How Does the Content of Deservingness Criteria Differ for More and Less Deserving Target Groups? An Analysis of Polish Online Debates on Refugees and Families with Children

Abstract The article analyses opinions on deservingness expressed by social media users in debates about social welfare granted to refugees and families with dependent children in Poland. The article’s focus is on the content of deservingness criteria. This term describes the variety of factual and specific expectations applied to beneficiaries within each of the deservingness criteria. Qualitative content analysis of Facebook comments led to the finding that when users evaluate beneficiaries’ deservingness, they take into account their control over their own neediness, attitude, reciprocity in relation to the general population, identity and the level of need. However, within each of these deservingness criteria there is a plenitude of diverse, specific, often contradictory concepts of what exactly the sign of (un)deservingness is. The study shows that in the case of refugees, a group deemed less deserving, those content categories are more demanding and exclusive. In particular, the content of the need category proved broad and biased toward favouring a generally ‘more deserving’ group. The understanding of families’ need was often based on collective relative deprivation and the assumption that those who are needy have been neglected in previous social welfare programs, whereas refugees’ ‘real need’ was often a logically empty category.


Introduction
Deservingness theory seeks to answer the question of 'who should get what and why' by researching the public evaluations of target groups and their impact on the legitimacy of social welfare programs (van Oorschot, ; ; Laenen, ; Meuleman et al., ).Within this stream of analysis, scholars have provided evidence that the ranking of the perceived deservingness of groups of welfare beneficiaries remains rather stable in various countries (Laenen and Meuleman, ; Jensen and Petersen, ) and people apply the criteria of control, attitude, reciprocity, identity, need (CARIN) to evaluate these populations.It has also been found that these criteria are influenced by: welfare state regimes, the time and place-specific severity of social problems, and the class to which the opinion holder belongs (van Oorschot, ; Jeene et al., ; Meuleman et al., ).
In addition to the development of public perceptions about deservingness into a precise and quantitatively measurable concept, qualitative works on the topic have also been published (Herke, ; Heuer and Zimmermann, ; Laenen, ; Nielsen et al., ; Osipovic, ).They show that when people talk freely about who deserves what in a welfare state, they apply CARIN criteria, albeit in a highly tangled form that exhibits various, context-specific manifestations of the criteria, which ultimately are applied differently to different target groups.
The goal of this article is to better understand which specific context-related situations, behaviours and attitudes people find to be manifestations of deservingness of a target population and how these manifestations differ for target groups perceived as more and less deserving.The article looks at Poland, where deservingness is fiercely debated and is subject to major cleavages (Gugushvili and van Oorschot, ).It compares Poles' opinions on the (non)deservingness of families with dependent children and refugees, as they represent two groups deemed more and less deserving (Heuer and Zimmermann, ).The article adds to deservingness research by analysing online user comments.Moreover, while some research on factual deservingness criteria as they emerge in a bottom-up fashion has been conducted (Heuer and Zimmermann, ; Nielsen et al., ; Laenen, ; Osipovic, ), I am not aware of any studies which systematically compare specific expectations and evaluations applied to different target groups.

Theoretical background
One of the core findings of deservingness theory is that the general public deems various target groups more or less deserving of social welfare.Research on deservingness proves that this is the result of public perceptions of how these groups meet CARIN criteria.The position of families with dependent children and refugees, the two target groups this article looks at, is different in terms of those criteria.Overall, the general deservingness of immigrants is found to be lower than that of other groups (van Oorschot, ; Reeskens and van der Meer, ).This is consistent with more specific comparisons, such as those by Heuer and Zimmermann, (, p. ) who show that in terms of identity, families are seen in a more positive light than immigrants.
Moreover, research confirms that although the beneficiaries' complex characteristics are taken into account by the general public (Buss, ), the overall deservingness of the target groups is not the result of the same process of adding       evaluation outcomes for all groups in all the CARIN criteria.Instead, it results from 'deservingness valuations' (Laenen, )that is, applying various criteria to various target groups and the relative importance of these criteria.For instance, what makes immigrants 'less deserving' is, according to Heuer and Zimmermann (), evaluating them mainly through the prism of identity, reciprocity and to a lesser extent through need.At the same time, families' strong deservingness is equally based on assessments of their identity, rationale for social investment and their needs.
An under-researched aspect of deservingness is the concrete meaning of its abstract criteria, though this issue has begun to attract attention (Nielsen et al., ; Laenen, ; Osipovic, ).Hereinafter I will refer to this aspect of public opinions as the content of deservingness criteria.In this article that content stands for an array of opinions and claims about specific manifestations of deservingness criteria, together with their indicatorsthe concrete features, behaviours or situations which the general public recognizes as the signs of (non)deservingness.CARIN and other deservingness criteria measured by survey questions or vignettes are thus labels for groups of these manifestations.
Research suggests there are three features of deservingness criteria content which may be relevant when comparing diverse target groups.First, the qualitative analyses of deservingness in even single policy domains (Laenen et al., ; Nielsen et al., ) prove that each of the CARIN criteria stands for an array of diverse norms or ideals.The definition of CARIN criteria as 'trans-situational, abstract moral principles' (Meuleman et al., ) implies that they are meant to capture different manifestations of, for example, reciprocity, as in cases of a retired person and a young disabled veteran.However, the current literature shows that manifestations of these criteria expected of the members of a target group may vary significantly in terms of how demanding or exclusive the manifestations are.
For instance, the control is found to be applied both to beneficiaries' control over their status on the labour market and over their health (Laenen et al., ).In regard to the principle attitude, research on refugees' deservingness discourses show how specific expectations are applied to this group.They include cultural performance, adjustment and learning focused on integration (Zakariás and Feischmidt, ).In Herke's research on single mothers in Hungary (), their attitude turned out to be evaluated against the question of whether they commit welfare fraud by being single on paper only, while at the same time exploiting their ex-husbands financially.In regard to reciprocity, there is evidence that people take into account both beneficiaries' past and potential future contributions (Laenen et al., ).Nielsen and colleagues () showed that reciprocity in deservingness evaluations of immigrants includes considerations of their financial payments, functional contributions to the labour force in the    hosting country and behavioural reciprocity in the form of immigrants' positive energy and willingness to integrate.
The diverse meanings of the identity criterion have been well documented in migration studies.Zakariás and Feischmidt () speak of its construction along religious-cultural-geographical lines, while Carmel and Sojka () find that when evaluating migrants' deservingness British and Polish policymakers develop temporal-territorial, ethno-cultural, labourist, and welfareist understandings of identity criterion.Finally, need is found to have a range of manifestations in deservingness valuations, including financial needs, health needs and number of children (Laenen et al., ).Regarding financial needs, research shows that in some contexts people believe the wealthiest individuals are not needy and thus not deserving of social welfare.In other contexts, having children was found to be so expensive in general that it was deemed necessary that families with children receive support from the welfare state irrespective of their wealth (ibid).
Second, numerous authors (Heuer and Zimmermann, ; Nielsen et al., ; Laenen et al., ) have provided evidence that factual deservingness opinions are often based on evaluating target groups through the prism of amalgams of various CARIN criteria.For instance, opinion holders are interested in whether potential welfare beneficiaries are not 'work-shy'.This category links the criteria of control, attitude and reciprocity.In regard to immigrants, Nielsen and colleagues () argue that a hybrid of need and control criteria may be found in their interviewees' distinguishing between an economic migrant and a refugee.The former is deemed to have control over his or her status and a low level of need, while the latter is perceived as needy and lacking control over their own situation.
Qualitative studies on deservingness ideas held by welfare beneficiaries (Lavee, ; Herbst-Debby ; Kremer, ) provide a more detailed understanding of the factual interplay of deservingness criteria.They show how beneficiaries are reflexive agents who critically negotiate diverse public deservingness narratives.Users often accept such principles as reciprocity, yet assume that state intervention or labour market openness may be necessary for beneficiaries to have the agency needed to reciprocate.Thus, this stream of research indicates that CARIN criteria are hard to make sense of without addressing broader issues of social justice.
Third, the literature shows that the content of deservingness criteria is context-related.In interpretive terms this implies that some criteria have meaning only in a specific context.In causal terms, some studies suggest that the content of deservingness criteria is influenced by a context consisting of an institutional setting and the public image of a target group.For instance, Yoo () argues that constructing immigrants as undeserving of means-tested benefits in the US was based on the arguments raised during Congressional       hearings of their dishonesty and coming to the US only to receive welfare.Thus, the immigrants' alleged fraud was developed in this context as an indicator of their absent need.Policy feedback and public images are main explanations of a causal relationship between the context and deservingness opinions (Aaroe and Petersen, ).The latter involves constructing such stereotypes and criteria amalgams as 'welfare queen', 'the lazy poor' or 'the cheating migrant' (van Oorschot and Roosma, ).Overall, findings about the relationship between the context and the content of deservingness criteria are inconclusive and scarce.Moreover, causal explanations based on policy feedback and public images seem to be of little use in countries whose welfare state model is multi-layered, while the reshaping of target groups' images through media and political framing is a subject of highly contentious politics.
This article provides an analysis of the content of deservingness criteria as they appear in public debate in Poland.Poland should serve as a good case to investigate, given the complexity and contestations of various deservingness stances in the country.These stances are, foremost, the natural outgrowth of the hybrid model of the welfare state (Cerami, ), which includes the institutional legacy of communism, focused on universalist access to services and acknowledging people's needs, stark neoliberal policies on the labour market, and recent populist changes.Poles' adherence to these legacies is an aspect of strong sociopolitical cleavages in Poland which are also present in opinions on deservingness (Gugushvili and van Oorschot, ).Furthermore, processes described in the stream of analyses on the current right-wing populist shift in social policy (Herke, ; Kissová, ) apply to Poland, too.They entail active reshaping of deservingness opinions including appreciation of large middle-class families and the downgrading of immigrants (Lendvai-Bainton and Szelewa, ).

Research methods and context
The approach employed in this article, based on the analysis of Facebook comments, is intended to complement other methods used in research on deservingness.In ontological terms, public users' comments may be regarded as public opinion as, unlike survey responses, comments are public and created in social interaction (McGregor, ; Batorski and Grzywińska, ).In methodological terms, users' comments are helpful to analyse differences in factual deservingness opinions, as they are expressed spontaneously in users' natural language.In political terms, analyses which prove the use of social media comments by politicians and policymakers in developing policy proposals go a long way to explaining the relevance of those comments (McGregor, ).
The study seeks to answer the following research questions: How are deservingness opinions and their content factually formulated in online debates on policies targeted to refugees and families with dependent children?
What are the similarities and differences in the content of deservingness criteria in online debates on policies aimed at helping refugees and families with dependent children?
How are aspects of the debates, such as socio-political cleavages and past welfare policies, referred to in deservingness opinions?
Debates on both target groups were selected because of their salience in Poland.The issue on social welfare granted to refugees was a part of a broader discussion related to the 'refugee crisis' of .This topic was ubiquitous in the headlines and campaigning in the run-up to Poland's Parliamentary elections of October , when the 'Law and Justice' party (Prawo i Sprawiedliwość, or PiS, in Polish) employed a strong anti-immigrant rhetoric (Narkowicz, ) and argued for withdrawal of the decision to accept , refugees under EU relocation schemes.The share of Poles claiming that the country should accept no refugees increased from % in May  to % in April  (and remained at this level until research in ) (CBOS, ).Despite the increase of surveillance measures and the expanded scope of immigrant detention (Klaus, ) after PiS came to power, social welfare schemes granted to refugees remained unchanged.They include support granted in detention centres to people applying for refugee status, such as: living space, meals, pocket money ( € per month), a single benefit to buy clothing ( €), the refund of travel costs to immigration offices, healthcare services and access to public education.In rare cases, when applicants are allowed to live out of a detention centre, they are granted a monthly benefit of ca.- € per person.In any case, they are not allowed to work during the first six months after entering the country.After receiving refugee status (which takes approximately , months), individuals are entitled to a one-year integration programme organised as a part of the social assistance system.During this period they are granted a benefit of ca.- € per person per month., people applied for international protection in Poland in  and about , in , while the share of positive decisions on refugee status was about ,% in  (and about % in ) (NIK, ).
The debate on support to families with dependent children was triggered by Law and Justice's promise in electoral campaign of  to introduce a generous benefit for families.The 'Family  Programme' was launched in April  as the biggest social welfare programme in Poland.It entitles families with children aged less than  years to receive a monthly benefit of ca. € ( PLN) per child.Until July , it was granted upon means test to families with one child while no means-testing applied to families with more children.

     
The programme is different than the one Civic Platform (the ruling party that preceded PiS) administered, when only very low means-tested family benefits were granted.The programme quickly garnered strong public backing: in , % of Poles declared support for it and in  % supported it being extended to parents of first children without means testing (CBOS, ).
For the purpose of this study, the sampling of textual data was meant to mirror both online debates.At the same time, I strived to respond to challenges caused by employing users' comments.Research shows that social media users cannot be regarded as any kind of 'representative' polity (McGregor, ) and that social media discussions, including in Poland, become clustered in line with growing political polarisation: the feature is labelled 'the echo chambers'; use of algorithms curating targeted content for users also leads to 'filter bubbles' in social media (Batorski and Grzywińska, ).Thus, I sought to construct a sample of maximally diverse sites and assumed that qualitative analysis, focused on tracking the structure of arguments and not their frequencies, is appropriate to analysis of this kind of data.
The textual material I analysed comprised Facebook comments written below posts about refugees or refugee policy and the 'Family  Programme'.The sampling procedure was four-step.First, I selected major daily Polish newspapers from left-wing to right-wing. Next, using a key-word search ('refugee' and '') on Facebook, I selected from each outlet the  or  most-commented upon posts (posted in  in the case of refugee policy, and between the years - for family policy).The selection of periods to research mirrors the cycles when both issues were debated (see Pictures  and  in online appendix).Then, using Maxqda's Web collector function, I collected posts and up to the first  comments.Lastly, through reading comments one by one, I selected and coded only those which included opinions on deservingness.A detailed list of posts is provided in an online appendix.The length of textual material was similar in both debates, as was the number of coded text segments:  in the refugee debate, and  in the family policy debate.The proportion of coded segments to Facebook comments was, on average, ,:.
My operationalisation of a deservingness opinion followed the approach used elsewhere in qualitative studies on deservingness (Laenen et al., ; Nielsen et al., ).I coded both statements of the following structure: 'policy A is right/wrong, because people with feature X/meeting a criterion X (don't) receive a benefit' and of a structure: 'the people with feature X/meeting the criterion X should (not) receive a benefit', where X belongs to the content of deservingness category.While all of the comments analysed were public, to ensure anonymity, users' aliases are not referenced in the article.Data collection and analytical procedures have received a clearance of my university's ethical commission.

  
My research strategy was abductive.Thanks to recursive consulting of inductively obtained data with deservingness theory, I was able to refine the research questions and re-select textual material for the final sample.Accordingly, I applied qualitative content analysis to Facebook comments which meet the above operational criteria of a deservingness opinion, and the grounded theory approach to the coding procedure (Charmaz, ).At the stage of open coding, I labelled text segments with codes which were as close to specific deservingness ideas as possible, and I compared them continuously within and between both discussions.I then organized them into broader categories, which are referred to in the next section as deservingness content categories (see Table ).Simultaneously, I allocated the categories into the criteria of control, attitude, reciprocity, identity and need.Finally, I compared and refined the names of the content categories.During the course of my reading and coding, I wrote memos, mainly on the structure of the debate.Apart from CARIN criteria, I also uncovered deservingness justifications including: rejection of any deservingness criteria, criteria based on European social standards and on procedural justice.However, those were beyond this article's scope of analysis.

Research findings
The analysis showed that, in both debates, all CARIN criteria were discussed by Facebook usersalthough their salience differed, as presented in Table .In the case of families with children, reciprocity, need and attitude were almost equally salient in the debatethe evaluations of families' deservingness relatively often included these criteria but they were also presented in a very straightforward manner with a high number of comments and reactions (this translates into the number of 's in Table ).In the refugees debate, identity and attitude were the most important issues.
Deservingness criteria raised by users turned out to be contentious.This applies above all to users' different views about which criteria should be applied to the target population in general.There was a wider range of standpoints in the debate on family benefits, with three different viewpoints: an emphasis on universality of children's needs; a (neo)liberal expectation of reciprocity based on work; and the right-wing idea of beneficiaries' national identity together with  their reciprocity understood as childbearing.In the debate on refugees, deservingness criteria were often combined to characterise an 'ideal beneficiary.'A hypothetical 'most deserving refugee' would be a Christian woman who needed to escape from war and was eager to work and learn Polish.In both debates, users presented diverse, often contradictory understandings of deservingness criteria.Deservingness criteria were frequently brought up in amalgams.The most salient categories included: 'a Polish man doing his hard work every day', which encompasses the criteria of reciprocity, attitude and identity; 'a healthy, young and violent migrant pretending to be a refugee' (encompassing the criteria of need, attitude and control); 'a lazy, non-working mother who lives off welfare' (attitude and reciprocity); 'our Polish children shall never go hungry' (identity and need) or 'Poles' suffering, poverty and unmet needs' (identity, need and control).
Against this backdrop, users' comments included a variety of specific understandings of deservingness criteria.The categories they fall into are presented in Table .
Control is the least salient criterion in the discussions.But its content is very similar for both target groups.As Table  shows, in both cases categories focused either on the importance of external factors or on individual agency have been used.Of the former, users emphasised the low level of beneficiaries' control and their being forced to be in a situation of need.With regard to refugees, users argued that they had no choice but to flee from war.For families, the arguments of control referred mostly to living in poverty because of structural problems, such as regional unemployment or low wages.Here, it was argued that the government was responsible for the poverty they fell into and thus it should develop a systemic response to its own mistakes.The second category in the control criterion consisted of arguments that both refugees and families are fully responsible for their own fate because of their own decisions to migrate or to have many children.
As shown in Table , a salient category in the group of attitude-related opinions is work, which is regarded as proof the beneficiary is not lazy, does not cheat the system and is not too comfortable.The argument about the need to work is closely related to users' claim that a person deserving of social welfare should not 'live off welfare'.In debate on families, it was often argued that beneficiaries, mostly women, are undeserving if they are able to work but opt to take welfare instead.This comment is indicative: Only lazy asses who live off the state's mercy and social assistance do this [receive the benefit], shame on them!They [the government] should cease this benefit and report on the money, because it should go to children and not on those who don't know how to support them.(F_FK)       Facebook users also expressed attitudes focused on living off welfare that included: women divorce or don't get married in order to report lower household income, 'being passive' or 'doing nothing but waiting for the benefit' (F_RP).In both debates, having many children was regarded indicative of a desire to overuse welfare state.Of refugees, users often claimed that coming to Europe indicates a desire to 'live off welfare' and seeking to 'have an easy life on European welfare' (R_RP).
Further groups of attitude-related categories present in both debates refer to appreciating help or not being overdemanding, as well as making good use of benefits.Arguments about refugees included that those who throw food away or don't clean their flats should not receive welfare.A user expressed this idea in a following way: they should register them all and then help only those who really appreciate it, and those who only destroy everything should be sent back home (R_GW).
According to many users, a clear example of non-deservingness is wasting the goods or help they have received.A typical claim: If they waste food and don't want to eat it, let them die starving, [give them] no money, no flats, and quickly they will fuck off (R_FK).
In the case of families with children, 'being overdemanding' meant mostly claiming yet other benefits.Users were rather unanimous that those parents who spend their welfare assistance on drugs and alcohol shouldn't be granted it.
The most pronounced attitude-related expectation applied to refugees, who should respect the law, religion, culture and customs in Poland.Frequent ideas used in this context were: the need to assimilate, show respect to our culture, not building mosques, respecting Polish women, and not being aggressive.Arguments ranged from ideas of conditional helpgranting welfare only if a refugee proves a command of the language to full rejection of help, as expressed in this comment: they don't deserve a penny!Why the hell give them pocket money!?So that they buy a machete at the bazaar and assault Poles!? Fuck, no.They don't know what culture is or what our customs are, so piss off (R_FK).
In the debate on families with children, users expressed the opinion that families deserve social welfare because they are taking care of children as the following statements show: 'welfare [is] for those who have the courage to have many children' (F_FK), and those who 'are not too comfortable like those without children' should receive it (F_GW) or 'contribute to a beautiful idea of family with children' (F_PL).It was also emphasised that childrearing is    difficult and should be acknowledged as proper work and thus makes parents deserving of social welfare.
In both debates there were arguments that beneficiaries need to reciprocate by contributing to a common pool of resources with their work.In the debate on families, divergent points of view and active disagreements occurred as to whether the work of one parent is sufficient, or if both parents should work to deserve a benefit.Here, a group of arguments included the idea that the beneficiary needs to pay taxes to be deemed deserving.As concerned refugees, the logic of this argument was the same as with the need to work.However, families with children 'paying taxes' took on a different meaning and was close to the plea for universalism.Namely, it was argued that all Poles pay (some) taxes, so they are deserving of family benefits.Accordingly, some users expressed the idea that wealthy families are especially deserving of benefits because they pay high taxes.
The manifestation of reciprocity which Facebook users expected exclusively from refugees was that their skills be useful for the economy and their entrepreneurship.In the debate on family benefits, in contrast, it was childbearing which was broadly understood as a sign that an individual deserves welfare. Users frequently argued that if someone has  or  children, she ( : : : ) simply deserves that [benefit] (F_FK).
An argument about family benefit recipients' non-deservingness was their criticising the government who granted it.Users argued that those who claim government assistance shouldn't protest against it or insult members of the government.
A significant share of identity claims in both debates worked according to a binary logic which may be summarized as follows: All Poles are deserving of social welfare because of being 'us' and no refugees are deserving of welfare because they are alien.Apart from this radical understanding of identity criterion, in the debate on refugees users gave examples of cases which drive up refugees' deservingness.They may be placed on an axis of both geographical and cultural-religious closeness.Thus, those with Polish origins, Ukrainian citizens or Catholics were found more deserving than Muslims.It was frequently argued that there are also Christians among those people, so we should help them, if anyone (R_GW).
Although overall, identity concerns resulted in deeming Poles highly deserving of social welfare, in some claims it was argued that those who deserve social welfare the most are the 'good' Polish families who adhere to traditional values and cultivate 'Polishness.'It was often spoken here of 'all our Polish       children' as a 'national treasure' which should be taken care of.This deservingness logic, focused on othering those who don't adhere to traditional values, is present in the following comment: : : : pathology is the family without children and sense of development, not even of Poland but the whole white nation. Such children-free families are a threat to our future generations (F_GW).
The concrete meaning of the criterion need differs significantly between the two debates.Facebook users are unanimous that disability is a manifestation of neediness, which undoubtedly makes a person deserving of social welfare, as does poverty.The latter is, however, understood differently when speaking of refugees and Polish families with children.Above all, users often expressed opinions that Poles' needs are by definition more important than those of foreigners.The identity criterion is thus often framed as a need criterion which results in either an assumption that 'refugees may never be deemed deserving until Poles' needs are met,' or 'refugees may never have easier access to social welfare than Poles have.'The following exemplify these points of view.
Poor Poles deserve help first (R_FK), People in Poland are starving and they're gonna get houses and benefits?I think that's not fair!(R_GW).
Facebook users also indicated that sociodemographic status and the ability to work were proof of refugees' neediness.Thus, women and children were deemed needy, whereas users claimed frequently that 'young, healthy men' are not needy by any means.Their lack of need was described with such expressions as, 'they would make it there' (R_GW), or 'they look well-nourished and not very tired' (R_FK).Other signs that refugees were not in need included the fact that they managed to come to Europe and the suspicion that they use forged documents.According to numerous commenters, the former speaks to refugees being in fact well-off while the latter shows that they want to hide the fact of not having escaped from the cruelty of war.In contrast to the arguments for refugees' non-deservingness, a small number of comments indicated that 'war refugees do deserve help' (R_RP).
In the debate on family policy, commenters discussed such proof of deservingness as a family's sociodemographic status, family income and experiencing general difficulties.It was argued that single parents and parents of children with disabilities or illnesses are clearly in need and thus deserve social welfare.Users disagreed as to whether financial need means that only the poorest, whose income is means-tested, are deserving or whether all but the richest deserve social welfare.A group of indicators of Poles' deservingness was focused on their low wages, the high cost of living and rising prices in the country.Moreover, children having various needs which must be fulfilled also fell into the general    category of need expressed by Facebook users.Thus, all parents who finance some of their children's needs from family benefits were recognized as deserving of social welfare: Families really need this money!If they buy a car with it, then great, they will need it to take children to the doctor and you make a scandal out of that!(F_FK).
Finally, neediness was frequently associated with collective deprivation.Here, users emphasised that those who deserve social welfare were systematically neglected by previous governments.It was often claimed that the Civic Platform party systematically took advantage of Polish families.However, these claims go further than merely legitimizing a programme which should have been introduced earlier.They suggest that members of the target population deserve to be repaid both in moral and economic terms.

Conclusions and discussion
The goal of the study was to analyse the content of opinions on public deservingness appearing in Facebook debates about the social welfare granted to refugees and families with children.This was done in Poland, which employs a hybrid welfare state model and has witnessed contentious identity politics (Lendvai-Bainton and Szelewa, ), potentially increasing the complexity and polarisation of opinions on deservingness (Gugushvili and van Oorschot, ).This analysis contributes to a growing body of qualitative research about how in practice people make deservingness claims (Herke, ; Heuer and Zimmermann, ; Laenen, ; Nielsen et al., ; Osipovic, ).
The study sought to answer, first, how deservingness opinions are factually formulated in online debates.Overall, opinions in both debates turned out to be highly complex.In contrast to Heuer and Zimmermann (, p. ), who found only some CARIN criteria being applied to migrants and families, in my study users referred to all CARIN when evaluating the deservingness of migrants and families with children.This is consistent with assumptions about the role of a multi-layered welfare state regime and contentious identity politics.
I found that deservingness criteria coexist in public debates in two ways.The first is based on the construction of a 'most deserving beneficiary'.This followed a pattern described in earlier works, e.g. by Buss ().The second way encompasses conflicts between major ideological stances.The first mechanism was more present in debates on refugees, which speaks to users' tendency to scrutinize refugees more exhaustively than families when talking about welfare programmes.However, both debates do use this mechanism.Powerful hybrids of deservingness criteria are also used in both debates.Many of them are identical to those found by other researchers, such as 'the deserving refugee' (Nielsen et al., ) or 'lazy, non-working mother living off welfare' (Herke,       ).My study also revealed other amalgams, such as the figure of 'Poles' needs taking priority,' which turned out to be a salient mix of identity and need criteria, the application of which results in the full denial of refugees' deservingness.
The findings demonstrate significant differences in specific understandings of all deservingness criteria.Many were consistent with the findings of other researchers.In the case of refugees, these included i.e. cultural performance, adjustment and learning focused on integration within an attitude criterion (Zakariás and Feischmidt, ), payments, functional contributions and usefulness on the labour market within the criterion of reciprocity (Nielsen et al., ).In the case of families, the understanding included a lack of welfare fraud and gratitude (Herke, ), beneficiaries' past and potential future contributions in reciprocity (Laenen et al., ; Daly, ), and financial needs, health needs, number of children or experiencing high costs of children's' upbringing in general in need criterion (ibid.).
My findings point to two important aspects of the content categories.First, some content categories are contradictoryfor example, attitudes towards the mothers of toddlers.According to some users, proof of an appealing attitude in a mother is her desire to stay home, while others sought to ascertain that she worked.This shows that individuals' opinions differ not only in terms of deservingness perceptions and valuations (Laenen, ) but also in regard to what exactly they understand as good attitude (control, reciprocity, etc.).Second, content categories vary in terms of exclusiveness.For instance, having any child-related costs to be covered, being not very rich, and absolute poverty were all understandings of neediness present in the family debate.
The second research question concerned similarities and differences in the content of CARIN criteria with regard to refugees and families.Categories including work, tax paying, not being overdemanding in claims for welfare and not wasting or misusing help once it is granted were identical in both debates.However, the exclusiveness of some content categories differentiated strongly in both debates.Numerous deservingness constructions in the refugee debate implied logically empty sets of beneficiaries.This applied to refugees having to know Polish very well, their coming to Europe or using supposedly forged documents being indicators of a lack of neediness or having to wait until all Poles' needs are satisfied before claiming help.
In contrast, where families with children were concerned, although many content categories were framed as reciprocity or need, many were close to universalism.This was the case for proof of deservingness such as: being a Polish child, paying taxes, making expenditures, experiencing hardships in life, or the collective being neglected by past policies.My analysis shows that under the label of need in particular there are different understandings of deservingness, which differ strongly between refugees and families in terms of their exclusiveness.

  
The third research question this article addressed was how such aspects of debate context as socio-political cleavages and past welfare policies are referred to in opinions on deservingness.The study revealed that apart from the implications of the context mentioned, there were also explicit references to current and past policies, linked to control and need.For example, users claimed that beneficiaries are deserving because the state had failed to guarantee them a certain standard of living and thus it needs to correct its own mistakes.This logic is consistent with some assumptions about the communist legacy in deservingness opinions (Gugushvili and van Oorschot, ).Secondly, specific contextrelated understanding of need assumed that the proof of families' neediness was their being overlooked in recent redistribution and recognition state policies.This finding suggests the need means something collective, relative and prone to populist discursive identity politics (Lendvai-Bainton and Szelewa, ) within public deservingness opinions.It also suggests there are links to existing policies that go beyond the influence of the welfare state model (van Oorschot, ).
In terms of the limitations of this analysis, its scope is restricted to a single country, in which specific target groups are comparedone whose deservingness is tangled with broader migration, integration and security policies and the other to whom universalist approaches are widely applied in Europe.Furthermore, an ethical and methodological issue here is the risk that the voices of socially excluded groupsin particular, those of refugees and parents who lack time, skills or other resourcesare absent in both debates.This is plausible, given the evidence (McGregor , p. ) that social media users are not representative of any larger publicthey are more partisan, polarized and uncivil, which is further aggravated by mechanisms of social media targeting (Batorski and Grzywińska, ).Finally, the literature on deservingness understandings held by welfare users (Lavee, ; Herbst-Debby, ; Kremer, ) reminds us that opinions and narratives present in public discourse are just a repertoire of stances out of which individuals form more complex and nuanced understandings of deservingness.These are beyond the scope of this study.
Overall, the study prompts the following research questions for further analyses.Do changes in the content of CARIN criteria shared by the general public impact perceptions of the target groups' deservingness?If so, how?How do political and media framing of target groups' public images influence the content of deservingness criteria?Might the content of deservingness criteria, in turn, better explain the legitimacy of programs?For instance, if the specific criterion of deservingness is not wasting goods, is the program perceived as better if it entails more control measures over how benefits are spent?This study opens up a space for a more fine-grained analysis of relations between political framing, deservingness opinions and programme legitimacy.
     

TABLE  .
The Salience of CARIN Criteria in Both Debates

TABLE  .
The Content Categories of CARIN Criteria