6	-6.7	-7.7	-4.4
	(-0.88)	(-0.73)	(-0.81)	(-3.03)	(-1.38)	(0.61)	(-1.99)	(-1.73)	(-2.03)
Politics: no vote	-5.6	-7.6	-11.3**	-10.5*	-12.9	-6.1	-7.8	-15.2	-10.0*
	(-1.06)	(-1.39)	(-3.30)	(-2.30)	(-1.72)	(-1.93)	(-1.85)	(-1.83)	(-2.60)
Politics: other response	-1.7	2.8	-4.7	-7.2*	-1.9	-1.6	-8.5	-11.2	-4.5
	(-0.41)	(0.54)	(-0.72)	(-2.19)	(-0.33)	(-0.66)	(-1.59)	(-1.63)	(-1.31)
Representation: agree	36.0***	24.4***	30.9***	36.6***	29.9***	30.8***	32.6***	26.6***	30.9***
	(6.59)	(4.90)	(6.77)	(9.04)	(5.43)	(4.38)	(6.21)	(5.08)	(7.80)
Fair share: agree	19.0***	24.8***	20.6***	20.4***	23.6***	15.0**	24.4***	19.5**	21.2***
	(4.69)	(5.10)	(4.97)	(4.38)	(6.27)	(3.67)	(6.09)	(3.13)	(7.19)
Undeserving recipients: agree	1.9	-3.4	1.1	-2.6	-0.0	1.6	1.7	3.7	0.5
	(0.67)	(-1.37)	(0.33)	(-1.03)	(-0.01)	(0.65)	(0.54)	(1.15)	(0.28)
# benefits received									1.8
									(1.77)
Constant	27.5**	44.0***	26.4**	30.9***	42.5***	34.7***	33.1***	44.9***	36.2***
	(3.94)	(7.58)	(3.59)	(6.14)	(5.22)	(5.76)	(4.32)	(6.60)	(8.94)
Observations	1019	1019	1019	1019	1019	1019	1019	1019	1019

Note: The table shows the coefficients of linear regression models with the individuals’ short-term concerns in different risk categories as dependent variable and the variables on the vertical axis as independent variables. The risk categories are: Becoming ill or disabled; Losing a job or self-employment income; Not being able to find/maintain adequate housing; Not being able to pay all expenses and make ends meet; Not being able to access good-quality childcare or education for your children (or young members of your family); Not being able to access good-quality long-term care for elderly family members; Not being able to access good-quality long-term care for young or working-age family members with an illness or disability; Being the victim of crime or violence; Having to give up my job to care for children, elderly relatives, or relatives with illness or disability; Accessing good-quality healthcare. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 6.3 percentage points less likely to be satisfied with social protection services than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

	Family support	Education	Employment	Housing	Health	Disability	Long-term care	Public safety	Average
Female	2.4	3.1	0.2	-1.1	5.6	-6.7	-1.1	-0.9	0.2
	(0.71)	(1.00)	(0.06)	(-0.44)	(1.64)	(-2.08)	(-0.33)	(-0.25)	(0.11)
Age: middle	-14.8**	-13.2*	-1.7	-11.4*	-8.5*	-5.9	-8.8*	-8.6*	-9.2**
	(-3.46)	(-2.85)	(-0.40)	(-2.22)	(-2.71)	(-1.54)	(-2.55)	(-2.46)	(-3.73)
Age: older	-19.0*	-19.9***	-8.4*	-7.9	-10.7**	-7.5*	-6.5	-7.1	-10.8**
	(-2.76)	(-4.28)	(-2.15)	(-1.69)	(-2.95)	(-2.41)	(-1.56)	(-1.36)	(-3.96)
Education: >= tertiary	2.3	7.0**	-0.4	0.4	4.7	3.8	3.6	3.6	3.0*
	(0.82)	(3.21)	(-0.20)	(0.12)	(1.82)	(0.92)	(1.55)	(0.89)	(2.60)
Income: medium	5.4	9.8*	10.2**	3.6	5.2	6.7	5.9*	7.9	6.8**
	(1.77)	(2.65)	(3.22)	(0.99)	(1.69)	(1.96)	(2.37)	(1.79)	(3.52)
Income: high	6.2	17.4*	13.8**	6.3	6.4	3.6	10.0	7.4	9.1**
	(1.53)	(2.82)	(3.47)	(1.62)	(1.51)	(0.84)	(1.78)	(1.28)	(3.47)
Parent	21.3***	11.2*	6.7	11.0**	4.4	10.2***	9.3*	9.1**	10.1***
	(5.13)	(2.30)	(1.77)	(3.55)	(0.91)	(4.21)	(2.69)	(3.29)	(5.25)
Unemployed	-1.9	-11.1*	-7.7*	-0.7	-9.9	0.4	-3.1	-2.2	-4.8
	(-0.26)	(-2.26)	(-2.23)	(-0.15)	(-1.74)	(0.09)	(-0.76)	(-0.41)	(-2.08)
Empl. stable	-0.8	-6.6	0.5	0.3	-5.6	-4.9	-4.7	-0.7	-2.6
	(-0.19)	(-1.65)	(0.15)	(0.08)	(-1.32)	(-1.38)	(-1.11)	(-0.23)	(-1.49)
Small city	3.8	2.1	4.6	3.0	-1.4	-4.8	-0.0	3.2	1.2
	(0.81)	(0.39)	(1.08)	(0.96)	(-0.25)	(-1.31)	(-0.00)	(0.68)	(0.36)
(Rural) village	4.3	0.6	3.3	0.9	-3.2	-0.3	0.8	2.9	1.1
	(1.22)	(0.14)	(0.82)	(0.36)	(-0.61)	(-0.08)	(0.24)	(0.80)	(0.44)
Benefit(s) received	1.8	3.0	16.4*	9.8	-0.9	31.5***			-1.9
	(0.32)	(0.42)	(2.87)	(0.84)	(-0.24)	(5.15)			(-0.75)
Politics: radical left	-10.5	-1.6	-4.8	0.8	-6.5	-6.0	0.6	-4.2	-4.2
	(-1.65)	(-0.23)	(-0.70)	(0.19)	(-0.94)	(-1.03)	(0.13)	(-0.65)	(-1.44)
Politics: radical right	-7.1	-10.1	-11.6**	-11.1*	-8.7*	-8.7	-7.8	-15.9*	-10.1*
	(-1.28)	(-1.37)	(-3.05)	(-2.53)	(-2.35)	(-1.72)	(-1.41)	(-2.48)	(-2.86)
Politics: other party	-1.9	2.0	-4.7*	-0.9	-6.0	-2.3	-2.5	-0.8	-2.3
	(-0.75)	(0.75)	(-2.76)	(-0.36)	(-1.97)	(-0.93)	(-0.57)	(-0.17)	(-1.21)
Politics: no vote	-11.7*	-26.9***	-14.8*	-9.6*	-18.7**	-15.1*	-5.6	-26.6***	-16.2***
	(-2.31)	(-6.41)	(-2.62)	(-2.41)	(-3.16)	(-2.87)	(-0.95)	(-4.49)	(-6.82)
Politics: other response	-3.9	-7.8	-11.2*	-0.4	-12.9	-6.2	-6.1	-3.6	-6.3*
	(-0.78)	(-1.84)	(-2.24)	(-0.07)	(-2.08)	(-1.56)	(-1.27)	(-1.17)	(-2.47)
Representation: agree	24.9***	16.8***	30.5***	33.9***	26.2***	26.6***	28.8***	18.1***	25.6***
	(5.63)	(4.29)	(6.25)	(10.15)	(8.13)	(5.23)	(9.81)	(4.16)	(10.66)
Fair share: agree	12.2**	13.9**	14.8***	11.6*	22.7***	14.6**	13.0**	19.3***	14.9***
	(3.15)	(3.32)	(4.38)	(2.92)	(8.53)	(3.80)	(3.92)	(5.47)	(6.29)
Undeserving recipients: agree	4.4	-0.5	4.8*	2.9	-2.4	0.3	1.3	0.2	1.3
	(1.94)	(-0.12)	(2.27)	(0.99)	(-0.67)	(0.12)	(0.35)	(0.04)	(0.73)
# benefits received									2.1
									(1.98)
Constant	23.6**	45.9***	22.2*	19.4**	49.8***	28.4**	22.2*	42.6***	31.5***
	(3.11)	(9.78)	(2.83)	(3.01)	(6.97)	(3.71)	(2.86)	(5.70)	(7.81)
Observations	1008	1008	1008	1008	1008	1008	1008	1008	1008

Note: The table shows the coefficients of linear regression models with the individuals’ in different satisfaction with social protection services in different policy areas (namely Family support; Education; Employment; Housing; Health; Disability/incapacity-related needs; Long-term care for older people; Public safety) as dependent variable and the variables on the vertical axis as independent variables. Each row shows the statistical relationship between the explanatory variable (e.g. female) and the outcome (satisfaction with social services) relative to its reference category (men) – which is illustrated by the vertical line = 0 – taking into account the statistical relationship between the outcome and all other explanatory variables (e.g. age, income,…). For example, on average French women are 6.3 percentage points less likely to be satisfied with social protection services than comparable men (i.e. identical in all other characteristics included in the regression apart from gender). Detailed definitions of the explanatory variables are reported in Box 4.1. Observations are weighted by sample weights. Standard errors are clustered at the regional (province) level. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Source: RTM 2022.

Annex Table 4.A.18. Satisfaction with social protection services in the United Kingdom, accounting for perceptions of representativeness and fairness

Statistical relationship between satisfaction with social protection services in different policy areas (e.g. family support, housing,…) as dependent variable and individual-level characteristics (vertical axis) as estimated in linear regression models, the United Kingdom, 2022

	Family support	Education	Employment	Housing	Health	Disability	Long-term care	Public safety	Average
Female	-2.3	1.5	-6.5	-2.4	-3.4	-0.4	-1.4	1.9	-1.5
	(-0.83)	(0.60)	(-1.87)	(-0.97)	(-1.28)	(-0.11)	(-0.53)	(0.52)	(-0.74)
Age: middle	-12.9*	-16.0**	-9.1*	-8.2	-8.1**	-5.6*	-7.2	-6.6*	-9.3**
	(-2.84)	(-3.14)	(-2.53)	(-1.87)	(-3.92)	(-2.60)	(-1.66)	(-2.43)	(-3.43)
Age: older	-24.7***	-23.6***	-16.9***	-17.6**	-17.7***	-6.2	-15.0**	-14.1**	-16.9***
	(-4.95)	(-5.31)	(-4.44)	(-3.78)	(-4.65)	(-2.05)	(-3.37)	(-3.88)	(-6.72)
Education: >= tertiary	-3.5	4.2	1.9	2.7	0.6	0.7	-1.2	-0.5	0.6
	(-1.20)	(1.29)	(0.56)	(0.79)	(0.23)	(0.26)	(-0.42)	(-0.14)	(0.28)
Income: medium	1.2	2.1	2.4	2.2	1.7	-1.6	1.3	7.5*	2.1
	(0.39)	(0.60)	(1.03)	(0.72)	(0.52)	(-0.65)	(0.40)	(2.33)	(1.07)
Income: high	6.9	3.6	8.0	4.9	4.7	3.4	8.1	11.3*	6.3
	(1.42)	(1.17)	(1.92)	(1.24)	(1.03)	(0.92)	(1.89)	(2.44)	(1.92)
Parent	7.7*	12.4*	6.7*	2.6	1.0	4.3	7.0*	1.6	5.5*
	(2.59)	(3.07)	(2.51)	(0.79)	(0.30)	(1.31)	(2.86)	(0.51)	(2.87)
Unemployed	1.7	6.2	1.9	2.5	1.2	3.4	-0.0	11.6	3.8
	(0.31)	(0.86)	(0.38)	(0.61)	(0.25)	(0.86)	(-0.01)	(1.89)	(0.88)
Empl. stable	5.7	8.1	8.0*	6.1*	6.8	8.2*	1.3	11.1*	6.8*
	(1.72)	(1.84)	(2.30)	(2.45)	(1.61)	(2.31)	(0.36)	(2.48)	(2.35)
Small city	-0.2	4.2	1.8	5.8	1.6	3.1	4.8	1.9	3.0
	(-0.04)	(1.50)	(0.43)	(1.55)	(0.37)	(1.38)	(1.40)	(0.50)	(1.23)
(Rural) village	0.1	0.9	0.4	6.7*	2.9	3.7	6.4	1.5	2.9
	(0.02)	(0.32)	(0.09)	(2.29)	(0.69)	(1.29)	(1.49)	(0.38)	(1.27)
Benefit(s) received	2.4	1.4	-2.2	25.2**	4.5	0.7			-2.7
	(0.50)	(0.18)	(-0.36)	(4.10)	(1.37)	(0.17)			(-0.67)
Politics: radical right	-7.4	-14.1*	-11.7	-6.3	-7.4	-11.0	-10.8	-16.3**	-10.6*
	(-1.02)	(-3.01)	(-1.91)	(-1.06)	(-1.71)	(-1.82)	(-1.83)	(-3.60)	(-2.48)
Politics: other party	-10.5***	-6.4	-5.6*	-2.5	-3.4	1.1	0.2	-3.5	-3.8
	(-5.10)	(-1.20)	(-2.78)	(-0.99)	(-0.84)	(0.51)	(0.10)	(-1.28)	(-1.67)
Politics: no vote	-7.2	-15.0	-6.9	-7.5	-12.7*	-4.6	-4.0	-17.2**	-9.4*
	(-1.49)	(-1.97)	(-1.25)	(-1.79)	(-2.79)	(-1.08)	(-0.94)	(-3.22)	(-2.63)
Politics: other response	-12.0**	-14.0**	-6.8*	-11.9*	1.2	-6.6	-4.6	-7.1	-7.8**
	(-3.11)	(-4.05)	(-2.20)	(-2.53)	(0.21)	(-2.03)	(-1.69)	(-1.64)	(-3.27)
Representation: agree	27.2***	17.2**	24.4***	28.7***	31.4***	29.1***	37.0***	27.5***	27.8***
	(5.59)	(3.39)	(5.14)	(8.16)	(5.58)	(7.39)	(9.52)	(7.43)	(8.15)
Fair share: agree	17.9**	14.8*	19.1***	19.6***	16.3**	19.9**	12.5**	22.9***	18.2***
	(3.85)	(2.96)	(4.93)	(5.48)	(4.17)	(4.26)	(3.69)	(6.97)	(5.89)
Undeserving recipients: agree	4.5*	7.7***	-1.8	5.4*	2.2	3.9	3.9	3.0	3.3*
	(2.41)	(4.95)	(-0.59)	(2.45)	(1.02)	(1.82)	(1.33)	(1.44)	(2.55)
# benefits received									1.0
									(0.70)
Constant	32.6*	39.0***	34.6**	19.6***	34.5***	11.9	17.2*	24.1**	27.4***
	(3.04)	(5.77)	(4.04)	(4.55)	(4.65)	(2.19)	(2.36)	(3.61)	(4.93)
Observations	1037	1037	1037	1037	1037	1037	1037	1037	1037

Source: RTM 2022.

Annex 4.B. Figures

Note: The figures report the share that agrees with the following statements: “I feel the government incorporates the views of people like me when designing or reforming public benefits and services” (Panel A); “I feel that I receive a fair share of public benefits, given the taxes and social contributions I pay and/or have paid in the past” (Panel B); “Many people receive public benefits without deserving them” (Panel C).

Source: RTM 2022.

Note: The figure above reports the fraction of respondents who worry about short-term risks (left panel, by category) and are satisfied with social protection (right panel, by policy area) in each cluster in France.

Source: RTM 2022.

Note

← 1. Radical right parties are defined using Weisstanner, de Romémont and Bargu, (2021[3]). This is detailed further in Box 4.2.

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Countries

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Publications

Publications

Featured publications

Data

Data

Featured data

News & Events

News & Events

Featured Events

About OECD

About

Engage with us

Work with us

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment