Show Less
Open access

The Generational Welfare Contract

Justice, Institutions and Outcomes

Simon Birnbaum, Tommy Ferrarini, Kenneth Nelson and Joakim Palme

This groundbreaking book brings together perspectives from political philosophy and comparative social policy to discuss generational justice. Contributing new insights about the preconditions for designing sustainable, inclusive policies for all of society, the authors expose the possibilities of supporting egalitarian principles in an aging society through balanced generational welfare contracts.
Show Summary Details
Open access

Chapter 7: Contracts for trust

Justice, Institutions and Outcomes

Simon Birnbaum, Tommy Ferrarini, Kenneth Nelson and Joakim Palme

Citizens’ trust in each other as well as their trust in political institutions has been subject to quite intense academic debate. Low levels of trust may not only have adverse consequences for how individuals can cope with various aspects of their lives. It may also destabilize democratic government and make economic markets less efficient. With low trust, people are also more likely to abstain from mutually beneficial cooperation that underpins our ideas about positive-sum solutions in the generational structure of social citizenship. Bringing trust into the theoretical framework of comparative welfare state analysis is central for our analysis on generational welfare contracts, but it also gives rise to intricate questions about how institutional structures influence people’s willingness to rely on the actions of other citizens and of government. While these issues have received increased empirical scrutiny, perspectives that more explicitly focus on trust and generational aspects of welfare states remain unexplored.

Research on welfare states and trust has flourished, in part due to increased availability of comparative data for a large number of countries. Most studies tend to discuss trust in relation to the socio-economic position of individuals in the social structure, including the role of social class. By contrast, less is known about trust and age (Kocher, 2015). In this chapter, we will move into this largely uncharted terrain and analyse trust in relation to the generational welfare contracts discussed in previous chapters of this book. More specifically, we will investigate whether trust in other people and in governmental institutions reflect how countries have organized their generational welfare contracts. Are balanced generational welfare contracts associated with higher levels of trust, thus providing further evidence of positive-sum solutions in policymaking and cooperative exchange between people of different ages?

Besides adding another important piece of the puzzle to our understanding of central distributive processes in welfare states, the analyses presented in this chapter are relevant to the wider discussion about the long-term feasibility of social citizenship in an era of population ageing and structural change. In this context, strengthening relations of trust enjoys strategic priority in promoting and maintaining the social preconditions of just institutions from one generation to the next, precisely because it is so intimately linked to state capacity in matters of social justice. Arguments about possible links between age-related social citizenship rights and trust speak directly to this concern. Where trust is low, the state’s capacity to balance social citizenship across generations, prevent poverty and combat social inequalities more broadly is likely to be seriously restricted.

The chapter is organized as follows. Next, we introduce the concept of trust and distinguish its different forms. We then discuss the ways in which beliefs in governmental institutions and fellow citizens may be related to the generational welfare contract. In the subsequent empirical sections we first carry out descriptive analyses. Thereafter we use multilevel regressions to further investigate possible links between the generational structure of social citizenship and trust.


Trust is usually described as an attitude or judgment concerning the reliability of other people or institutions. When people trust other citizens (or government institutions), they may confidently leave control over decisions in the hands of others, and anticipate sensible judgments. Trust is thus associated with expectations of honesty, fairness or benevolence. Discussions of trust are often developed around the idea that people build up expectations about future scenarios that include actions of other persons, groups or governmental institutions towards oneself (Sztompka, 1999). Expectations of actions that are likely to yield positive, fair or beneficial outcomes for oneself are accordingly associated with higher levels of trust (Gambetta, 1988).

It is reasonable to believe that positive-sum solutions in the development of social citizenship depend on sufficiently strong and widespread trust in other people and society. Thus, in highlighting how mutual (positive or negative) expectations are formed and reproduced, questions about trust are likely to play a central role in understanding the sustainability of generational welfare contracts. The idea that trust is crucial for a cohesive society that works towards the well-being of all its members is of course not new to the social sciences. A century ago, Simmel pointed out the central role of trust for social development, in a well-cited quote identifying trust as “one of the most synthetic forces in society” (Simmel, 1950, p.326). Conceptually, trust has more recently also become closely linked to the wider academic discussion about social capital (Coleman, 1990). Although social capital in itself lacks a clear and undisputed meaning, it is often used in sociology and political science to address the value of vibrant social networks, as well as the virtues of bridging social ties between people of different backgrounds and identities (Dekker and Uslaner, 2001). While trust refers to people’s attitudes, social capital concerns the nature and value of social networks. Fukuyama (1995, p.26) illustrates the close affinity between trust and social capital, describing the latter as “a capability that arises from the prevalence of trust in a society or in certain parts of it”. In quantitative empirical analyses trust is also the most frequently used indicator of social capital (Torpe, 2003). Thus, questions about trust (and social capital) very much concern the extent to which individuals can relate to each other and to society in a spirit of civility and solidarity, irrespective of (for example) age, gender or socio-economic background.

While trust is inherently multidimensional, a key distinction can be made between social (or horizontal) trust and political (or vertical) trust. Broadly speaking, social trust concerns beliefs in the actions of other people, whereas political trust is linked to expectations of the functioning of governmental institutions, such as people’s confidence in parliaments, voting systems or in the legal system (Newton, 2001).1 While being conceptually distinct, trust between citizens and trust in institutions are often connected: if citizens do not trust a particular institution to enforce contracts, they may not trust persons to fulfill agreements, and vice versa. Thus, social trust and political trust are likely to feed into each other and may also contribute to virtuous or vicious cycles in generational politics. Such processes may either encourage gainful cooperation across age groups and reinforce mutual understanding for the common good or, on the contrary, cause divisions, hostility and lack of cooperative actions that cut across generational interests.

In view of potential positive-sum solutions and generational trade-offs in politics, it is important to acknowledge that questions about the exact causal relationship between political and social trust are far from settled, despite considerable scholarly efforts (Uslaner, 2002). It is beyond the scope of this book to deeply engage in the complex chicken-and-egg question about which type of trust that tends to appear first and feeds into the other. However, it should be noted that both Levi (1998) and Rothstein (2000) argue that political trust, reflecting beliefs that governmental institutions are trustworthy and non-corrupted, is a key factor for establishing cooperative behavior and strengthening social trust between citizens. Some recent empirical studies also support the idea that political trust promotes social trust (Rothstein and Eek, 2009; Sønderskov and Dinesen, 2016).


There is much to suggest that citizens need to have sufficient belief in the functioning of government and in the fair operation of the state’s legal and administrative institutions in order to also develop social trust, and it seems to be precisely here that the welfare state and social policy become highly relevant to the discussion. The idea that social trust may be promoted vertically from above, for example, by the ways in which welfare states reinforce cooperative exchange between people, has inspired research that takes on a cross-national perspective (Gelissen et al., 2012; Kääriainen and Lehtonen, 2006; Kumlin and Rothstein, 2005; Larsen, 2007; Rostila, 2013). Several of these studies indicate that social policymaking is closely intertwined with issues of trust, including beliefs in government institutions as well as expectations of honesty and fairness of fellow citizens.

The exact causal pathways linking welfare states and social policies to different types of trust are complex. Despite numerous studies investigating how trust varies between welfare states, explanations remain indistinct. Nonetheless, a few studies have tried to pinpoint important principles in the design of social policy that may explain why some welfare states generate not only high levels of political trust, but also why they are successful in building up social trust. Universalism appears to be a key institutional feature, affecting trust by means of transparency and efficacy of institutional arrangements (Kumlin and Rothstein, 2005), as well as equality of their outcomes (Uslaner, 2002). It should be noted that universalism is a tricky concept in social policy research. Often it quite narrowly refers to programs where eligibility is based on citizenship, while in other cases universalism more loosely refers to programs that adequately cover the needs of a large majority of citizens (Beland et al., 2014). Universalism in the debate on trust is mostly used in the latter, broader sense.

Coalition-building and risk-sharing in welfare states come in different forms and are intimately connected to institutional designs. State-corporatist principles in social policymaking (as further discussed in Chapter 4) may create high levels of solidarity within professions, but segmentation of interest is likely to form around status differences originating in occupational positions. In addition, conflicting interests may be expected to appear between “insiders” and “outsiders” on the labor market. Stratification associated with more market-oriented welfare states – with stronger elements of selectivity and means-testing in social policy – may instead reinforce conflicting interests between poor (net receivers) and better off (net contributors) citizens. Universalism, however, carries great potential to support patterns of solidarity that transcend both status and class differentials.

Universalism is widely associated with a particular form of benign cooperation resembling Levi’s (1998) ideas about contingent consent, where high levels of political and social trust play crucial roles for enabling stable majorities to embrace relatively high taxes and social spending to equalize life chances and create more egalitarian societies (Rothstein, 1998). The basic idea is that citizens more willingly will accept an expansion of government (including social policy) when they are confident that people receive their fair share of public commitments, and that burdens are shared equitably through impartial and non-corrupted institutions.2 It is reasonable to think that universalism can play an important role in satisfying and reproducing these conditions by providing all (or most) members of a society protection against various types of social risks without heavy reliance on the discretionary power and bureaucratic involvement associated with means-tested benefits (Rothstein and Uslaner, 2005). Universal welfare states may therefore sustain the everyday experience of well-functioning and impartial government institutions (political trust), thereby also supporting mutual cooperation between members in society (social trust). In this way, universalism can be expected to reinforce a sense of solidarity in society, supporting coalition-building and risk-sharing, with consequences for both political and social trust.

In line with our hypothesis about positive-sum solutions in generational politics, we suggest that similar processes and virtuous cycles of universalism, equality and trust also may appear in relation to the age-related structure of social citizenship. The basic idea is that generational welfare contracts guided by a reciprocity-based rationale matter greatly to the generational patterning of trust, and that this is important for understanding when and why universal programs are associated with higher levels of political and social trust (White, 2003). Balanced generational welfare contracts may here support beliefs about the impartiality and fairness of government institutions more generally by signaling that no age group is left behind, and that social needs and risks of all life stages are addressed equitably. Political trust in the functioning of governments (i.e. social policy) may also create and support virtuous policy feedbacks of high social trust and cohesion between people of all ages, where no age group attempts to benefit from the welfare state at the expense of another, even when they for demographic reasons constitute a larger part of the voter constituency.

A universalistic strategy of smoothing social citizenship rights across major age-related social risks, similar to what we have observed in countries with balanced generational welfare contracts (see Chapter 4), not only offers a promising basis for mutually supporting the welfare of all citizens irrespective of age and promoting relational equality between current, overlapping generations. It may also be viewed as a major investment in the common assets of trust and state capacity to provide public goods, with repercussions for the long-term sustainability of welfare states and social justice. In this perspective, distributing social citizenship rights evenly across age groups serves the double role of supporting the social and institutional bases of trust between citizens and in relation to the political community. In the following, we empirically investigate the relationship between generational welfare contracts and trust, and examine whether data supports our claims about possible links between the generational structure of social citizenship and people’s confidence in government institutions as well as fellow citizens.


The Generational Patterning of Trust

There are several ways to measure social and political trust. To increase country coverage, we use data from the same social surveys as in Chapter 6. For the European countries we use the 2010 European Social Survey (ESS), and for the non-European countries we rely on data from the sixth wave (2010–14) of the World Values Survey (WVS). In the ESS we use the following question for social trust: “Generally speaking, would you say that most people can be trusted, or that you can’t be too careful in dealing with people?” Respondents were asked to choose between a score of 0 (cannot be too careful) up to a score of 10 (most people can be trusted). In the WVS we use the following question: “Generally speaking, would you say that most people can be trusted or that you need to be very careful in dealing with people?” Respondents could choose to answer, “Most people can be trusted” or “Need to be very careful”. In order to harmonize the two datasets, we dichotomized the response categories in the ESS (scores 0–5 5 cannot be too careful, and scores 6–10 5 most people can be trusted).

For political trust we use the following question in the EES, “How much do you personally trust each of the following institutions? Parliament, political parties and the legal system.” Also here, a response scale between 0 and 10 was used for each item (ranging from no trust at all to complete trust) and responses were dichotomized as above. In the WVS we rely on the following question, “For each of the following organizations, could you tell me how much confidence you have in them: parliament, political parties and the courts.” The WVS applies a four-point scale, ranging between a great deal of confidence to none at all, and responses thus again were dichotomized to achieve comparability with the ESS (scores 1–2 5 no trust at all and scores 3–4 5 complete trust). For both datasets, political trust is a dichotomized index of trust in the parliament, political parties and the legal system.

Our variables on political and social trust are thus dichotomized at the individual level and vary between 0 and 1 at the country level. Values close to zero indicate lack of trust, whereas values close to 1 show almost complete trust. Figure 7.1 shows levels of political and social trust in three age-related social risk groups by type of generational welfare contract in 18 OECD countries around 2010. Only country averages for each generational welfare contract are shown. Similar to our analyses in preceding empirical chapters, the first category, characterizing the period of childhood, includes people with dependent children; the second category, comprising working-age risks, includes childless people aged 18 to 64; the third category of old-age risks contains elderly people (65 years or older).

Note: The balanced contract includes Austria, Belgium, Denmark, Finland, France, Sweden and Norway. The pro-work contract includes Germany, Japan, the Netherlands and Switzerland. The pro-old contract includes Australia, Canada, Ireland, Italy, New Zealand, the United Kingdom and the United States.

Source: The European Social Survey and the World Values Survey.

Figure 7.1a–b  Social and political trust in three age-related social risk categories by generational welfare contract in 18 OECD countries around 2010

Social and political trust differs quite extensively across countries. For each age-related social risk, political and social trust is higher in countries with balanced generational welfare contracts, followed by welfare states of the pro-work type. Countries with unbalanced generational welfare contracts of the pro-old orientation achieve by far the lowest levels of political and social trust, even among the elderly. A closer inspection of data also reveals that political trust is more evenly distributed across age-related risk categories in countries with balanced generational welfare contracts, although differences in this regard are small in comparison with countries that have pro-work contracts. Social trust diverges somewhat from this pattern. In countries with balanced contracts, and to a lesser extent also in countries with unbalanced contracts of the pro-work type, there is a slight negatively skewed generational gradient in social trust. In these countries, social trust is slightly lower among childless people of working age than among families with dependent children, even going down a little further among people in old age. It is difficult to provide a sound explanation for such a generational decline in social trust, but we will return to this issue in the upcoming regression analysis.

Our general understanding of how age is related to trust is limited. Earlier studies based on large cross-national social surveys, like the ones we use in this chapter, provide somewhat inconclusive evidence. Li and Fung (2012) observe that social trust tends to increase with age, while Fehr et al. (2003) and Naef et al. (2008) demonstrate a decline of trust in older age groups. In another study Fehr (2009) suggests that there is a U-shaped relationship, where social trust tends to decline among middleage people. From a sociological perspective, we would assume levels of social trust to be positively associated with the number of close interactions of an individual, and that social trust (people’s attitudes) is closely linked to social capital (types of networks). To address whether our results reflect real differences in access to social capital of older citizens vis-à-vis younger ones would require other types of data than those we have at our disposal. However, it is important to point out – in the context of exploring potential positive-sum solutions in the generational structure of social citizenship – that social trust among elderly persons in countries with balanced welfare contracts by far exceeds that of countries with unbalanced contracts. The latter includes countries with generational welfare contracts that we categorize as pro-old.

A few countries depart from the main patterns above. Both the Netherlands and Switzerland, two countries with generational welfare contracts of the pro-work type, have levels of political and social trust that are on par with countries in the balanced group (Figures 7.2 and 7.3). Levels of political trust in Switzerland have previously been shown to be relatively high (Bühlmann et al., 2013), which may have to do with particularities in the Swiss political system, not least related to direct voting procedures. The relatively high levels of social trust in the Netherlands have in earlier studies been directly linked to fairly strong civic engagement and volunteering among citizens (de Hart and Dekker, 1999). Notably, both countries also have universal systems of old-age pensions.

Source: The European Social Survey and the World Values Survey.

Figure 7.2a–c  Political trust in three age-related social risk categories in 18 OECD countries around 2010

Source: The European Social Survey and the World Values Survey.

Figure 7.3a–c  Social trust in three age-related social risk categories in 18 OECD countries around 2010

Regression Analysis

The analysis above indicates that countries with different generational welfare contracts also have different levels of political and social trust. Countries with unbalanced generational welfare contracts (i.e. age-related social citizenship rights are disproportionally distributed, with lower levels of income replacement in major social insurance schemes) tend to perform less well in terms of political and social trust than countries with more balanced contracts. Next, we investigate whether this pattern also holds in more elaborate statistical analyses combining data at both country and individual levels.

Table 7.1 shows multilevel regressions of income replacement in major age-related social insurance schemes on levels of political and social trust in 18 OECD countries around 2010. The methodological approach is the same as in Chapter 6, and the focus is thus on the overall level (arithmetic mean) of income replacement in social insurance for our three age-related social risks (childhood, working age and old age). In order to capture whether people differ in their likelihood of having high levels of political and social trust, we again use dummy variables for each age group, with childless persons of working age as reference (working-age risk category). At the individual level we use the same adjustments for confounding factors as in Chapter 6: education, gender and family status (couple or single person). Due to the small number of country observations, we once again only adjust for GDP per capita at the aggregate level, and whether survey data is from the WVS or not.

The multilevel regressions clearly show that social citizenship is associated with trust at the individual level. People are more likely to develop high levels of both political and social trust in countries where the overall level of income replacement in age-related social insurance is higher. At the individual level, some interesting generational associations are observed. Compared to childless respondents in working age (reference category in the regression), elderly respondents (old-age risk category) tend to have greater trust in political institutions, whereas respondents with children (childhood risk category) are inclined to have somewhat lower levels of social trust. For similar reasons as in Chapter 6 on subjective well-being, the latter result deviates somewhat from the preceding descriptive data analysis due to the inclusion of family status as an individual-level variable in the statistical regressions.

When investigating whether the overall level of income replacement in major age-related social insurance schemes is linked to the generational gradients in trust noted above, we followed a similar procedure to the previous chapter and introduced a number of cross-level interactions as independent variables in the statistical regressions. The BIC statistic (see Chapter 6) shows that only the regression model including interactions of elderly respondents provides better fit to the data. Notably, the sign of the interaction term between elderly respondents (old-age risk category) and the overall level of income replacement in major age-related social insurance schemes is negative. Thus, the overall comprehensiveness of social citizenship is not only related to the likelihood of individuals reporting high levels of political trust more generally, but also linked to the generational gradient in political trust that is associated with old age at the individual level. The difference in political trust between elderly persons and childless working-age respondents tends to be lower in countries where income replacement in age-related social insurance is higher.

Table 7.1  Multilevel regressions of political and social trust on the overall level of income replacement in age-related social insurance in 18 OECD countries around 2010

Note: * p < 0.05, ** p < 0.01. Coefficients shown as log odds. Constants are not shown. Cluster robust standard errors in parentheses. Reference categories: working-age risk, female and tertiary education. BIC 5 Bayesian Information Criterion.


This chapter moved our analytical spotlight to social and political trust. Whereas social trust is often considered to be an important factor in shaping social ties between citizens, political trust has more to do with the perceived legitimacy of the state and its institutions. Both facets of trust and their relationships to different generational welfare contracts contribute important pieces to the puzzle of understanding how generational politics can receive broad popular support, as well as to how just welfare state institutions can be promoted and maintained for the sustainable future.

We observed clear relationships between type of generational welfare contract and both forms of trust, lending further empirical evidence for the presence of positive-sum solutions in generational politics. Balanced generational welfare contracts are related to higher levels of political and social trust. Differences in political trust between age-related risk groups also tend to be smaller in countries with balanced generational welfare contracts.

We also showed that elderly people tend to develop higher levels of political trust than childless persons in working age. However, this age-related gradient in political trust is closely entwined with how countries have organized their generational welfare contracts. Our analyses on individual-level survey data showed that in countries where the overall level of income replacement in age-related social insurance is higher – as is often the case in countries with balanced generational welfare contracts – the difference in political trust between elderly and childless respondents of working age tends to be lower.


1.  Sometimes reference is also made to a third type of trust, so-called thick (or particularized) trust, which is primarily formed between close family members and others with whom we share strong ties through daily contact (Putnam, 2000; Williams, 1998). Thick trust has less direct relevance for our present concerns of addressing positive-sum solutions in generational politics.
2.  An interesting illustration of the importance of political trust in attitudes towards redistribution is provided by Svallfors (2013). People with a preference for economic equality are less likely to support high taxes or social spending if they do not trust in the fairness and efficiency of government agencies. Thus, when people disapprove of increased taxes and social spending, we should not jump to the conclusion that they necessarily reject economic equality and poverty prevention as desirable objectives. Instead, the likely explanation is often that people have low trust in the functioning of governments, and therefore lack sufficient assurance that their contribution to institutions of redistribution will be reciprocated or that it will be used for its intended purposes (rather than ending up elsewhere, for example, due to corruption or inefficiency).