Responsibility for impairment shapes the perceived deservingness of welfare claimants with disabilities

When do people support government assistance for people with disabilities? Disability welfare programs account for large shares of national welfare budgets, but little is known about public attitudes toward disabled welfare claimants. Drawing on psychological research in stereotype content, we argue that attitudes toward welfare for people with disabilities are likely to be more conditional than previously acknowledged. In two nationally representative, preregis-tered survey experiments in Wales ( N = 3393) and Scotland (1707), we ask respondents to evaluate the deservingness of a fictitious disabled claimant to government assistance. We manipulate the claimant's outgroup status and the manner in which they acquired their impairment. We find that disabled claimants perceived as even somewhat responsible for their impairments are considered substantially less deserving of government assistance than those perceived not responsible, even when their needs for assistance are identical. Contrary to expectations, we find relatively modest and inconsistent outgroup penalties in perceived deservingness. Finally, we find large heterogeneous treatment effects among respondents holding to more authoritarian social values. These results challenge conventional wisdom regarding the universality of support for disability welfare and


I N T RODUC T ION
Disability impacts the lives of millions of British citizens.Roughly 20% of the British population reports some form of disability, and around one third of British households include one or more persons with a disability (Heslop & Gordon, 2014;Reher, 2021).Disability has farreaching implications for the social and material well-being of individuals.People with disabilities 1 are disproportionately unemployed and underemployed (Kruse & Schur, 2003;Powell, 2021), face discrimination in the labor market (Ameri et al., 2018;Kruse et al., 2018), are far more likely to live in poverty and food insecurity (Brucker, 2016;Coleman-Jensen & Nord, 2013), and report significantly higher rates of loneliness, social isolation, and discrimination than their nondisabled peers (Emerson et al., 2021;Mattila & Papageorgiou, 2017).Many of these inequalities are compounded by intersecting sources of marginalization, such as membership in ethnic or cultural minority groups, or geographic distance from government resources and services (Krahn et al., 2015).
Given these myriad forms of social disadvantage, many disabled citizens rely on financial assistance from the government to meet their basic needs.Disability welfare programs are some of the most costly features of the British social safety net.In 2016, U.K. governments spent roughly GDP£37 billion on disability, incapacity, and injury benefits, accounting for more than 15% of the total U.K. welfare budget, and around 2% of national income-more than 10 times the share of national income dedicated to unemployment benefits (Niemietz, 2016).Anxiety over the perceived fiscal burden of disability welfare programs has led to widespread retrenchment (Banks et al., 2012;Baumberg Geiger, 2017;David & Duggan, 2006).Between 1995 and 2010, U.K. governments conducted seven major reforms to disability benefits schemes, aiming to reduce expenditure by tightening eligibility criteria, reducing the generosity of payments, and increasing the burden of evidence placed on claimants to demonstrate the extent of their functional limitations (Banks et al., 2012(Banks et al., , 2015)).
However, while disability welfare programs are important sites of policy reform and political conflict, researchers have only a thin understanding of public attitudes toward disabled welfare claimants.On the one hand, people with disabilities are stereotyped as in need through no fault of their own and therefore deserving of assistance (Cuddy et al., 2007;Fiske et al., 2002).Thus, disability welfare programs are ubiquitous across diverse welfare regimes (Carpenter, 2012), and disabled citizens are considered more deserving than almost any other recipient group (Van Oorschot, 2000, 2006).On the other hand, disability is a diverse social category, and disabled people vary along characteristics that may shape their perceived deservingness.People with disabilities vary dramatically in the nature and severity of their impairments (Krahn et al., 2015), and disability routinely intersects with other marginal or stigmatized social attributes and identities known to shape perceived deservingness (Ford, 2016;Nario-Redmond, 2019;Schur & Adya, 2013;Van Oorschot, 2000).Research in welfare attitudes has not yet considered how common individual differences between people with disabilities may contribute to differences in perceived deservingness, or how such perceptions may vary among people with different political values and predispositions.
This article addresses this gap.In two nationally representative survey experiments in Wales (N = 3393) and Scotland (N = 1707), we examine how two common sources of heterogeneity among people with disabilities-the circumstances in which they acquired their 1 Terminology around disability is continually evolving and varies with cultural context and individual preference (Nario-Redmond, 2019;Reher, 2021).In the United States, a preference for "person-first" language-that is, "person with a disability"reflects a desire to avoid defining people by their disabling health conditions or impairments.This terminology is commonly expressed in the acronym "people with disabilities."In the United Kingdom, where this study is based, "disabled person" is more common, reflecting a commitment to the social model of disability (Oliver, 1996;Shakespeare, 2006).We affirm the plurality of opinion within the disability community and use both terms interchangeably throughout this article.

| 3
PERCEIVED RESPONSIBILITY AND DISABILITY WELFARE impairment and ethnicity-shape the perceived deservingness of disabled welfare claimants.Specifically, we examine attitudes toward a fictitious male subject with a brain injury acquired either from birth, in a motorcycle accident, or through illicit drug use.These studies yield three key findings.First, we find that claimants perceived as even somewhat responsible for their impairments are seen as substantially less deserving of assistance than those perceived as not responsible, even when their needs for assistance are identical.We improve on prior work by modeling the relationship between perceived responsibility and deservingness, providing us with greater certainty as to the psychological mechanism behind these results.Second, we find that in some instances, ethnic minority and migrant people with disabilities are seen as somewhat less deserving than native-born and ethnic majority people with disabilities, although results vary with perceived responsibility for impairment.Finally, we show that these disparities are particularly pronounced among respondents holding to more conservative and authoritarian social values.
These findings represent an important contribution to the literature in welfare attitudes.While research on perceived deservingness is a mainstay in the welfare-attitudes literature (Petersen et al., 2012;Van Oorschot, 2000), existing studies tend to focus on attitudes toward members of stereotypically low-deserving groups, such as immigrants (Van Der Waal et al., 2013;Van Oorschot & Uunk, 2007), the unemployed (Aarøe & Petersen, 2014;Petersen, 2012), and ethnic minorities (Ford, 2016;Gilens, 1996Gilens, , 2009)).Fewer studies have examined variation in perceived deservingness within relatively high-deserving groups, such as the elderly or people with disabilities.Whereas prior studies find the baseline of support for disability welfare is relatively high (Van Oorschot, 2006), we find that deservingness attitudes vary dramatically with relatively small changes in the characteristics of individual disabled claimants.Understanding attitudes toward individual disabled citizens is particularly important in the context of disability welfare, where determinations of eligibility are made at the level of the individual claimant (Banks et al., 2012(Banks et al., , 2015)).While such determinations are theoretically made on the basis of need, the discretion of street-level bureaucrats tasked with making these decisions may be biased by extraneous perceptions of the deservingness of particular claimants based on their individual characteristics.Our findings suggest that equally needy disabled claimants may face differential risks of exclusion from support based on their perceived responsibility for their impairments, and to a lesser extent their ethnicity.
Finally, these results have implications for mass political behavior.While prior work emphasizes the relative uniformity of support for disability welfare in the mass public, our results suggest that such attitudes may be destabilized by elite rhetoric that seeks to undermine or discredit disabled claimants.Thus, our findings may help explain why political elites who make cuts to welfare programs for even stereotypically high-deserving groups are not always punished at the ballot box.

T H E DE SERV I NG POOR?
Disabled citizens are widely thought to epitomize the "deserving poor" (Hampton, 2016;Van Oorschot, 2000, 2006).Numerous observational studies find that in advanced democracies, no social group besides the elderly is considered more deserving of assistance from the government than people with disabilities (Larsen, 2008;Van Oorschot, 2000, 2006;Van Oorschot & Roosma, 2017).This finding replicates across states with diverse welfare regimes, differential levels of welfare spending, and cultural attitudes toward welfare recipients (Alesina et al., 2004;Esping-Andersen, 1990;Larsen, 2008).
The regularity of support for disability welfare is explained in part by the content of disability stereotypes.Research in social cognition finds people with disabilities are ambivalently stereotyped as high in warmth and low in competence (Cuddy et al., 2007;Fiske, 2018;Fiske et al., 2002Fiske et al., , 2007)).While people with disabilities are broadly perceived as friendly, cooperative, and trustworthy (warmth), they are stigmatized as low in personal efficacy, skill, and social status (competence) (Fiske et al., 2007).At the core of this stereotype is the perception that people with disabilities lack responsibility for their circumstances, which tends to elicit pity, compassion, and a willingness to help 2 (Cuddy et al., 2007;Goetz et al., 2010;Nario-Redmond, 2019;Petersen et al., 2012).This basic stereotype profile is found across diverse cultures and is applied to a variety of impairment groups (Cuddy et al., 2007;Fiske, 2015;Nario-Redmond, 2019). 3In the absence of specific information, people tend to rely on cognitively accessible schemas and stereotypes when making deservingness judgments about members of social groups (Aarøe & Petersen, 2014;Conover & Feldman, 1984;Jensen & Petersen, 2017).Thus, studies that rely on undifferentiated category cues tend to elicit responses that reflect the content of common group stereotypes.
However, disability stereotypes are likely to be unstable predictors of attitudes toward disabled welfare claimants in many everyday political contexts.While stereotypes shape attitudes and behavior when information is scarce, people are less likely to rely on stereotypes when specific information is available about the motivations and intentions of particular individuals (Aarøe & Petersen, 2014).Thus, while people may express support for disability welfare in response to a category cue, support is likely to vary substantially in the face of more fine-grained information about specific disabled claimants (Jensen & Petersen, 2017).A number of experimental studies have examined how deservingness attitudes vary in response to the characteristics of individual claimants (Aarøe & Petersen, 2014;Gilens, 1996;Petersen, 2012), but no studies (to our knowledge) have applied these insights specifically to people with disabilities. 4 We believe this is an important oversight for two reasons.First, different impairment characteristics may be perceived as signaling different levels of deservingness.Disabled people vary dramatically as to the type and intensity of their impairments, their experience of impairment acquisition, and the extent of functional limitation they experience (Krahn et al., 2015;Nario-Redmond, 2019).To the extent that these characteristics signal different levels of need, or responsibility for one's circumstances, they should also shape the perceived deservingness of disabled welfare claimants (Van Oorschot, 2000, 2006).Secondly, disabled citizens are demographically diverse.Disability is common to every social and demographic category and readily intersects with other characteristics known to impact perceived deservingness, such as gender, ethnicity, sexuality, migrant status, and social class (Bogart et al., 2017;Nario-Redmond & Oleson, 2016;Schur & Adya, 2013).While outgroup penalties in perceived deservingness have been repeatedly demonstrated in the welfare-attitudes literature (Ford, 2016;Gilens, 1996;Kootstra, 2016;Van Oorschot, 2000), scholars have yet to examine how these characteristics might shape attitudes toward disabled welfare claimants.

W HO DE SERV E S H ELP ?
What specific characteristics of people with disabilities are likely to shape their perceived deservingness?Existing studies argue welfare recipients are more likely to be considered deserving when they satisfy several key deservingness criteria (Meuleman et al., 2020;Van Oorschot, 2000, 2006): namely, when they are in acute or pressing need (Delton et al., 2018), when they are perceived as lacking responsibility for their circumstances (Petersen, 2012;Petersen et al., 2012), when they are seen as grateful and docile (Cook, 1979;de Swaan, 1988), when they are perceived as having contributed to society in the past or likely to contribute in future (Huddy et al., 2001), and when they are perceived as more proximate in identity terms (e.g., ethnicity, citizenship) to dominant social groups 5 (Ford, 2016;Gilens, 1996;Larsen, 2008;Van Oorschot & Roosma, 2017).While people with disabilities are widely stereotyped as satisfying many of these criteria (Cuddy et al., 2009;Meuleman et al., 2020), they vary substantially along many of these dimensions in ways which may shape their perceived deservingness of assistance.
We argue that two common sources of intragroup heterogeneity among disabled citizensthe circumstances under which they acquired an impairment, and their ethnicity-should shape the perceived deservingness of disabled claimants to government assistance. 6Specifically, we argue that the manner in which a disabled claimant acquired their impairment will shape their perceived responsibility for their circumstances, and in turn their perceived deservingness (Jensen & Petersen, 2017;Petersen et al., 2012).Further, we expect that when people with disabilities are migrants or belong to ethnic minority groups, they will be less likely to be perceived as members of a shared political community, which will erode their perceived deservingness relative to ethnic majority people with disabilities (Meuleman et al., 2020).

Disability and responsibility
An extensive literature in welfare attitudes finds that the degree of perceived responsibility a recipient has for their circumstances powerfully shapes deservingness judgments (Aarøe & Petersen, 2014;Alesina et al., 2004;Gilens, 2009;Petersen, 2012;Van Oorschot, 2000).People tend to oppose benefits for those who appear to be in need through what is perceived to be their own laziness or neglect, but they support those who are perceived to be in need through no fault of their own (Jensen & Petersen, 2017;Petersen, 2012;Petersen et al., 2012).This finding operates both between and within recipient groups.Stereotypically low-responsibility groups (e.g., the elderly, people with disabilities) are considered more deserving than high-responsibility groups (e.g., the unemployed) (Van Oorschot, 2000, 2006).Likewise, deservingness attitudes vary within groups in response to cues about the level of responsibility of individual group members (Aarøe & Petersen, 2014;Jensen & Petersen, 2017;Petersen, 2012;Petersen et al., 2012).Several studies find that members of the same recipient group (e.g., the unemployed) receive vastly different deservingness evaluations when framed as, for example, "a cheater" versus "a reciprocator" or as "lazy" versus "motivated" (Petersen, 2012).Other studies suggest respondents are also attuned to more implicit responsibility cues, where the attitude or intention of the recipient is unclear.Jensen and Petersen (2017) find that due to an ingrained psychological bias to view sickness as randomly distributed, people are inclined to see the sick as more deserving than the unemployed.However, not all sick people are considered equally deserving of assistance.Respondents are less inclined to agree that the government should provide support for people with health conditions that appear more related to behavior (e.g., obesity) than disease (e.g., cancer).As these findings suggest, cues that signal even the possibility of responsibility for one's circumstances may substantially diminish an individual's perceived deservingness.
We expect that respondents will respond similarly to differences in impairment acquisition among people with disabilities.While some impairments are congenital, most are acquired, and the acquisition 5 These criteria are commonly summarized under the acronym CARIN-care, attitude, reciprocity, identity, and need (Meuleman et al., 2020;Van Oorschot & Roosma, 2017). 6While there are other characteristics that may shape perceived deservingness, we select these characteristics for two reasons: (1) Perceived responsibility is widely cited as the single strongest determinant of deservingness attitudes across diverse cultural contexts (Petersen et al., 2012;Van Oorschot, 2006), and (2) minority ethnicity is an important driver of welfare attitudes in multiethnic democracies like Great Britain (Ford, 2016).
of particular impairments is often related to individual behavior and experience.Workplace and stress-related injuries are common (Hampton, 2016;Heffernan, 2020;Turner & Blackie, 2018), as are those acquired through misadventure, risk taking, disease, or natural processes of aging and bodily decay (Krahn et al., 2015).Whereas some impairments may be more directly linked to individual behavior than others, a relatively small proportion of disabled people possess impairments that are likely to be perceived as randomly distributed.Rather, many impairments will be seen as related to the behavior or lifestyle choices of individuals and will be interpreted as implying varying degrees of personal responsibility.For example, Weiner et al. (1988) finds that while most disabling impairments (such as blindness, HIV/AIDS, paralysis) are considered equally permanent and unalterable, respondents are more supportive of job training for individuals with mobility or visual impairments than for those with HIV/AIDS.As the authors argue, HIV/AIDS is more commonly attributed to individual behaviors and is therefore seen as less deserving of support.On the other hand, interpreting these results is complicated by the presence of societal stigma and prejudice toward certain kinds of impairment.For example, as Skitka and Tetlock (1993) argue, the view that people with HIV are undeserving of government assistance reflects both perceptions of responsibility for circumstances on the one hand and homophobic attitudes on the other.Likewise, Druckman et al. (2021) find that university administrators are less willing to provide disability accommodations to college students with attention-deficit/ hyperactivity disorder (ADHD) than college students who are blind, primarily because the former are stigmatized as lacking a strong work ethic.As we explain later, we address this difficulty by holding impairment type constant, varying only the manner in which the impairment was acquired.
We argue that given the immense diversity of circumstances in which a disabling impairment might be acquired, perceived responsibility for impairment is likely to vary substantially between individual disabled claimants.Further, given the strong inverse relationship between perceived responsibility and perceived deservingness, we expect that claimants perceived as more responsible for their impairments will be seen as less deserving of assistance from the government.These expectations inform our first hypothesis 7 : H1.Perceived deservingness of government assistance will be lower for claimants perceived as more responsible for their impairments.
Importantly, attributions of responsibility are not merely straightforward reactions to factual information cues.Rather, they are normative judgments that vary among individuals with different political values and predispositions.For example, Skitka and Tetlock (1993) find that while conservatives are inclined to deter free-riders and punish violations of social norms, liberals tend to avoid making value judgments of individual behavior by allocating resources equally across claimants, even when they are depicted as responsible for their predicament or violating social norms.Thus, we examine the degree to which political values moderate the relationship between perceived responsibility and deservingness in greater detail later in this article.

Disability and intersecting minority identities
The perceived deservingness of claimants also varies with their group identity.People are often reluctant to redistribute resources toward those with whom they feel little cultural affinity or political obligation, such as recent migrants (Kootstra, 2016;Van Oorschot, 2006;Van Oorschot & Roosma, 2017) or toward members of racial and ethnic minorities, especially when they are stigmatized as lazy, indigent, or low in social status (Ford, 2016;Gilens, 1996;Nelson & Kinder, 1996;Winter, 2006).These findings are echoed in stereotype 7 All hypotheses were preregistered at aspredicted.org.
14679221, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License content studies.Migrants and poorer ethnic minorities are typically stereotyped as low in both warmth and competence, which elicits anger, contempt, and a reluctance to provide assistance (Cuddy et al., 2007;Fiske et al., 2002;Petersen et al., 2012).However, while disability routinely intersects with other minority identities (Krahn et al., 2015;Nario-Redmond, 2019), no studies have yet examined whether ethnic minority or migrant people with disabilities are perceived as differently deserving than their ethnic majority or native-born peers.
How should we expect deservingness attitudes to vary at the intersection of disability and other minority identities?Some scholars argue that in certain social contexts disability may serve as a "master" status that eclipses other dimensions of an individual's identity (Fine & Asch, 1988).If this is true, we may expect attitudes toward disabled welfare claimants to be relatively indifferent to their ethnicity or migrant status.However, there is little evidence elsewhere in the experimental literature that respondents consider only one identity in isolation when making deservingness judgments about individual claimants.Rather, people are very capable of adjudicating between different characteristics of individuals when evaluating, for example, their deservingness of government assistance (Ford, 2016;Gilens, 1996;Kootstra, 2016;Petersen, 2012) or their desirability as incoming migrants (Hainmueller & Hopkins, 2015).Thus, we expect that respondents will consider both disability and other intersecting minority identities simultaneously in making deservingness judgments.According to the stereotypecontent model, the interaction between disability and minority ethnicity could elicit a range of attitudinal responses.On the one hand, prejudice toward ethnic minorities may lead respondents to view ethnic minority claimants with disabilities as low in both warmth and competence, which would theoretically reduce their willingness to lend assistance (Cuddy et al., 2007;Fiske et al., 2002).By contrast, if respondents possess more egalitarian views toward ethnic minorities, then the minority ethnicity of a disability welfare claimant should have no impact on their perceived deservingness.Given the persistence of the ethnic and migrant penalties observed elsewhere in the welfare-attitudes literature, we expect that migrant and ethnic minority people with disabilities will be seen as less deserving on average than ethnic majority and native-born people with disabilities, all else being equal.However, given existing research on the strength of responsibility cues in shaping deservingness attitudes, we also expect any ethnic differences to be less determinant of deservingness attitudes than perceived responsibility. 8This discussion informs our second hypothesis: H2.Ethnic minority claimants with disabilities will be considered less deserving, all else equal, than ethnic majority claimants with disabilities.

R E SEA RC H DE SIGN
We test these hypotheses in two nationally representative preregistered survey experiments in Wales (N = 3393) and Scotland (N = 1707). 9The experiments were carried out as part of the Welsh Election Study (WES) (Wyn Jones et al., 2021) and the Scottish Election Study (SES) (Henderson et al., 2021), fielded by YouGov in April-May, 2021.In both studies, YouGov recruited large samples of Welsh and Scottish voters from their online panel of over 1 million British adults.YouGov uses a sampling frame to approximate the demographic composition of the Welsh and Scottish population and provides poststratification weights so that model estimates can be interpreted as nationally representative. 10Deservingness attitudes exhibit substantial cross-national variation depending on the structure of national welfare institutions (Esping- Andersen, 1990;Larsen, 2008) and cultural values toward welfare recipients (Aarøe & Petersen, 2014;Alesina et al., 2001;Van Oorschot, 2006).To maintain the comparability of our samples, we examine two countries with near-identical demographic compositions, welfare regimes, and cultural attitudes toward immigration. 11

Experimental design
The experiments share most major design features.Both experiments use a 3 × 2 factorial design, in which we randomly assign respondents to narrative vignettes describing a fictitious male subject who is unable to work due to an acquired brain injury (see Table 1).To directly test our hypotheses, the vignettes vary the group identity of the subject 12 and the manner in which they acquired their impairment.This allows us to manipulate the perceived social proximity of the subject to the respondent and the degree of perceived responsibility the subject has for their impairment.The design yields three impairment responsibility conditions and two group identity conditions.Finally, we hold degree of perceived need for assistance constant across conditions by specifying the functional limitation associated with the subject's impairment (inability to work).Importantly, our design does not allow us to observe any potential differences in attitudes toward a female claimant depicted in similar circumstances.We chose to frame the claimant in this study as male, as some of the behaviors included in the vignettes (e.g., riding a motorcycle) are more stereotypically associated with men than women.This design choice may limit the generalizability of our results.For example, it is possible that a female claimant may be judged more harshly or held to a higher degree of responsibility for 10 All analyses are weighted.Unweighted analyses yield highly similar and substantively identical results.

11
While the Scottish Government has some limited welfare powers, in both Wales and Scotland the vast majority of welfare powers remain reserved to the centralized U.K. government.Furthermore, recent polling data suggests Wales and Scotland have similar baseline attitudes toward immigration.
12 Study 1 varies both migrant status and ethnicity, whereas Study 2 varies only the latter.We describe this in greater detail in subsequent sections.
T A B L E 1 Experimental manipulation used in Study 1.

Injury type Treatment wording
Birth David is 28 years old and lives in Cardiff.engaging in behaviors that are more likely to be perceived by respondents as counter-stereotypic.While we expect our results would be substantively similar in response to a female claimant, our design does not allow us to determine this conclusively.

Measures
Both studies use identical measures of all variables.We estimate the effect of our treatments on two dependent variables, each measured using a single survey item on a 7-point scale.The first asks, "To what extent do you think [name] deserves some financial assistance from the government?"(not at all-to a very large extent).The second asks, "To what extent do you think [name] was responsible for their injury?"(not at all-to a very large extent).Questions were asked on separate pages of the online survey.We also collected measures of relevant demographics including gender, age, education, income, disability status, and whether the respondent has immediate family members with disabilities.To control for ideological heterogeneity, we included two fiveitem scales of a respondent's left-right and liberal-authoritarian values.Item wording for these scales is common across election studies in the United Kingdom (e.g., see British Election Study [BES]; Fieldhouse et al., 2020) and is provided in Section B of the online supporting information.

Empirical strategy
Our empirical strategy is identical in both studies.We estimate the main effects of our treatments using ordinary least squares (OLS) regression.We first model the bivariate relationship between each dependent variable and a categorical variable for the treatment: Following this, we estimate the relationship between perceived responsibility and perceived deservingness within each subsample by interacting perceived responsibility scores with the treatment:

ST U DY 1: M IGR A N T C U E Study 1: Procedure
Study 1 was carried out as part of the 2021 Welsh Election Study (WES) (N = 3393), fielded by YouGov from March 19 to April 6, 2021. 13Participants were randomly assigned to one of six narrative vignettes. 14As previously discussed, the vignettes vary the group identity 15 of the subject and the manner in which they acquired their impairment. (1) (2) 13 Study 1 was preregistered at AsPre dicted.org: https:// aspre dicted.org/ blind.php?x= iz9iv8. 14Random assignment is successfully demonstrated in Table 13 in the online supporting information. 15Our samples are very ethnically homogenous (reflecting the real populations of Wales and Scotland): the Welsh sample is 94% White British born whereas our Scottish sample is 93% White British born.We use the terms "ingroup" and "outgroup" reflecting this majority status.Robustness checks that control for ethnicity/interact with the treatment do not lead to any changes in results.Similarly, running the analysis including only respondents who are White Welsh/Scottish/English/Northern Irish lead to no differences in results.
The outgroup condition describes a man named Khalid who emigrated from Yemen 16 to Wales with his family when he was 5 years old.Thus, while Khalid is depicted as foreign born, he was raised in Wales and was not plausibly responsible for his family's decision to emigrate.The ingroup condition describes a man called David and makes no mention of birthplace or ethnicity. 17In each condition, we specify that the subject is 28 years old to avoid differential perceptions of old age, which may impact perceived deservingness (Huddy et al., 2001).In the first impairment condition, the subject acquired their injury due to complications during childbirth (no responsibility); in the second, the subject acquired their injury in a high-speed motorcycle accident (moderate responsibility); in the third, the subject acquired their injury after using illicit drugs (high responsibility).
Finally, we include an open-ended hard manipulation check at the end of the survey asking respondents to recall how [David/Khalid] sustained their injury.The vast majority of respondents were able to accurately recall the relevant information from the vignettes (see Figure 2 in the online supporting information).Excluding respondents who "failed" this manipulation check makes no substantive changes to results. 18

Study 1: Results
We first examine the effect of our manipulations on perceived responsibility for impairment.As illustrated in Figure 1, our impairment manipulation performed as expected: Respondents believed David and Khalid were most responsible for their disability when it was acquired by using illicit drugs, and they were least responsible when it was acquired at birth, with motorcycle accidents falling in the middle.For example, the perceived responsibility for the David (drugs) treatment was 4.4 points higher on a 7-point scale compared to the David (birth) treatment.We also observe differences in participants reactions to the different names, with Khalid viewed as more responsible than David in the birth condition (although substantively the difference is negligible).
Having established the success of our manipulations, we examine their effect on perceived deservingness.In Hypothesis 1, we predicted that perceived deservingness would diminish as perceived responsibility increased.The results, displayed in Figure 1, confirm this pattern for David who respondents saw as most deserving in the Birth treatment and least deserving in the Drugs treatment.This was not the case for Khalid, with respondents saying that he was less deserving in the birth treatment compared to the motorbike treatment.Outside of the birth treatment, respondents did not perceive Khalid to be any more or less deserving than David.As a result, we are unable to confirm our second hypothesis (H2).
To get a clearer estimate of the relationship between responsibility and deservingness, we model the direct effect of responsibility on deservingness within each treatment group.As illustrated in Figure 2, respondents who viewed David or Khalid as having more responsibility for their impairment considered them to be less deserving of receiving government assistance.However as detailed in Table 2, only in the Khalid (birth) condition does the treatment indicator remain negative (β = -.85)once responsibility is included as an explanatory variable.This suggests that respondents were using considerations other than perceived responsibility to judge the deservingess of Khalid in the birth treatment.One 16 Wales has one of the largest and oldest Yemeni communities in the United Kingdom. 17Names were randomly selected from the 10 most common U.K. and Yemeni names according to forebears.io,a site that geographically maps the distribution of forenames and surnames across the world.In 2019, 94.8% of the Welsh population identified as White, and 1.8% identified as Muslim (see https:// gov.wales/ equal ity-and-diver sity-stati stics -2017-2019).In the absence of a clear ethnicity cue, respondents should perceive "David" (a common name in Wales)-as Welsh born and White.

18
Replication code for analysis excluding these respondents is provided in the online supporting information.possibility is that respondents believe Khalid's family to be guilty of "benefit tourism" (Verschueren, 2015).If respondents believe that Khalid's family, after learning of his disability, chose to emigrate to the United Kingdom because of greater availability of government support, they may be inclined to see Khalid's claim to government assistance as opportunistic or exploitative.Thus, the seemingly punitive reaction to Khalid may be   understood as a displaced judgment on the actions of his family, rather than a direct evaluation of his deservingness as an individual.This interpretation is further evidenced by the fact that we observe no differences in perceptions to David and Khalid outside of the birth treatment.In these conditions, Khalid is depicted as acquiring his impairment several years after his family emigrated, and respondents therefore have no obvious reason to attribute their decision to emigrate to the pursuit of government benefits. 19 These results also illustrate the psychological dominance of behavioral cues in deservingness judgments (Petersen, 2012;Petersen et al., 2011Petersen et al., , 2012)).The motorbike and drugs conditions differ from the birth condition in that Khalid is depicted as having some agency in the acquisition of his impairment.Given that the penalty observed for Khalid in the birth condition appears not to extend to these conditions, we can assume that respondents in the motorbike and drugs conditions are focused on evaluating the relationship between Khalid's own behavior and his deservingness of assistance.Further, in the absence of a behavioral cue (as in the birth condition), respondents may place more weight on other considerations relevant to deservingness, such as migrant status or membership in an ethnic minority group (Meuleman et al., 2020).Finally, these results point to an important design limitation in Study 1.Given that we vary both ethnicity and migrant status in the outgroup condition, we are unable to make clean inferences about the independent causal effect of either.We address these limitations in Study 2.

ST U DY T WO: NA M E -ON LY C U E Study 2: Procedure
Study 2 was carried out as part of the postelection wave of the Scottish Election Study (SES), fielded by YouGov in May 2021.A large, nationally diverse sample of Scottish adults was recruited by YouGov to participate in SES (N = 3442), around half of whom were randomly assigned to our experiment (N = 1707). 20 In Study 1, we deployed a two-dimensional outgroup cue that elicited very different responses across impairment conditions and prevented us from drawing clean inferences about the effect of minority ethnicity on perceived deservingness.In Study 2, we conducted a direct replication of Study 1 with a more precise outgroup cue: We removed any reference to migrant status and varied only the name of the subject [David/Khalid]. 21In doing so, we aimed to observe how deservingness attitudes vary at the intersection of disability and ethnicity in particular, and the degree to which these attitudes are moderated by perceived responsibility for impairment. 22All hypotheses and measures in Study 2 are identical to Study 1, and we use the same empirical strategy. 19Given the absurdity of assigning responsibility to a baby for their disability, an alternate possibility is that a small number of respondents are responding in a deliberately humorous or insincere manner-in other words, "trolling" the survey (Lopez & Hillygus, 2018).While trolling may explain some respondents' decision to assign responsibility to either claimant in the birth condition, it does not explain why we observe such a stark gap in perceived deservingness between David and Khalid in the birth condition in Study 1.While we are unable to test it directly, we believe the benefits tourism hypothesis provides a more intuitive explanation for this result. 20Random assignment of treatment is successfully demonstrated in Table 14 in the online supporting information.

21
Name-only cues are commonplace in exposure studies examining race and ethnicity.For a review, see Sen and Wasow (2016).Aside from the name of the subject, the vignettes for each impairment condition in Study 2 read identically.Complete wording is available in the online supporting information. 22Research by (Kootstra, 2016) has shown that Europeans' differentiate between native-born ethnic minorities and those born elsewhere when evaluating deservingness of employment welfare.

Study 2: Results
Figure 3 illustrates the main effects of our treatments on perceived responsibility and deservingness of government assistance.Respondents reacted to the responsibility cues in the same way as Study 1, with David and Khalid seen as least responsible in the birth treatment, followed by the motorbike treatment, and finally the drugs treatment.Importantly, we observe no differences in the perceived deservingness of David and Khalid, suggesting the migrant cue was likely the source of the observed differences between our subjects in Study 1.
Figure 4 shows the relationship between perceived responsibility and deservingness.Consistent with our first hypothesis (H1), those who saw David and Khalid as more responsible thought of them as less deserving.On the other hand, we find inconsistent support for our second hypothesis (H2).Respondents considered Khalid to be less deserving than David in both the birth treatment and the drugs treatment.On the one hand, these results suggest that some significant ethnic penalty may be elicited with a relatively subtle outgroup ethnicity cue (a name).However, the magnitude of the effects observed here are negligible.
In sum, results from Study 2 provide mixed support for our hypotheses.While we find strong evidence that claimants with disabilities are seen as less deserving of government assistance when their perceived responsibility for their impairment is greater, confirming Hypothesis 1, we do not observe a consistent outgroup penalty across responsibility conditions, therefore failing to confirm Hypothesis 2. Importantly, we observe stark disparities in perceived deservingness across responsibility conditions despite the claimant's clear and unambiguous need for government assistance.

T H E ROL E OF POL I T ICA L VA LU E S
Thus far we have detailed the main effects of our treatments on David/Khalid's perceived responsibility for their injury and perceived deservingness of government assistance.
F I G U R E 3 Coefficient plot of Study 2 main effects with 99% and 95% confidence intervals.Full results provided in Table 7.All estimates are relative to reference category (David × Birth).Consistent with our expectations, we find that perceived responsibility for impairment substantially diminishes claimants' perceived deservingness of government assistance.We also find that migrant status and outgroup ethnicity have some effect on perceived deservingness, although effects are smaller and less consistent than expected.However, as previously discussed, research in welfare attitudes consistently finds that deservingness attitudes vary not only with claimant characteristics, but with the political values of respondents (Feldman & Steenbergen, 2001;Feldman & Zaller, 1992;Jensen & Petersen, 2017;Sniderman & Brody, 1977).Thus, it is that some of the small main effects we observe in response to our ethnicity manipulation may be masking more polarized attitudes among respondents with different political values.
Here, we examine the extent to which individual-level political orientations moderate the observed treatment effects using two common measures of political values.The first is a fiveitem measure of left-right political ideology, and the is a four-item measure of liberalauthoritarian values.The British Election Study measure of political ideology is suitable for the current research in that it focuses largely on attitudes toward socioeconomic hierarchy and preference for redistribution.By contrast, the authoritarian-values measure emphasizes rule following, attitudes toward punishment, and social conformity.Both measures have featured prominently in British Election Studies since the 1980s and tap theoretically and empirically distinct constructs (Evans et al., 1996). 23 How might we expect left-right and authoritarian values to shape attributions of responsibility and deservingness toward welfare claimants with disabilities?First, existing studies find that those on the political right tend to have stronger negative emotional responses to welfare claimants framed as responsible for their circumstances, make stronger attributions of personal responsibility, and are more inclined to withhold assistance than liberals (Gilens, 2009;Skitka & Tetlock, 1992, 1993).By contrast, right-wing values appear to have less of an impact on deservingness attitudes when claimants are perceived as lacking responsibility for their circumstances (Jensen & Petersen, 2017).Thus, we expect economically right-wing 23 In the WES data, the Pearson's r correlation between authoritarianism and left-right values is .16.In the SES data, the correlation is .09.

F I G U R E 4
Study 2: Within-impairment group effects of perceived responsibility for injury on perceived deservingness of government assistance.Full results provided in Table 7. PERCEIVED RESPONSIBILITY AND DISABILITY WELFARE respondents to attribute higher levels of responsibility to David/Khalid in the motorbike and drugs conditions and to view them as less deserving of government assistance, relative to more economically left-wing respondents. 24However, we do not expect ideology to significantly moderate perceived responsibility in the birth condition, where David/Khalid is depicted as having no responsibility for their impairment.Second, both political conservatives and those holding to more authoritarian social values are more inclined than liberals to punish perceived violations of social norms (Adorno et al., 1950;Skitka & Tetlock, 1993;Stenner, 2005).Thus, we expect both right-wing ideology and authoritarian values to have a particularly deleterious effect on perceived deservingness in the drugs condition, where the subject is depicted as both more responsible, and as having engaged in an ostensibly illegal activity (consuming illicit drugs).

Results
To examine the degree to which political values moderate the effect of our treatments on perceived responsibility and perceived deservingness, we calculate the effect of treatment assignment on our outcome variables, conditioned by respondents' pretreatment political values.To do this, we estimate an OLS model of the following form: where Y i is perceived responsibility or deservingness, Treat denotes a categorical variable with six values for the treatment group to which respondent i was assigned, and PolVal is a respondent's left-right ideology or authoritarian-values score.We then interact the treatment condition with both left-right and authoritarian values. 25Additional models provided in the online supporting information include relevant control variables but are not presented in main text.Inclusion of controls results in no substantive changes.

Left-right values
To aid interpretation of results we plot the effects by injury type (although note the model uses a six-way interaction).In each plot, higher values on the x-axis correspond to more authoritarian or right-wing values, and higher values on the y-axis refer to higher levels of perceived responsibility or deservingness.Figure 5 illustrates the relationship between left-right values and perceived responsibility within each treatment group.The role played by left-right values in moderating the effect of our manipulations on perceived responsibility was consistent across treatment conditions and across studies.Contrary to expectations, more right-wing respondents were not more likely to view the candidate as responsible for their impairments in any of the treatment conditions in either study.Furthermore, those higher in right-wing values did not assign greater responsibility on average to Khalid than David in any impairment condition.
Likewise, we observed no heterogeneous treatment effects on perceived deservingness among respondents with different left-right values.As depicted in Figure 6, while respondents higher in right-wing values were less likely than left-wing respondents to view the subject as deserving of government assistance, this relationship did not depend on treatment assignment.24 These expectations were not preregistered. (3) 25 Model tables are provided in the online supporting information (Sections D and F).

Authoritarian values
Figure 7 illustrates the role of authoritarian values in moderating the effect of the treatments on perceived responsibility.In both studies, we observe substantially large and significant conditional effects in the illicit drugs treatment, consistent with our expectations of authoritarians being more sensitive to apparent rule breaking.By contrast, more authoritarian respondents did not consider the subject more responsible for their injuries in any of the other conditions in either study.Figure 8 illustrates the role of authoritarian values in moderating the effect of the treatments on perceived deservingness.The role of authoritarianism varied across treatment conditions.In Study 1, authoritarians were significantly less likely to consider the subject deserving of assistance in all but one experimental condition (David × motorbike).In Study 2, more authoritarian respondents only assigned lower levels of deservingness to the subject in the drugs conditions, where a movement from the bottom to the top of the authoritarian-values scale was associated with a drop of between 2.83 and 3.07 scale points (on a 7-point scale).
Responses to the outgroup manipulation were again mixed and inconsistent.In Study 1, authoritarians reacted strongly to the name treatment in the birth injury, with those at the higher end of the scale considering Khalid to be substantially less deserving than David, with approximately a 3.3-scale-point difference between the two.This suggests that the

Implications
Three aspects of these results are noteworthy.First, these results suggest that political values play only a limited role moderating the effects of our responsibility manipulations on perceived responsibility.Contrary to expectations, we find no evidence that right-wing values moderate levels of perceived responsibility for impairment in response to any of our responsibility conditions.Similarly, we find that authoritarian values moderate perceived responsibility only in the drugs condition, where the claimant is depicted as engaging in illegal activity.
The lack of a significant relationship between political values and perceived responsibility is particularly striking in the motorbike condition given the greater degree of ambiguity in the responsibility cue.Intuitively, we may expect political values to play a greater role in shaping perceived responsibility in the absence of clear information signaling the claimant's degree of responsibility for their impairment.Our findings do not support this intuition.Second, these results suggest that right-wing and authoritarian values play different roles in shaping deservingness attitudes.On the one hand, the deservingness attitudes of more authoritarian respondents were sensitive to responsibility cues-especially in conditions where the claimant was engaged in illegal activity.By contrast, right-wingers rated the claimant as less deserving on average, but such attitudes did not depend on the treatment to which they were assigned.These results suggest that factors beyond the individual-level characteristics of claimants may be driving a preference for more conservative redistributive policy among right-wing respondents.
Finally, the interaction between political values and outgroup cues also yields inconsistent results.Right wingers and authoritarians tend not to assign higher levels of responsibility to Khalid than David and are not consistently more punitive toward Khalid than David in their deservingness evaluations.There are some important exceptions.For example, authoritarians are far more punitive toward Khalid than David in both the birth and motorbike conditions in Study 1.However, we observe no significant effects of this kind in Study 2. Thus, while right-wing and authoritarian values are often associated with lower levels of perceived deservingness across treatment conditions, such values are not consistently associated with more punitive attitudes toward claimants depicted as members of ethnic minority groups.

DI SC US SION A N D CONC LUSION
When do people support welfare for people with disabilities?Despite accounting for the majority of welfare spending, stereotypically high-deserving claimant groups such as people with disabilities have received little attention in the literature on welfare attitudes.This article has detailed two studies examining the effects of perceived responsibility for impairment and minority group membership on the perceived deservingness of disabled welfare claimants in Great Britain.Our results suggest that stereotypical perceptions of people with disabilities as the deserving poor mask far more conditional attitudes toward many disabled welfare claimants.We find that claimants perceived as even somewhat responsible for their impairments are considered significantly less deserving of government assistance than a claimant who acquired their impairment from birth.Even when the claimant's actual degree of responsibility is ambiguous, perceived responsibility increases and perceived deservingness diminishes substantially relative to when the claimant is depicted as having no responsibility for their impairment.Further, our findings suggest that perceived responsibility for impairment has a deleterious impact on perceived deservingness when the claimant has a clear and unambiguous need for assistance.
On the other hand, contrary to expectations, we found that respondents adjusted their deservingness evaluations modestly and inconsistently when the claimant was depicted as a member of an outgroup-namely, as an ethnic minority migrant.Indeed, the largest outgroup effects were observed in the most theoretically unlikely of conditions.In Study 1, respondents expressed punitive attitudes toward Khalid in the birth condition, where he is depicted  as having no responsibility for his impairment, or the decision of his parents to emigrate to Wales.By contrast, in conditions where responsibility cues were provided (motorbike, drugs), the claimant's status as an ethnic minority migrant had no effect on their perceived deservingness.Having considered the possibility of inattentive or insincere responding, we think it is more likely that respondents are seeking to punish Khalid or his family for perceived "benefits tourism" in this condition.These findings complicate our understanding of the role of migrant status in deservingness evaluations.Whereas migrant status alone has no significant impact on the perceived deservingness of a disabled claimant, migrant cues which signal the possibility of benefit tourism significantly reduce respondents' generosity toward even stereotypically high-deserving claimants.
In Study 2 we conducted a cleaner test of our minority ethnicity hypothesis, with mixed results.While we observed a modest statistically significant outgroup penalty in the birth and drugs conditions (p < .05),we observed no such penalty in the motorbike condition.On the other hand, we observed these results using a relatively subtle ethnicity cue.It is possible that larger or more consistent ethnic penalties may be observed in response to a stronger outgroup cue.These results suggest that while outgroup status may shape deservingness evaluations in some instances, the effects of these cues are substantively small and inconsistent relative to responsibility cues.
Finally, the effect of our treatments on deservingness attitudes varied among respondents with different political values.On the one hand, respondents with more authoritarian values reacted more strongly to our responsibility manipulations.In several conditions, authoritarians assigned more responsibility and less deservingness to claimants framed as more responsible for their impairments, especially when the claimant acquired their impairment through illegal drug use.By contrast, respondents with more right-wing economic values were less likely to consider the respondent deserving of assistance on average regardless of treatment condition.
These findings have a number of implications for our understanding of the politics of disability welfare.Understanding the impact of individual characteristics on perceived deservingness is particularly urgent in the context of disability welfare, where eligibility decisions are made at the level of the individual claimant (Banks et al., 2012;Baumberg Geiger, 2017).To the extent that such decisions are shaped by perceived deservingness, our results suggest that equally needy disabled claimants may face differential barriers to support based on their perceived responsibility for their impairment, and to a lesser extent their ethnicity.Further, our findings suggest that welfare programs targeted at even relatively high-deserving groups may be vulnerable to discursive frames that cast doubt on the deservingness of claimants or their membership in a shared political community.As previously noted, the susceptibility of disability welfare attitudes to framing effects could have significant implications for mass political behavior.Specifically, our findings suggest that voters may be disinclined to punish politicians who make cuts to welfare programs for even needy claimants if they believe them to be undeserving of support.
These findings also hint at several avenues for future research.First, our design only examines responses to individuals with intellectual disabilities.Prior work finds that social attitudes toward people with disabilities vary somewhat with impairment type (Fiske et al., 2002(Fiske et al., , 2007;;Nario-Redmond, 2010, 2020), and that deservingness attitudes vary with the perceived neediness of the recipient (Delton et al., 2018).Impairments that are more visible or stereotypical are often interpreted as more authentic or legitimate than less obvious impairments (Dorfman, 2019a(Dorfman, , 2019b;;Nario-Redmond, 2019) and may elicit stronger deservingness attitudes.However, existing research finds that people apply the same basic stereotype profile to a diverse range of impairment categories (physical, intellectual, and sensory) (Fiske et al., 2002(Fiske et al., , 2007)), and such stereotypes tend to elicit a similar willingness to extend assistance (Cuddy et al., 2009).Thus, we expect that while the results presented here may vary in magnitude, they are likely to generalize to other impairment groups.Second, our studies only examine attitudes toward male claimants engaged in risk-taking behaviors, such as riding a motorcycle, that may be interpreted as stereotypically masculine.As previously noted, it is possible that respondents may react more punitively to a female claimant engaged in these kinds of risk-taking behaviors.While we do not expect the direction of results would substantially change in response to a female claimant, future work should more deliberately account for the possibility of gender bias in the perception of disabled welfare claimants.
Finally, our design examines responses to claimants who are unable to work and therefore have a clear justification for seeking government assistance.More work is needed to determine how deservingness attitudes might shift when the work capacity of disabled claimants is ambiguous.Similarly, whereas we examine responses to individual claimants, social policy is often framed in terms of target groups (Schneider & Ingram, 1993).Further research is needed to determine whether attitudes toward people with disabilities as a group are similarly malleable, or whether information about individual disabled claimants also shapes perceptions of people with disabilities as a group.
Understanding the political implications of disability is an increasingly urgent task for researchers and policymakers.While improvements in population health have prolonged human life, people are spending a greater proportion of their lives with disability and ill health (Carpenter, 2012).These demographic shifts impact both individual quality of life and the share of government spending dedicated to social insurance and welfare for the sick and people with disabilities.The fiscal implications of these trends are likely to be exacerbated by the COVID-19 pandemic.Physicians estimate one in five patients hospitalized with COVID-19 is discharged with a new disabling health condition or impairment, and numerous countries have moved to formally classify "long COVID" as a disability (Briggs & Vassall, 2021).With a growing share of the population set to acquire a personal stake in redistributive policies for people with disabilities, disability welfare programs are likely to remain an important locus of political conflict.

AC K NOW L E DGM E N T S
We are grateful to Paul Atwell, Ted Brader, Christian Fong, Ailsa Henderson, Vincent Hutchings, Rob Johns, Donald Kinder, Hannah Orban, Stefanie Reher, Rachel Statham, Susan Thorp, Kris-Stella Trump, Jacob Walden, Sarah Warren, Richard Wyn Jones, and to the three anonymous reviewers whose comments and critiques greatly improved this article.Previous iterations were presented at the 2022 and 2023 Midwest Political Science Association annual conferences, the 2021 U.K. Political Studies Association ECR Conference, University of Manchester, and Cardiff University's Wales Governance Centre.All remaining errors are our own.

F
Coefficient plot of Study 1 main effects with 99% and 95% confidence intervals.Full results provided in Table2in the online supporting information.All estimates are relative to reference category (David × birth).
Study 1: Within-impairment group effects of perceived responsibility for injury on perceived deservingness of government assistance.Full results provided in Table2.
Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License | 15

F
Predicted values of perceived responsibility across the distribution of left-right values by treatment condition.Plots include 95% confidence intervals.Histogram plots density of respondents along leftright scale.All bars plotted with width of .1.Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License main effects observed in Study 1 are driven almost entirely by respondents higher in authoritarian values.Authoritarians also viewed Khalid as substantially less deserving than David in the motorbike condition in Study 1.By contrast, we observed no significant differences in deservingness between David and Khalid in the drugs condition.The relationship between authoritarian values and the ethnicity manipulation was far more uniform in Study 2. Here, authoritarians did not consider Khalid any less deserving of assistance than David in any experimental condition.

F
Predicted values of perceived deservingness across the distribution of left-right values by treatment condition.Plots include 95% confidence intervals.Histogram plots density of respondents along leftright scale.All bars plotted with width of .1.

F
Predicted values of perceived responsibility across the distribution of liberal-authoritarian values by treatment condition.Plots include 95% confidence intervals.Histogram plots density of respondents along liberal-authoritarian scale.All bars plotted with width of .1.

F
Predicted values of perceived deservingness across the distribution of liberal-authoritarian values by treatment condition.Plots include 95% confidence intervals.Histogram plots density of respondents along liberal-authoritarian scale.All bars plotted with width of .1.
David sustained a brain injury due to complications during childbirth.As a result, he has a cognitive impairment and is unable to work Khalid is 28 years old and emigrated to Wales from Yemen with his family when he was 5. Khalid sustained a brain injury due to complications during childbirth.As a result, he has a cognitive impairment and is unable to work Motorbike David is 28 years old and lives in Cardiff.In 2014, David sustained a brain injury in a high-speed motorcycle accident.As a result, he has a cognitive impairment and is unable to work Khalid is 28 years old and emigrated to Wales from Yemen with his family when he was 5.In 2014, Khalid sustained a brain injury in a high-speed motorcycle accident.As a result, he has a cognitive impairment and is unable to work Drugs David is 28 years old and lives in Cardiff.In 2014, David sustained a brain injury after using illicit drugs.As a result, he has a cognitive impairment and is unable to work Khalid is 28 years old and emigrated to Wales from Yemen with his family when he was 5.In 2014, Khalid sustained a brain injury after using illicit drugs.As a result, he has a cognitive impairment and is unable to work 14679221, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 14679221, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 14679221, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 14679221, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 0 .1 .2.3.4.5 .6 .7 .8.9 1Lib-auth values 0 .1 .2.3.4.5 .6 .7 .8.9 1 Lib-auth values 0 .1 .2.3.4.5 .6 .7 .8 .
Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/pops.12978 by Welsh Assembly Government, Wiley Online Library on [27/03/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License