Mind the gap: Young people and welfare-state related knowledge in deservingness and welfare attitude research

Abstract

Welfare deservingness opinions help explain welfare attitudes and hence are crucial for understanding the social legitimacy of the welfare state. However, even when considering deservingness alongside other explanatory frameworks, many open questions remain in the welfare attitude framework. This article argues that a novel research agenda focusing on welfare-state related knowledge and young people could considerably enrich current debates in deservingness and welfare attitude research. Deservingness assessments are made heuristically and could greatly depend on what people know (especially when they are misinformed). Studying this with young people is highly relevant, as the formative years are crucial for welfare attitude formation and change, even later in life. Research with young people provides unique opportunities for disentangling causal mechanisms between welfare-state related knowledge, deservingness and welfare attitudes. Moreover, it could help challenge welfare-state related misinformation and build resilience against disinformation. This thematic review outlines benefits, blind spots, and research trajectories when focusing on knowledge and young people in deservingness and welfare attitude research.

Introduction

In times of frequent welfare state reform, understanding the social legitimacy of the welfare state requires investigating welfare attitudes on different social benefits and obligations but also finding out what drives attitudinal differences (see Van Oorschot, 2000). In addition to looking at contextual factors (for example, Blekesaune, 2007; Larsen, 2006), previous studies trying to explain individual welfare attitudes focused on values and self-interest (Hasenfeld and Rafferty, 1989; Jæger, 2006; Kangas, 1997) or, more recently, welfare deservingness (Laenen, 2020; Van Oorschot et al., 2017). This is not surprising as who should get what, and why? (Van Oorschot, 2000) is back on the agenda, which justifies studying the social legitimacy of targeted welfare (Van Oorschot and Roosma, 2017). In turn, this puts focus on deservingness as a strong predictor of targeted welfare attitudes (Laenen, 2020; Van Oorschot and Roosma, 2017). However, even when considering deservingness alongside other explanatory frameworks, ‘the search for the individual-level determinants of welfare policy preferences is far from over’ (Laenen, 2020: 177).

Political knowledge might be a critical determinant outside the attention of previous research. After all, ‘the more informed people are, the better able they are to perform as citizens’ (Delli Carpini and Keeter, 1996: 219), which includes the task of critically assessing the status quo. Indeed, previous studies found significant changes in deservingness perceptions and welfare preferences when simulating a better-informed public (Althaus, 2003; Geiger, 2017; Kuklinski et al., 2000). However, so far, it is unclear why scholars should even assume that knowledge about the welfare state could influence deservingness and welfare attitudes and, if it does, how and with whom to pursue research on this matter.

Through a review of relevant literature, this article argues that focusing on welfare-state related knowledge and young people has the potential to enrich debates about deservingness and welfare attitude research. In what follows, the article (a) demonstrates the connection of welfare-state related knowledge to deservingness and welfare attitudes (knowledge–deservingness–attitudes nexus – referred to as the nexus) and its social policy implications, and (b) introduces future research trajectories. A central argument developed is the need to focus on young people, as this could help elucidate the causal mechanisms in the nexus and help address problems like welfare-state related mis- and disinformation. Ultimately, pursuing the novel agenda could lead to a better understanding of the social legitimacy of the welfare state.

Open the case: The knowledge–deservingness–attitudes nexus

Why should political knowledge matter for deservingness opinions and welfare attitudes? Earlier work argued that low political sophistication results in unstable and random attitudes (but see: Achen, 1975; Converse, 1970, 2006 [1964]; Erikson, 1979). However, later work demonstrated that ambivalence and on-the-spot answer processes were more plausible reasons for attitude instability (Zaller, 1992), meaning that attitudes of less sophisticated individuals are probably not random. On the contrary, an ill-informed citizenry is prone to systematic biases (Kuklinski and Quirk, 2000), which could mean that policy preferences differ if citizens are better informed (for example, Althaus, 2003).

Another perspective is questioning whether political knowledge is even sufficient or necessary for political competence. After all, people could use heuristics to arrive at valuable preferences and decisions despite limited expertise (Lupia, 1994; Popkin, 1994; Sniderman et al., 1991). The difference between less and more sophisticated individuals then lies in different rationales (Sniderman et al., 1991). Political competence would then not be about knowledge but arriving at the same result as if better informed, questioning what people actually need to know (Lupia, 2016). Indeed, people lacking specific knowledge can learn necessary information from their political environment (for example, via interest groups and political parties: Lupia and McCubbins, 1998). Two questions remain: what heuristics guide welfare attitudes, and why would it still matter to be well-informed?

Deservingness as social policy heuristic

Deservingness considerations are a common heuristic used to evaluate social policies. This is related to people focusing on beneficiaries and victims in policy evaluations, which is influenced by group perceptions (see Nelson and Kinder, 1996). For example, support for assistance then depends on whether individuals or groups are perceived to be responsible for their problems or seen as victims of circumstance (Sniderman et al., 1991). Similarly, Petersen et al. (2012) explain welfare support with perceptions of people in need as lazy or unlucky. Moreover, with deservingness-relevant cues present, people heuristically rely on classifications into cheaters and reciprocators in welfare judgements (Petersen et al., 2011). This is traced back to ancestral small-scale help relationships (Petersen et al., 2011; Petersen, 2012), implying that asking who should get what, and why is anchored in our evolutionary history. Accordingly, it makes sense that deservingness heuristics are used independent of knowledge, political ideology or cultural heritage (Aarøe and Petersen, 2014; Petersen et al., 2011, 2012).

Assessing deservingness in modern times is not as simple as distinguishing the lazy from the unlucky, however. For example, many people rely on us-versus-them categorizations, best visible in the immigrant deservingness gap (Reeskens and Van der Meer, 2019). The CARIN typology is the most comprehensive criteria-set people allegedly use to assess deservingness, employing a control, attitude, reciprocity, identity and need criterion. Deservingness perceptions (CARIN-criteria scorings) and deservingness valuations (CARIN-criteria importance) influence the support for social rights and obligations of target groups (Laenen, 2020; Van Oorschot, 2006; Van Oorschot and Roosma, 2017).

Deservingness and the problem of limited knowledge and misinformation

Why focus on knowledge if (deservingness) heuristics can lead to sound welfare attitudes? Most importantly, even though people use heuristics independent of knowledge, how and what shortcuts they use depend on what they know and the available information (Kahneman, 2012; Sniderman et al., 1991). Additionally, there is evidence that structural changes stand in the way of learning from the political environment as dealignment processes lead to less-solid ties between people and political parties (Biezen et al., 2012; Dalton, 2014). Moreover, class cleavages and positions of class-related actors are less straightforward than before (Cronin et al., 2011; Kitschelt, 1994), which could make class consciousness (see Korpi, 1983; Stephens, 1979) less important for welfare support. This reduction in the ability to rely on cues from the political environment increases individual responsibility and the relevance of knowledge as more sophisticated individuals are more likely to accept messages and decide corresponding to their political values (Lau and Redlawsk, 2001; Zaller, 1992).

This brings us back to the initial argument that information and knowledge are decisive when using heuristics, which might also be true for deservingness. For instance, deservingness assessments differ significantly in the presence of cues about why a person became unemployed, with only people missing such information resorting to stereotypes (Aarøe and Petersen, 2014). Thus, understanding the reasons for unemployment or knowing the unemployment rate could be important. Indeed, the unemployed are perceived as more deserving in times of high unemployment (Larsen, 2006). Additionally, group-specific perceptions cannot be used as a heuristic when people cannot recognize the beneficiaries or victims of a policy (Piston, 2018).

More important, however, is that ‘incorrectly’ used shortcuts (Lupia and Johnston, 2001: 196) can lead to serious mistakes in judgement. Such errors are most likely when heuristics are based on misinformation, defined as firmly holding false information (Kuklinski et al., 2000). Misinformation of welfare-state related facts is associated with welfare attitudes (Kuklinski et al., 2000) and deservingness perceptions. For example, overestimating the number of benefit claimants or fraudsters is significantly associated with lower perceived deservingness (Geiger, 2017). Such misinformation could result from disinformation in the media framing recipients as less deserving (for example, by highlighting fraud and using stereotypes: Devereux and Power, 2019), influencing support for retrenchment (Slothuus, 2007).

Particularly alarming in this regard is Kuklinski et al.’s (2000) finding that those furthest from an objectively-correct answer were most convinced of being right, which was mainly the case in questions on target-group related knowledge (for example, percentage of Black welfare recipients). This means misinformation could also occur regarding information logically connected to deservingness perceptions, as deservingness is a target-group focused approach. Misinformation guiding deservingness and welfare attitudes is incompatible with any definition of political competency. However, such a claim requires a causal connection between knowledge, deservingness, and welfare attitudes.

Causality and the nexus

Research on the causal relationship between political knowledge and attitudes could be biased when not accounting for bidirectional causality. The previous sections demonstrated how knowledge and information could influence deservingness opinions and welfare attitudes. Previous studies have already relied on the assumption that knowledge and misinformation might causally influence deservingness and welfare attitudes to simulate the effect of a better-informed citizenry (Althaus, 2003; Geiger, 2017; Kuklinski et al., 2000). However, bidirectional causality is not only possible but likely given that people are motivated reasoners when processing (political) information (Lodge and Taber, 2013; Redlawsk, 2002; Taber and Lodge, 2006). When confronted with new political evidence, people primarily consider information supporting current views, while contradictory information is questioned and argued against (Ditto and Lopez, 1992; Edwards and Smith, 1996). The effect strengthens with stronger attitudes and partisan or political identity (Taber and Lodge, 2006).

These biases help explain why simple presentations of facts are unlikely to change knowledge – let alone attitudes – regarding partisan, racial and ideologically-loaded topics (for example, Abrajano and Lajevardi, 2021; Kuklinski et al., 2000), and even if proper interventions change knowledge, this does not necessarily lead to changing attitudes (Green et al., 2011). Nevertheless, experimental studies also show that people can incorporate information to update their beliefs, group and issue attitudes as well as related policy support (Abrajano and Lajevardi, 2021; Jensen and Kevins, 2019; Kuziemko et al., 2015; Lawrence and Sides, 2014). Overall, this means that bidirectional causality is likely in the nexus. Consequently, misinformation-guided deservingness and welfare attitudes could indeed be problematic.

Social policy implications of a flawed knowledge–deservingness–attitudes nexus

Misinformation guiding deservingness and welfare attitudes is worrisome because flawed deservingness perceptions could drive welfare state reform for targeted social policies via policy responsiveness mechanisms (see Burstein, 2003; Brooks and Manza, 2006). Indeed, there is a strong correspondence between deservingness and social policies, that is, groups seen as more deserving enjoy more generous and less conditional social protection than groups seen as less deserving ¹ (also: Larsen, 2008; Laenen, 2020). It then gets very problematic in cases where misinformation guides deservingness opinions in a way that leads to more disadvantageous policies for already worse-off groups.

Simultaneously, distortions could affect policy feedback mechanisms (see Korpi and Palme, 1998; Rothstein, 1998), which can be exemplified for the institutional logic of welfare attitudes that links institutional welfare-policy structures to individual deservingness perceptions (Larsen, 2006). As people outside academia are likely unaware of welfare regimes, ² it is more convincing to assume that meso-level structures (for example, income programmes) influence deservingness perceptions (Jordan, 2013; Laenen, 2018). However, this would still require at least some basic knowledge of income programmes. For example, assuming implicit higher reciprocity scorings and thus deservingness for unemployment benefit recipients than for social assistance recipients only makes sense if a person knows the difference between those programmes. The argument is not that there is no institutional logic but rather that individual attitudes may not be shaped by policies or institutions themselves but by perceptions of them, which could be influenced by framing from policymakers and the media (Larsen and Dejgaard, 2013). For example, following changes to Danish social policies, anti-immigrant sentiments increased as natives who lost their benefits due to the changes both blamed immigrants for their losses and considered them contenders for increasingly limited support, which populist politicians used to advance their welfare chauvinism agenda (Jørgensen and Thomsen, 2016). Overall, good reasons exist to focus on knowledge in deservingness and welfare attitude research. The remainder of the article will outline future research trajectories, including demonstrating why focusing on young people would be a fruitful approach.

Trajectories for studying the knowledge–deservingness–attitude nexus

Conceptualization and operationalization

Investigating the nexus requires clear conceptualization and operationalization of political knowledge, deservingness, and welfare support. For welfare support, scholars can rely on the welfare attitude module of the European Social Survey (ESS) as an excellent reference point accounting for the multidimensionality of welfare attitudes (Roosma et al., 2013). There exists no similarly well-tested set of deservingness measurements (Meuleman et al., 2020), which has limited the comparability of previous work. For instance, many studies investigating deservingness actually measure target-group specific welfare support (for example, the government’s role in unemployment protection), making it hard to compare to studies measuring target groups’ CARIN-criteria scores. Laenen (2020) suggests that better insights could be gained by clearly differentiating between welfare deservingness and welfare support, relative and absolute deservingness, and deservingness valuations and perceptions. Another idea worth pursuing is juxtaposing measurement alternatives to find out what better explains welfare support: general deservingness principles without reference to policies and target groups or public-image approaches asking for the deservingness of specific target groups (see Meuleman et al., 2020). Relying on a common vocabulary (Laenen, 2020) and being transparent in measurements would immediately improve clarity and comparability.

Conceptualizing and operationalizing political knowledge poses the biggest problem. Although many contributions exist on how to assess political knowledge, there is wide ‘diversity in the kinds of questions researchers use to operationalize this concept’ (Barabas et al., 2014: 840) – which does not even consider that measuring political knowledge should ideally also aim at procedural memory (see Lupia, 2016) or visual forms of knowledge (Prior, 2014). Even when using the narrower, more tractable definition as the ‘range of factual information about politics stored in long-term memory’ (Delli Carpini and Keeter, 1996: 10), this is accompanied by many tasks. For example, researchers must clarify what constitutes a fact and whether to focus on general or policy-specific information (Gilens, 2001).

Moreover, after deciding and justifying a set of factual questions, analysing answers is more complicated than it might initially seem. For example, researchers must transparently define what constitutes a correct answer when questions cannot simply be answered with a right-or-wrong one (Geiger, 2018). They must also consider possible interpretations of don’t-know answers (see Luskin and Bullock, 2011; Mondak and Davis, 2001). Moreover, incorrect responses might not mean misinformation but can represent a lack of numeracy (Ansolabehere et al., 2013), partisan cheerleading or congenial inference (Bullock et al., 2015; Prior et al., 2015; Schaffner and Luks, 2018). Good advice in this regard is using incentives and certainty measures for answers, which also allows for distinguishing the uninformed from the misinformed (Kuklinski et al., 2000).

Lastly, being more transparent about what is measured and avoiding big terms like political knowledge improves comparability (Lupia, 2016). Focusing on welfare-state related rather than general political knowledge is reasonable for the nexus. More specifically, when interested in unemployment, researchers should focus on unemployment-related information (for example, spending, benefits, outcomes), the deservingness of the unemployed and attitudes toward unemployment protection. The latter must be done systematically for various programmes and target groups because people rely on different deservingness criteria when asked about different policies and target groups (Heuer and Zimmermann, 2020; Meuleman et al., 2020). However, enhancing methodology is only the first step in determining what (mis-)information matters.

Welfare-state related (mis-)information: What information matters?

Finding out what information matters is not only about what people (don’t) know about the welfare state but whether it affects their deservingness opinions and welfare attitudes. The crux is detecting information necessary or sufficient to be competent at political tasks (Lupia, 2016). The few studies on welfare-state related knowledge show poor performance, even among political science students (Jensen and Zohlnhöfer, 2020). Although people can be correct, they are often wrong irrespective of being asked about welfare state input, output or outcomes. For example, people often think that spending on unemployment is higher than on pensions, are inaccurate about benefit design or overestimate the number of benefit claimants and fraudsters (Geiger, 2018; Kuklinski et al., 2000; Taylor-Gooby et al., 2003; Taylor-Gooby and Martin, 2008).

A promising path forward to detect instances where knowledge or misinformation influences deservingness is testing knowledge logically related to the deservingness criteria. For example, deservingness opinions might be distorted by false beliefs on the unemployment and fraud rate as this confounds the control, reciprocity and attitude criteria. On the contrary, it is less clear how being wrong about who provides social assistance should alter deservingness. More research is needed since some false beliefs are significantly associated with lower/higher deservingness, and others are not (Geiger, 2017). Doing so for different programmes and target groups could help determine who is affected most by welfare-state related misinformation, which also requires disentangling causal effects.

Approaching causality in the nexus

Clarifying causality is an important issue in the nexus, requiring finding out where potential causal relationships between (factual) knowledge and deservingness and welfare attitudes might lie and then testing those with approaches able to do so (Antonakis et al., 2010). An initial step could be conducting cross-sectional studies to test various plausible associations. Afterwards, scholars could implement survey experiments manipulating the potentially influential information (for example, Jensen and Kevins, 2019), preferably in a randomized block design, where participants are allocated to blocks based on their knowledge or misinformation (see Abrajano and Lajevardi, 2021). Also useful are instrumental variables (for example, Jaeger, 2008) or multiple measurements, for example, in panel designs (for example, Jæger, 2006), difference-in-difference designs (for example, Jerit and Barabas, 2017) and randomized-control trials. Those would allow testing time-dynamic relationships, long-term outcomes or intervention effects. However, the best approach to causality would mean studying ‘individuals who initially hold no beliefs or preferences about an issue and then track them over time’ (Kuklinski et al., 2000: 801). While such data does not exist, it is one of many reasons why studying the nexus with young people of different ages could be rewarding and should be considered critical to future nexus research.

Youth and the knowledge–deservingness–attitudes nexus

The role of welfare-state related knowledge in younger years

While there are many reasons young people should be included in deservingness and welfare attitudes research, this article asserts that the potential to disentangle causality in the nexus and better prevent misinformation from influencing deservingness and attitudes makes them the key to future research. Studying young people over time in their development offers potential for uncovering causal mechanisms and the role of external influences on knowledge, deservingness, and welfare attitudes. For example, such research could help explore whether political ideology is developed before deservingness opinions as proposed in current deservingness models (Laenen, 2020; Van Oorschot and Roosma, 2017).

Additionally, young people could rely even less on cues from traditional political channels than adults, further increasing individual responsibility. Moreover, (social) media socialization processes (for example, Barberá et al., 2015; Prior, 2005) could facilitate and reinforce misinformation through exposure to disinformation. In turn, this sheds light on the extent to which (mis-)information is already connected to deservingness and welfare attitudes at an earlier age.

Asking this is highly relevant as the formative years are central to developing political orientations and knowledge (Jennings, 1990; Neundorf and Smets, 2017). While it is right to assume life-long political learning and attitude change (Neundorf and Smets, 2017), recent research shows that socialization experiences in the formative years are crucial for welfare attitude formation, stability, and change even later in life (Neundorf and Soroka, 2018). Earlier work suggested that the formative years lie between 17 and 25/26 (Jennings and Niemi, 1981). However, as seven-year-olds are already sensitive to political events (Bartels and Jackman, 2014) and show signs of political orientation and knowledge (Deth et al., 2011), those years probably begin much earlier than previously thought (Neundorf and Smets, 2017). Thus, young people from early on must be aware of their social and economic context so that their political baseline is not built upon misrepresentations. This gets especially difficult in the context of ample disinformation, which is a perceived and actual threat facilitated by modern media consumption trends (Newman et al., 2022).

The problem of misinformation and disinformation: Has the (adult) train left the station?

Young people in their formative years are often seen as particularly vulnerable to disinformation due to their media consumption habits and cognitive development processes (Middaugh, 2019); however, there could be a crucial advantage in challenging misinformation and building resilience against disinformation in younger years. Cognitive biases such as motivated reasoning hinder challenging political misinformation as corrections compete with internalized partisan and racial identities and rigid ideological reasoning (Lawrence and Sides, 2014; Nyhan and Reifler, 2010). Such biases are stronger with more firmly held political identity, partisanship or attitudes. Assuming that young people in the formative years are less politically entrenched (Alwin and Krosnick, 1991; Neundorf and Smets, 2017), it is likely that they are more open to new welfare-state related information and less biased by motivated reasoning and selective exposure (see Hart et al., 2009).

Similarly, young people may better build resilience against disinformation (also, Middaugh, 2019). Increasing age is related to higher exposure, susceptibility and sharing of fake news (for example, Brashier and Schacter, 2020; Grinberg et al., 2019; Guess et al., 2019), explained by cognitive decline, low digital literacy and cognitive biases such as motivated reasoning. Accordingly, developing digital literacy skills when young could be crucial to combating disinformation even in later life, yet young people seem unprepared for this task (for example, Breakstone et al., 2021; McGrew et al., 2018). Pairing this observation with insights from the literature on the making of citizens (Neundorf and Smets, 2017), formal education may be a venue for this kind of work.

The problem of misinformation and disinformation: The role of schools and educators

Schools could be ideal for combating youths’ welfare-state related misinformation, increasing knowledge and fostering resilience against disinformation. Success in challenging or debunking political misinformation can be reached by relying on credible and trustworthy sources and presenting evidence in an appealing and coherent framework instead of only pointing out false information (see Guillory and Geraci, 2013; Geiger and Meuleman, 2016; Lewandowsky et al., 2012, 2020). Ideally, schools could meet these criteria. First, 14-year-old students’ trust in schools is higher than in governments, political parties or the media (Schulz et al., 2009). Moreover, trained educators are probably better suited than other socialization agents to promote welfare state conceptions and knowledge appealingly and with context.

Promoting welfare-state related knowledge in schools requires developing effective educational interventions, however. What sounds like a typical task is oddly not well-researched, as reflected in calls for studies assessing the causal effect of civic education (interventions) on political knowledge (Campbell, 2019; Geboers et al., 2013). Although well-designed studies exist aiming to do so (for example, Campbell and Niemi, 2016; Green et al., 2011), these are very rare. Instead, most influential studies on the topic have limits regarding causality (for example, Langton and Jennings, 1968; Niemi and Junn, 1998), are inconclusive or rely on questionable knowledge measurements and poor data (see Lupia, 2016). Future studies should focus on the nexus with youth, develop interventions based on the results, and experimentally test the causal effects of those interventions.

In addition, formal education could help build resilience against disinformation (Heyneman, 2021). Schools and educators could facilitate digital literacy skills (Wineburg et al., 2022) and support inoculation and prebunking, that is, ‘making people aware of potential misinformation before it is presented’ (Lewandowsky and Van der Linden, 2021: 356). First experimental studies show successful inoculation in a school context by playing the ‘fake news game’ (Roozenbeek and Van der Linden, 2019b), which allows users to experience persuasion techniques first-hand. However, more research is needed here, whereby schools could serve as vital research areas, especially because there is a need for research in real-world settings on how to achieve permanent inoculation effects and on social aspects (spreading) of inoculation (Lewandowsky and Van der Linden, 2021). As inoculation can be effective independent of culture, age, and partisanship (for example, Roozenbeek et al., 2020; Roozenbeek and Van der Linden, 2019a), schools could greatly help in building ‘herd immunity’ against disinformation by being active in prebunking.

Practical considerations for investigating the nexus with (underaged) young people

While the previous sections focused on why scholars should focus on the nexus with young people, the last section presents ways to do so, including methodological and ethical considerations. Age-wise, studies should entirely cover the formative years, that is, starting from six/seven, with special attention given to adolescence as a peak of formative experiences (for example, Bartels and Jackman, 2014; Ghitza et al., 2022). Methodologically, survey items must be adapted to the respondents’ cognitive and emotional abilities to gain valid results (for example, Deth et al., 2011). This should be complemented with qualitative approaches, especially with younger children. The option of follow-up questions and allowing children to express thoughts in their own words could facilitate data quality. Lastly, researchers must consider stricter ethical and data protection standards (Alderson and Morrow, 2020; also, Felzmann, 2009).

This directly influences sampling, for example, by increasing the complexity of consent processes with lower ages (Alderson and Morrow, 2020). Overall, in quantitative studies, reaching meaningful samples gets more difficult with minors (for example, Kahne and Bowyer, 2017). An alternative to costly survey companies and simple convenience samples is systematically sampling students in schools. It is possible to obtain meaningful samples through techniques used in large educational studies, such as drawing fixed units of students from schools and classes drawn with probability-proportional-to-size from defined strata (Rust, 2014). Potential hurdles are increased bureaucratic effort and maintaining good contact with gatekeepers (Kristjansson et al., 2013). Willingness to participate could be facilitated by cooperation with (research) institutions specialized in educational practice. Additionally, when interested in marginalized youth, scholars should rely on non-traditional sampling (Sanders and Munford, 2017).

Summary and conclusion

This article argues that deservingness and welfare attitude research could greatly profit from a new agenda focusing on the knowledge–deservingness–attitudes nexus, especially with young people in their formative years. Deservingness opinions are essential for understanding welfare support but could depend on people’s welfare-state related knowledge. People heuristically decide who deserves help, which could be influenced when people lack specific knowledge or information. Being uninformed or misinformed could lead to significant distortions of deservingness and welfare attitudes. Detecting such instances is crucial, as this could interfere with policy feedback and responsiveness mechanisms that drive social policy reform.

Pursuing this research agenda, however, requires thorough scholarly effort starting with a clear conceptualization and operationalization of the main concepts. For welfare attitudes, the ESS is a good guideline. For deservingness, future research could juxtapose different measurements (general deservingness principles vs public image approaches, Meuleman et al., 2020) and rely less on questions measuring target-group specific welfare support. When focusing on declarative memory as a legitimate starting point for political knowledge (Converse, 2000), scholars should investigate welfare-state related knowledge (for example, spending, benefits and outcomes) rather than general political knowledge. Information logically connected to the deservingness criteria could be of interest, as this could reveal misinformation.

Second, scholars need to disentangle causality in the nexus. Asking who should get what, and why is a partisan and ideologically-loaded topic triggering motivated reasoning. Consequently, bidirectional causality in the nexus must be assumed. More clarity can come from investigating where causal connections might lie and, afterwards, relying on approaches better able to detect causality (for example, survey experiments).

Disentangling causality in the nexus is just one of many reasons why focusing on young people in their formative years would be valuable. How knowledge or misinformation influences deservingness in younger years is unclear. The same is true for the connection between deservingness and welfare support. Exploring this is crucial as experiences in the formative years help to explain attitude stability and change even later in life (Neundorf and Soroka, 2018). Additionally, stronger cognitive biases in adulthood make challenging misinformation and building resilience against disinformation more difficult. Schools as key socialization actors could play a special role here. They could be ideal for combating youths’ welfare-state related misinformation, promoting knowledge and fostering resilience against disinformation. However, proving this to be true requires studying the causal effects of political education (interventions). Otherwise, this recommendation will be just one of many claiming that increased educational effort helps with societal problems without proper evidence (Campbell, 2019).

Although research with young people in the formative years presents methodological, ethical and practical challenges, it can offer unique insights as most studies focusing on ‘youth’ only include adults. Ultimately, taking on the novel agenda would clarify our understanding of the welfare state’s social legitimacy by showing how what we (don’t) know about the welfare state affect our deservingness opinions and welfare attitudes.