Stigma consciousness among the unemployed and prejudices against them: development of two scales for the 7th wave of the panel study “Labour Market and Social Security (PASS)”

Abstract

This paper develops scales to measure stigma consciousness in the unemployed and prejudices against them, using a quantitative survey. The requirement for these scales is illustrated in the introduction. The theoretical substantiation of the scale development includes a definition of stigmatisation, an overview of the current state of research regarding the consequences of stigmatisation for individuals, and the possible causes of stigmatisation processes. The argumentation is based on reflections on the significance of employment in people’s lives. How the prejudice and the stigmatisation scales are developed is the subject-matter of the main section. Using data from a pre-test of both scales, the final version of the scales is then carved out. We conclude with a summary of our theoretical considerations and some thoughts about possible applications for the scales.

Zusammenfassung

In diesem Beitrag werden Skalen zur Messung von Stigmatisierungsbewusstsein bei Arbeitslosen und zu Vorurteilen gegenüber Arbeitslosen entwickelt. In der Einleitung wird die Relevanz dieser beiden Skalen begründet. Die theoretische Begründung der Skalenentwicklung beinhaltet Definitionen von Stigmatisierungsprozessen, einen Überblick über den Forschungsstand zu den Auswirkungen von Stigmatisierung für Individuen und die möglichen Ursachen von Stigmatisierungsprozessen. Die Argumentation basiert auf Überlegungen zur Bedeutung von Erwerbsarbeit im Leben von Menschen. Wie die Stigmatisierungs- und Vorurteilsskalen entwickelt werden, ist der Gegenstand des Hauptteils des Beitrags. Mit Daten aus dem Pretest wird die endgültige Fassung der Skalen herausgearbeitet. Wir schließen mit einer zusammenfassenden Darstellung der theoretischen Betrachtungen und einigen Vorschlägen zu möglichen Forschungsfragen, die mit diesen Skalen untersucht werden könnten.

“I would say that sometimes I dress now not so extremely, as now, hmm, how I should say?… Like Hartz IV …. But I dress so that I do not catch anyone’s eyes, only because I am Hartz IV….”

“My neighbour constantly says to me ‘antisocial crap’ and tells me that my daughter is also becoming like me. Then I said to her that I have passed a vocational training too, and she looked very surprised.”

“If one has no job, then people think that one is just hanging around at home, sitting there with no desire to do anything.”

“There are many misconceptions: Anyone can find employment, if he wants to. I mean, there are thousands of prejudices: the unemployed are lazy, sit all day in front of the goggle-box, have too much money anyway and have far too many children, and also that you only have to put in a bit of effort and you will find the right job”

1 Introduction or Getting the idea

The call for questionnaire modules from the labour market and social security panel team (PASS) reached our mail boxes while the above-quoted paragraphs were still fresh in our minds. Joining the two together we came up with the idea of developing a module for PASS to measure the awareness among the unemployed of being the object of prejudices. The quotes originate from a small research project conducted in Kiel. First designated as an applied scheme, the topics studied in this qualitative research project were experiences of discrimination, social marginalisation, occupational and biographical prospects and the reliance by unemployed people or those within labour market schemes on institutions such as the Federal Employment Office.¹ Almost all of the participants in the qualitative survey reported various experiences of being a target of negative ascriptions regarding their inferior status. Three-quarters of the interviewees described situations in their daily lives when their actions had been driven by fear that they would otherwise confirm a negative stereotype of the unemployed. The statements refer to two critical and important dimensions of social stigma. First, visibility: if the stigma is visible, others could use it as a basis for condemnation. It is not so much important that the stigmatised property such as gender or age is really observable, but rather that people believe that others perceive it. Second, from the interviews, controllability seems to be very important. Individuals who have a stigmatised characteristic that is believed to be controllable are more disliked, rejected and harshly treated than those whose stigma is perceived as an ascribed characteristic such as gender or ethnicity. Most of the people in our sample saw their situation as not controllable, although they did not believe that others shared this opinion. Therefore, they were quite clear that the treatment they receive from others reflects prejudice and discrimination. Furthermore our preliminary findings illustrate that people obviously vary in the extent to which they refer to their stereotyped status depending on the situation and their personal characteristics. For example, we have evidence from qualitative data that the younger the unemployed persons and the shorter the period of unemployment, the more often they negotiate and correct the stereotypical views of their counterpart. But we did not find quantitative empirical data to analyse the incidence of stigma consciousness and the impact of stigma-induced stress among several subgroups.

Interpreting the data we discovered several patterns for coping with prejudices. Some of the subjects in fact became quite clever strategists in managing these prejudice-induced situations. Almost all of them were able in specific situations to distance themselves from the negative ascriptions associated with the role of an unemployed person. Their behaviour ranges from denial, distraction and retraction to attempts to negotiate the negative ascription in every situation it occurs. Apart from these rather ideographic results some questions emerged that could not be answered at this stage of research concerning the incidence of stigma consciousness among the unemployed and its effects on people’s lives. Overlooking the pertinent literature regarding these phenomena, there were indeed some ways to classify the patterns of behaviour we observed. Erving Goffman’s (1963) theory of stigmatisation seemed to be especially helpful. But in general, the state of research regarding perceptions of being excluded and stigmatised is lacking in theoretical reasoning and, even more so, in empirical substance. At present it seems unclear whether sociological theory has already achieved agreement and mutual consent for further consideration concerning these matters. But despite these important conceptual questions, our findings and hypothetical assumptions so far have still been explorative and several open questions remain: Is there a theoretical and empirical way to connect the collective representations of these people to their action strategies in daily life? Are there typical patterns of behaviour when coping with the threat of stigma? Is there a variation in the extent to which unemployed people refer to their stereotyped status? By developing and implementing a scale to measure stigma consciousness among the unemployed these questions can be answered.

On the other hand, one may ask whether these feelings and reactions mirror actual existing prejudices among other people. What negative attributes towards the unemployed prevail in society? What kind of negative characteristics are attributed to the unemployed? How are they distributed across the different groups of people in paid work or in education? How widespread are stereotypical beliefs about the unemployed among others? Therefore, to find some answers to these questions, we decided also to investigate these complementary views on the phenomenon.

As “stigma is a key component of processes of social exclusion” (Reutter et al. 2009: 298), the PASS survey seemed to be an appropriate place for gaining some new and substantial knowledge on many of these questions. Therefore we have proposed the inclusion of new scales in the survey to measure the concepts “stigma consciousness of the unemployed” and “prejudices against the unemployed”. In the following we constructed a scale to measure stigma consciousness and prejudices using a quantitative survey.

We first summarise the theoretical framework of the scale construction. We begin with a definition of stigmatisation. This is followed by an overview of the current state of research regarding the consequences of stigmatisation for individuals. We focus on stigmatisation in general because of any characteristics and in particular because of unemployment. A detailed report is given regarding the possible causes of stigmatisation processes as discussed in the literature. In the course of that section it is necessary to reflect the significance of employment in people’s lives and the social context within which negative attributions and deprivation take place.

In the empirical part of the paper, the development of the prejudice and stigmatisation scales are presented. The paragraph aims to offer a profound reflection of the difficulties we encountered in our research. The process of constructing the scales, applying methods of factor and correlation analysis, is reported in detail to provide a suitable outline for similar research projects. We show data from a pre-test of both scales and introduce the final scales that were implemented in PASS. The paper summarises the theoretical considerations by formulating a heuristic model of stigmatisation and concludes with some prospects concerning future applications of the scales.

2 Stigmatisation

2.1 Definition of stigmatisation

There have already been some investigations into the negative impact of being a member of a stigmatised group in other domains using different methods, in different settings, with different purposes and with partly surprising and contradictory results. Most of these studies on stigmatisation refer to the vague but widely prevalent definition by Goffman. According to him, stigma is an “attribute that is deeply discrediting” and that reduces the bearer of this attribute “from a whole and usual person to a tainted, discounted one” (Goffman 1963: 3). Besides some studies that aim to identify the various dimensions of the concept in a phenomenological way, a variety of definitions of stigma have been articulated. Stafford and Scott (1986) stated: “One of the curious features of literature concerning stigma is the variability that exists in the definition of the concept” (quoted in Link and Phelan 2001). Particularly with regard to the search for a conceptual framework, Link and Phelan (2001: 363–385), as well as Major and O’Brien (2005: 393–421), contribute—at least at a first glance—some interesting considerations for conceptualising stigma. For Link and Phelan, stigma exists “when elements of labelling, stereotyping, separation, status loss, and discrimination occur together in a power situation that allows them” (Link and Phelan 2001: 377). But in particular, the inclusion of features such as power, discrimination and status loss raises some new questions, especially since the authors represent aspects such as loss of status and discrimination as it turns out (Link and Phelan 2001: 378 ff). It is also currently unclear how extensive the discrediting effect of a negative attribute has to be to actually speak of a stigma. Although we think that Link and Phelan’s attempt to conceptualise this phenomenon is generally helpful, we do not share their self-praise that “the value of their definition rests in its full utility” (Link and Phelan 2001: 377). In our view, some important theoretical work still needs to be done on preliminary conceptual clarifications regarding the genesis and use of the concept.

2.2 Effects of stigmatisation

However, aside from these conceptual ambiguities, according to Crocker et al.—in very general terms—“the people who are stigmatised tend to experience more negative outcomes in their work lives and in their personal lives than do the non-stigmatised” (Crocker et al. 1998: 521). A review written by Major and O’Brien (2005) shows that stigma and its perception affect self-esteem and performance in the education system, and have negative effects on health (see also Schofield 2006; Tausendpfund 2005). They conclude that “members of stigmatised groups are discriminated against in the housing market, workplace, educational settings, health care, and the criminal justice system” (Major and O’Brien 2005: 396). Link and Phelan (2001) concede that results from discrimination may, in turn, engender further discrimination. Major and O’Brien (2005: 396) describe different adverse effects for stigmatised persons. This involves a negative impact of disadvantage and discrimination for members of stigmatised groups in the housing market, in education, in the justice system and in the area of health care, and can also be seen even in family contexts.

Up to now, only a few studies have examined whether stereotypes exist towards the unemployed. These studies, in contrast to our contribution, are more focused on what is called the ‘stigma effect’ and rely—besides an explanation by heterogeneity—on “work history explanation” (Omori 1997: 394) for longer unemployment. Vishwanath (1989: 488) states that this worker employment history effect “…generally means that a firm is less inclined to hire a worker with longer unemployment duration.” They raise the question of whether the unemployed face systematically lower chances of being hired because of the longer duration of their unemployment. In these terms firms interpret a longer duration of unemployment as a negative signal for their productivity. Omori’s findings (1997: 414) support “the hypothesis that stigma effects and heterogeneity explain the observed relationship between past and future unemployment durations” (Omori 1997: 414). Furthermore, the stigma effect varies with the local unemployment rate. People who experience unemployment during a time of high unemployment suffer less from stigma than people who are unemployed when only a small number of people experience unemployment.²

2.3 Stigma-consciousness and its effects

When focusing on the consequences of stigma on self-esteem Crocker and Major came to the conclusion “that prejudices against members of stigmatised and oppressed groups do not result in lower self-esteem” (Crocker et al. 1998: 531). Other research surprisingly suggests that members of stigmatised groups report low levels of personal experience of discrimination. Some stigmatised individuals seem to be reluctant to attribute negative outcomes to prejudices and discriminations. They believe in the controllability of the stigma itself and negate the intensity of social comparisons with other individuals and groups. To cover these differences, Elizabeth Pinel (1999) developed and validated a 10-item ‘Stigma Consciousness Questionnaire’ (SCQ) for several stereotyped groups (e.g. women, lesbians, gay men, African Americans, Latinos/Latinas, etc.). According to her, stigma consciousness reflects the extent to which individual targets of specific stereotypes “focus on their stereotyped status and believe it pervades their life experiences” (Pinel et al. 2005: 482). Stigma consciousness negatively correlates with disengagement in school and self-esteem. Pinel (2002) shows that women who believe that their male counterparts have sexist attitudes behave more critically and thus also received appropriate responses that confirm the perceived prejudices. Furthermore, Pinel et al. (2005) discovered that increases in stigma consciousness have a negative impact on stigmatised students’ abilities to succeed in college. Studies concerning the educational system show significant performance gaps between academically stigmatised (African Americans and Latinos/Latinas) and non-stigmatised (Whites and Asian Americans) students. Members of stigmatised and non-stigmatised groups differ substantially in measures of academic achievement (Major and O’Brien 2005: 408). Sidanius and Pratto (1999) indicate that children who are members of ethnically stigmatised groups receive a smaller proportion of public education funding than children who are not members of these groups. The practice of funding school districts through property taxes perpetuates educational disparities between whites and stigmatised ethnic minority groups in the United States (Sidanius and Pratto 1999). “Consequently, the schools attended by the children of dominants are generally much better equipped, are in better physical condition, are staffed by more experienced, more competent, and better-paid teachers, have smaller class sizes…” (Sidanius and Pratto 1999: 182).³ In a review on health-related stigma, van Brakel (2006) examined which discriminations are related to health, how stigma is perceived, which discriminatory or stigmatising practices can be found in health care, media or educational material and how they affect people. The effects of health-related stigma, which appear in the analysed studies, are varying and include different areas, e.g. starting a family, personal relationships, employment, education, mobility, leisure activities and participation in social and religious functions. For example, depression seems to be more prevalent among members of stigmatised groups. Crocker argues that these group differences often weaken if the socioeconomic status is controlled (Crocker et al. 1998: 532).

Despite these findings some questions remained: What kind of negative evaluations regarding the unemployed are widely shared and well known among members of society? Is there a highly salient difference between the employed and the unemployed, as the sociological contributions dealing with the public and media debates imply?

In our view many contributions in this specific area neglect the inevitable question of the genesis and distribution of negative attributes associated with an inferior status. So the next chapter will contribute some general thoughts regarding the given semantics of inequality, classifications and hierarchies.

2.4 Where does stigma come from?

Sidanius et al. (2004) addressed these concerns with their “arbitrary-set distinctions”: Group-based hierarchies and status differences exist within societies. They are created on the basis of salient characteristics along specific affiliations (class, race, religion). The selectivity, the genesis and the interaction between these systems still remain open. So the authors concede (2004: 871) that a “more detailed understanding of alternative forms of gender, arbitrary set, and adult-child relations and how these systems intersect is needed.” Some of the authors dealing with stigma refer to specific functions of a social stigma. First: by stigmatising or militating against others in out-groups, persons in the in-group can enhance their personal self-esteem; second: according to Tajfel (1982), derogating others is executed to maintain a positive social or collective identity; and third: according to social dominance theory (see Sidanius et al. 2004), derogating and stigmatising others is a strategy to legitimate the advantages of a group of people with a higher status. Thus this helps to preserve and maintain existing inequalities between groups. It enables the privileged people to believe that they earn their privilege, that the system is fair, and that people get what they deserve and deserve what they get. One of these “hierarchy-legitimising myths” in modern societies is that they are widely seen as meritocratic systems, where talent and real effort work out perfectly. This system-justifying ideology, which promotes the superiority of one group versus another, attributes unemployment and poverty to a lack of merit.

Stigma refers to some attributes that, in a given social context can lead to devaluation. In order to find proper social representations, patterns of interpretation and/or negative attributions towards the unemployed, and to use them for the construction of the prejudice scale, it was necessary to take a more intensive look at the literature on this topic to extract the main ideas.

Within the current situation of a de- and reconstruction of the German welfare state institutions and the European crisis scenario, the risks of social exclusion are discussed intensively in mass media and in various, quite unconnected, social scientific discourses. The issues debated include equal opportunities of social participation, the difficulties of equitable distribution opportunities, the uprising menace of a persistent exclusion of socially disadvantaged people, and the consequences for the labour market and corporate policies of organisations. Most of these studies identify the problematic labour market integration as a central theme cautioning against challenging consequences for individuals and for social and political cohesion. The smallest consensus beyond the work which has been done in this field is the crucial impact of employment. For a long time and for the majority of the scientists in this field, employment has not been simply a “threshold of respectability” (according to Bourdieu and Sennet, cf. Bescherer et al. 2009: 155). In modern industrialised societies, employment is the key mode of social integration (Durkheim 1977). Employment acts, at least indirectly, as a central socialisation instance (Kronauer 2002: 156). We assume that Germany is characterised by a deeply meritocratic social structure, where social esteem (Honneth 1994; Kaletta 2008) is rewarded by the principles of performance fairness. For people who are temporarily or permanently unable to fulfil the varied requirements of the labour market the situation is linked with several negative outcomes. In a society where skills and accomplishments are permanently subjected to some kind of scrutiny, an inferior social position due to lacking access to jobs can be connected to painful social experiences. This has an impact on the self-confidence of the jobless and their social interactions with others. Uske (1995), Oschmiansky (2003), Neckel (2006), Kessl et al. (2007) and Chassé (2010) refer to the downside of the public debate about unemployment in Germany and make an issue of the existence of specific signs of disapproval towards the unemployed. These contributions gave us the opportunity to find items for the prejudice scale. According to them, it is not just a glorification of work that takes place. In particular, signs of public contempt and negative stereotypes regarding the unemployed are expressed in different discourses. By increasing the importance of paid work, those who are unemployed will be excluded from many areas of social life (Bonß 2006). The long-term unemployed are particularly affected, because the longer they are out of work, the greater is the likelihood of being stigmatised as too lazy or incompetent. In Germany, the categorisation “unemployed/employed” is, according to Barlösius (2002), an effective line of demarcation and thus part of collective representations. In this context, Bourdieu (1992: 146) points out that there is a hierarchy of individuals and groups according to their specific value and utility for society (see also: Blau 1977: 6 ff; Berger 1988: 512; Neckel 2003: 163 ff). Furthermore, he argues that our society is organised, structured and constantly renegotiated under the logic of the “differential distance” (Bourdieu 2006: 359). He (1992: 103) assumes that actors experience social inequalities mediated by images, classifications and categories. By taking account of social categories they (even unconsciously) reproduce them. With reference to Goffman he states that there is “a sense of one’s place” (Bourdieu 1985: 17). Social hierarchies are—to a certain extent unconsciously—reproduced and perpetuated via the restrictions, differences, opportunities and distances between them and other social positions. Additionally, pejorative characterisations of specific social groups are spread by mass media (van Brakel 2006). Thus, stigmatisation has a tangible impact on the life of those concerned. One challenging question is: Although there is an enormous variability within the group “unemployed”—and the designations do not apply to all unemployed—why do depreciated traits matter in social interactions? Unemployment is a given category, a social label that is linked to several negative connotations. Those, in particular, who are concerned by unemployment, seem to know about the negative attributes that might be attributed to them.

In this initial phase of research we developed a brief theoretical frame of reference to point the way for further theoretical considerations, for generating items and formulating appropriate hypotheses (sensitising concepts).

3 The scales

3.1 Modification and development of the stigma-consciousness scale

The initial idea was the application of a 10-item scale of stigma consciousness developed by Elizabeth Pinel (1999) for our purposes. To apply this scale in a different context it needs to be modified in order to meet the specific purpose of the investigation and the assumed traits of the social group in focus (Brown and Pinel 2003: 627 or Mosley and Rosenberg 2007). “To customize the questionnaire for use with a particular group, one must only change the appropriate in-group and out-group to which each item refers” (Mosley and Rosenberg 2007: 87).⁴

With Elizabeth Pinel’s permission we translated and modified this scale for the group “unemployed”. In discussions about these items with colleagues from the Research Department E3—Panel Study ‘Labour Market and Social Security’ at the Institute for Employment Research—we were confronted with a lot of objections concerning the desirable qualities of the translated items. Some of the wording problems occurred because the German language often requires longer expressions than does the English original. The items were not as simple and straightforward as necessary and would have led to misunderstandings, would probably have annoyed respondents and would have made more instructions necessary. In the translated version of this adapted scale there were too many vague words. Phrases were not sensitive enough, and most importantly, seemed inappropriate for the intended target audience.⁵ So, we had to modify these items so that they matched our disposition and expectations. A four-point Likert scale was applied to adjust to the usual response style in the survey. In the original version the Likert scale ranges from 0 to 6 thus giving a neutral midpoint of 3 with ‘neither agree nor disagree’. Since it was not our aim to compare the strength of stereotypes against the different social groups explored in Pinel’s studies (gays, women …) we judged the similarity of scaling to be less important than fitting into the surrounding questionnaire. The new items were evaluated using cognitive interviews with students at the University of Kiel. In particular, we used common probing techniques asking the students whether they as respondents could possibly have any difficulties understanding the meaning of the question or particular words, whether they could potentially have a different understanding of what the question refers to and whether they could have any difficulty choosing an answer. Furthermore for some selected issues, it seemed more appropriate to use some specific probing techniques (Prüfer and Rexroth 1996; Rothgeb et al. 2007). To identify problems with the comprehension of survey items we used the “concurrent think aloud technique” in which the respondent verbalises thoughts while going through the scale and “paraphrasing techniques” whereat respondents are asked to repeat the items in their own words. With the results of these probing techniques we reduced or adjusted ambiguous items that were coded as unclear, too long or complex, that were classified as using an awkward syntax, having vague terms or requiring further details and information for the respondent.

For the second part of the stigma consciousness scale dealing with the strategies and beliefs regarding stigma, we started to collect items which covered nearly all of the above-mentioned topics in answer to stigmatisation. At this stage, we had to ensure that all strategies of coping with stigmatisation would be included. To avoid what Mohler calls an “item-tinkering exercise” (Mohler 2006: 10), and to clarify our decisions regarding the application of appropriate items, we used some steps of the facet theory (Borg and Mohler 1993; Borg and Shye 1995) using the mapping sentence technique. We got the first approximation of the right items through a simple adaption of this approach for the purpose of clarifying the different items. The mapping sentence should distinguish between the universes of persons, reactants, situations and different responses. Each item was classified according to these dimensions. Using this technique new items can be developed by seeking other possible values of each dimension or by rearranging combinations. Additionally, redundant items may be identified. So we were able to rule out some of the items on logical reasons. An illustration of the mapping sentence technique is provided in Table 1.

Table 1 Examples for mapping sentence technique applied to beliefs and responses to prejudices and negative stereotypes

Bearing in mind that this was a strictly qualitative judgement and that it had to be supported with objective data, we picked out 13 items for the pre-test procedure which were related to the underlying theoretical construct.

3.2 Development of prejudice against the unemployed scale

For the “prejudice against the unemployed” scale, we tried to crystallise the meaning regarding these issues from the sociological literature already summarised above. For item phrasing we collected prejudices from the literature. Using a discourse analysis Uske (1995) has identified a so called “quitter discourse”. As part of this kind of ideology stigmatisation of the unemployed as skiving off work becomes obvious. The unemployed are often incriminated as unwilling to work, as slackers or lazy fainéants. For example this pattern of interpretation was placed in Item C of the pre-test scale: “What do you guess; how many of 100 unemployed are unwilling to work?” We searched for patterns of pejorative attributions, collected them and phrased several questions. At this stage we assumed that these items would cause many problems and so we had several productive meetings and frequent correspondence with colleagues from the Institute of Labour Market Research (IAB) and took the time to discuss the appropriate measurement strategy and wording of the items.

3.3 Pre-testing of both scales

Pre-test interviews were conducted in spring 2012. Due to limited resources, only a restricted set of items and other scales for validating could be used in the pre-test. Besides the newly developed scales of stigma consciousness of the unemployed and prejudices against unemployed, the questionnaire included scales of the awareness of social exclusion and sense of coherence to determine convergent and discriminant validity. The participants were recruited from among the visitors of several job centres in Germany. The pre-test questionnaire was answered by 178 respondents. Seven interviewers carried out at least ten and at most 35 interviews. The mean duration of the interviews was 15.8 minutes. The interviewer reported that the readiness to answer the questions was almost always good (only one out of 178 interviewees answered with moderate readiness). A tabulation of the employment status shows that 121 interviewees are unemployed and 57 persons are employed, in education or retired. Thus, 121 persons were filtered to the questions about stigma consciousness, social exclusion and sense of coherence, and 57 persons were asked about their prejudices against unemployed.

3.3.1 Stigma consciousness scale

For each item in the new scales the interviewers were able to take a note in the case of a problem emerging and also to report what kind of problem this was. In Table 2 the missing values and the problems of the stigma consciousness items for the unemployed are summarised. The items have been split into two questions in the pre-test questionnaire, with experiences from interactions with the out-group (employed) and awareness of stigmatisation in the first question, and identification with in-group (unemployed) and strategies of action in the second question. We conducted the scale testing with all these items pooled.

Table 2 Stigma consciousness items: missing values and number of problems (N=121)

A first glance at the results for the stigma consciousness questions shows that some items have a couple of missing values and some caused more than a few problems in the interviews. Therefore, in a first attempt we excluded items that were problematic for these reasons, since the most important criterion is that all respondents should be able to understand and answer the questions. The items that are excluded are 1F, 1G, 2A, 2C und 2L. In the next step, a factor analysis was computed to derive the number of dimensions of the stigma consciousness scale that can be distinguished and to select the items that have the highest factor loadings on these factors. We used principal component analysis with varimax rotation after testing for several other specifications. In the literature principal component analysis is recommended for the purpose of data reduction whereas the principal factor method is more useful for confirmative data analysis (summarised in Wolff and Bacher 2010). Data reduction and extraction of dimensions was also the aim of our application. The number of components is derived from the number of eigenvalues larger than 1. We used varimax rotation to ensure orthogonality of the components, which is important in case the dimensions of the scale are used in regression models. The items with high uniqueness (i.e. little contribution to the factors) were excluded stepwise. These were the items 1B, 2H, 2J und 2K. The result was a factor model with five dimensions. However, one dimension consisted of only one item (1C), so we decided to exclude this item also. Additionally item 1A was left out because the wording of item 2F was almost identical. The results of the final version of the factor analyses are summarised in Table 3. This model now has four components. The fifth eigenvalue is only 0.66 indicating that the solution with four components is appropriate. The four components explain 74.3 % of variance showing that the model fits the data pretty well.

Table 3 Factor analysis stigma consciousness scale (N=104)

The final solution has four dimensions. The first factor deals with the social relations of employed people (maintaining relationships with the employed and feeling obliged to other unemployed people). The second dimension is the avoidance of incriminatory social situations (hiding unemployment, avoiding situations where prejudice could occur). Being under pressure to take action (everyday life is more difficult for the unemployed, feeling burdened by unemployment, trying to find a job as quickly as possible) constitutes the third dimension. The forth factor summarises two items regarding the awareness of the unemployed individual that she/he is affected by prejudices (thinking that people have prejudices, being affected by prejudices against unemployment). The Cronbach’s alpha values of all four dimensions are reported in the last row to test the internal consistency. With the exception of the forth factor the alpha values are sufficiently high even though so few items remained for each dimension.

Our initial theoretical point that distinguished between (a) experiences of interacting with the out-group, (b) presumed perception of the in-group by the out-group, (c) identification with the in-group and (d) strategies of action, is not reflected in these results. The reason is that particular experiences correspond to particular strategies, intensity of identification or kinds of presumed perception. Another reason is that the questions regarding what the unemployed think others are thinking about them are too complicated to get across. Therefore most of these items were skipped because they were problematic in the interviews. The correlations (Spearman’s rho) of the selected items of the scale are summarised in Table 4. The items are ordered along the dimension they belong to so that we see whether the items are highly correlated within each dimension and have a low correlation to items from other dimensions. All the within correlations are significant (shaded in grey). However, some of the between correlations are significant, too. Therefore, we conclude that the dimensions are not very sharply bounded.

Table 4 Correlations between stigma consciousness items (Spearman’s Rho)

Then we analyse convergent validity with sense of exclusion measured by a suitable question from the standard PASS questionnaire and a scale developed by Bude and Lantermann (2006). The PASS item of social exclusion reads as follows:

• Now we turn to some different questions. In society, you can have the feeling that you can take part and belong to it or you feel cut out. What about you? To what extent do you feel a part of society or do you in fact feel excluded? Please use the numbers 1 to 10 for rating …

  (Kommen wir jetzt zu ein paar anderen Fragen. Man kann das Gefühl haben, am gesellschaftlichen Leben teilzuhaben und dazuzugehören oder sich eher ausgeschlossen fühlen. Wie ist das bei Ihnen? Inwieweit fühlen Sie sich eher dazugehörig oder eher ausgeschlossen? Verwenden Sie zur Einstufung bitte die Zahlen von 1 bis 10 …)

The sense of exclusion scale by Bude and Lantermann (2006) originally had six items. The authors provided us with a new version with 11 items from which we selected three:

• I think that I have found my solid position in society. (Ich glaube, dass ich meinen festen Platz in der Gesellschaft gefunden habe.)

• I feel barred from my social environment. (Ich fühle mich von meinem sozialen Umfeld ausgeschlossen.)

• I know that I matter to my social environment. (Ich weiß, dass ich für die Menschen in meinem sozialen Umfeld wichtig bin.)

This scale is one-dimensional (proofed by factor analysis). The correlations between stigma consciousness factors and both exclusion scales are summarised in Table 5.

Table 5 Correlations between stigma consciousness factors and social exclusion scales (Spearman’s rho)

Since low values of the PASS exclusion scale represent exclusion and high values mean inclusion, the negative sign of the correlations to the stigma consciousness dimensions confirms the expected convergent validity. Two of the four correlations are (highly) significant; the other two are almost significant. A similar interpretation applies to the exclusion scale developed by Bude and Lantermann. The only difference is that the correlation of their scale to the factors 3 and 4 is absent.

Whether discriminant validity can be found between the stigma consciousness of the unemployed and a sense of coherence, is tested in the next table. To measure the sense of coherence the following scale was used (Schumacher et al. 2000):

• 1. I don’t really care about what is happening around me. (Es ist mir ziemlich gleichgültig, was um mich herum passiert.)

• 2. Often I cannot understand the behaviour of people around me. (Ich kann das Verhalten von Menschen, mit denen ich zu tun habe, oft nicht verstehen.)

• 3. I have often been disappointed by people on whom I actually counted. (Mich haben schon häufiger Menschen enttäuscht, auf die ich eigentlich gezählt habe.)

• 4. I often feel unfairly treated. (Ich habe oft das Gefühl, ungerecht behandelt zu werden.)

• 5. I have the impression that things I do every day have little meaning. (Ich habe das Gefühl, dass die Dinge, die ich täglich tue, wenig Sinn haben.)

• 6. In many situations I do not understand my own feelings. (In vielen Situationen verstehe ich meine eigenen Gefühle nicht.)

A factor analysis of this scale uncovers two dimensions: the second, third and fourth item load on the first factor displaying disappointment at others; the first, fifth and sixth load on a second factor measuring internal resignation. In Table 6 the correlation to the stigma consciousness factors is computed separately for both dimensions of the sense of coherence.

Table 6 Correlations between stigma consciousness factors and sense of coherence (Spearman’s rho)

The results show that both dimensions of the sense of coherence correlate significantly to the first factor of the stigma consciousness of the unemployed scale dealing with social relations. The second dimension “avoidance of situations” is positively correlated to the internal resignation dimension of sense of coherence. Obviously, both conceptions are independent from each other only with respect to the dimensions of pressure to act and awareness of prejudices, and partly in avoidance of situations. However, we do not interpret this result as a refutation of our scale. Both dimensions of the sense of coherence scale point to broken relations with others, so the correlation to the social relations dimension of stigma consciousness seems appropriate. The correlation of the second dimension of the sense of coherence that we named “internal resignation” with the avoidance of situations dimension of stigma consciousness is also easy to follow.

Finally, we analysed whether the scores of the stigma consciousness scale vary between subgroups. We found that the third dimension (pressure to act) is significantly higher for women. The awareness of prejudices (4th dimension) rises with age. The higher the income (household equivalence) the lower are difficulties in social relations with the employed and identification with the unemployed (1st dimension), and this is also true for the awareness of prejudices (4th dimension). Education, vocational training, nationality and marital status are not correlated to the stigma consciousness of the unemployed (in our small pre-test sample).

The final version of the scale that is included in the 7th wave of PASS reads as follows:

In everyday life you can observe different attitudes and behaviours of other people. Now I read some statements to you. Please tell me, whether the statements “fully apply”, “apply to a large extent”, “apply to a small extent” or “do not apply at all” to you personally?

Im alltäglichen Umgang mit anderen Menschen kann man verschiedene Einstellungen und Verhaltensweisen beobachten. Ich lese Ihnen nun einige Aussagen vor. Bitte sagen Sie mir, ob die folgenden Aussagen auf Sie persönlich „Voll und ganz zutreffen“, „Eher zutreffen“, „Eher nicht zutreffen“ oder „Überhaupt nicht zutreffen“?:

3.3.2 Prejudices against unemployed scale

The scale to evaluate existing prejudices against the unemployed applied the per cent rating method developed by Brigham (1971). He proposes asking questions about the perception of how many people perform the disapproved behaviour. This technique gives the respondent the impression that he/she is asked about empirical facts but not about their own prejudices. Thus the danger of social desirability in the answering behaviour should be lower. To quantify the degree of prejudice against the unemployed further considerations are necessary: Low values of the mean of each item indicate that the respondents on average think that fewer unemployed people behave in the disapproved way; high values indicate widespread attitudes about the unemployed. The highest mean value was computed for the item “How many of 100 unemployed people would find a job immediately if they were to search intensely and were ready to compromise?”, the lowest for the item “How many of 100 unemployed people get a higher unemployment benefit than they could earn by working?”.

To compute the absolute intensity of prejudices against the unemployed the sum (or mean) of all items could be computed for each respondent (using 100 minus the reported number for positively formulated items). In other cases a relative approach could be appropriate arguing that the mean of each item (see Table 7) can be interpreted as a benchmark of its stereotyping value. The individuals who think that a higher number of unemployed people—compared to the mean rating of all respondents—can be characterised by a statement have above average prejudices. The deviation from the mean can be used to derive a stereotyping score for each person. In cases where the item has a positive connotation the difference between individual and mean rating is multiplied by −1. By summarising the differences between individual ratings and mean values the stereotyping score can be computed for each respondent.

Table 7 Prejudices against unemployed items (N=57)

The scale was introduced by the following sentence: “In the following, we would like to talk about your experiences with unemployed people or what you can read or hear about them. Now, I will read some statements to you. You can rate your opinion from 0 to 100.” The items are specified in Table 7.

The table shows that there were very few missing values. A first factor analysis produced three factors. Item B was removed because the interviewers reported some more problems with this item. In the next step, item D was removed because of ambivalent factor loadings on the resulting two dimensions. Lastly, in order to obtain a one-dimensional scale the item G was also eliminated. The results of the one-dimensional scale are reported in Table 8. The explained variance is 50.6 %. The value of Cronbach’s alpha is 0.73.

Table 8 Factor loadings of the prejudices against the unemployed scale (N=55)

To illustrate how the items correlate to each other Table 9 is inserted here.

Table 9 Correlations of prejudices against unemployed items (Spearman’s rho)

Does the prejudice scale discriminate between subgroups? Although women have lower prejudice scale scores than men the difference is not significant. Age correlates negatively to prejudices against unemployed; that means that older persons have fewer prejudices. There are no differences for subgroups of different marital status or by household income. Those who have achieved high school graduation (Fachabitur or Abitur) have significantly fewer prejudices than people with a lower secondary education (Hauptschule or ohne Abschluss). The most obvious differences can be found between people with no vocational training and those with vocational training or college/university degree, with less prejudice in those with higher education. The scale in the PASS questionnaire is now phrased as follows:

In the following, we would like to talk about your experience with unemployed people or what you can read or hear about them. Please estimate for how many of 100 unemployed people the following situations apply. You can rate your answers with numbers from 0 “no unemployed person” to 100 “all unemployed persons”.

(Im Folgenden geht es um Ihre Erfahrungen mit Arbeitslosen oder was man über Arbeitslose liest oder hört. Schätzen Sie bitte jeweils, für wie viele von 100 Arbeitslosen treffen die nachfolgenden Situationen zu. Ihre Antworten können Sie mit Zahlen zwischen 0 „kein Arbeitsloser“ und 100 „alle Arbeitslosen“ abstufen.)

4 Conclusion

The above-mentioned contributions made quite clear that we know very little about the distribution of specific prejudicial attitudes and beliefs concerning these stereotypes. Therefore we propose a concise model based on theoretical concepts of appreciation and stigmatisation (Honneth 1994; Goffman 1963). This model assumes, through mediation by the elements of the self (Mead 1934), a relationship between social self-esteem, stigmatisation (enacted/perceived) and the (dis-) integration of people. The difference between enacted and perceived stigma has far-reaching consequences, especially on the micro level of action. The fact that somebody has already perceived or actually experiences discrimination or even merely fears being discriminated against makes some important differences both in actions and in social identity. With the limit of the scales in mind, we suspended the more detailed thoughts regarding the different situations, in which somebody is at risk of being negatively stereotyped. We assume that there are certain situations which are particularly suited to generating a sense of inferiority, where group differences in particular occur (e.g., degree of formalisation, hierarchy, formulation of normative claims). Crocker (1999: 91) states: “I argue here that the effects of stigma on the self are negotiated, created, and acted upon in the situation. In other words, self-worth, or the lack of it, in the stigmatised is not a stable, deep-seated personality characteristic. Rather, it emerges in the situation and is a function of the meaning given to that situation.”

Stigma consciousness is an experienced disregard of signs of mutual acceptance. As part of some kind of self-evaluation, this feeling will be associated with the perception of an individual’s own abilities and the percipience of a negative assessment in the social environment. The consequences of this self-evaluation can retroact back to the social context (Mead 1934: 202). That would mean that the social esteem is reduced or adjusted. In the absence of this correction, perceived disrespect will have an impact on the different dimensions of the self and also on the behaviour after the confrontation with the perceived rejection and exclusion. These feelings will increase as a function of past experiences. There are different possible consequences—that we focused on in our reflections—related to the construction of the different items. Recent stigma research following Goffman gave some attention to the variety of ways in which stigmatised people behave. In particular, Major and Eccleston (2004), along with Miller and Kaiser (2001), identify some overlapping patterns of responses to stigma. We used these suggestions to design some of the items of our pre-test scale. The spectrum of the reactions we found in our qualitative data set was also utilised. They ranged from social withdrawal, denial or reactions of avoidance of comparisons with non-stigmatised people, through to distraction, adaption or acceptance of the unfair treatment as not controllable and as a “fact of life” (Miller and Kaiser 2001: 82) until cognitive restructuring of certain situations took place.

This paper summarises theoretical rationales and the practical development of two scales: a “stigma consciousness of the unemployed scale” and a “prejudices against the unemployed scale” to be provided to others i.e. employed, students or inactive population. The newly developed scales are included in the seventh wave of the PASS survey. Data will be at the disposal of the scientific community for analysis not before 2014. The research can then inspect whether stigma consciousness and prejudices against the unemployed are directly connected to labour market-specific concepts such as the interactive-pragmatic employability framework (Promberger et al. 2008). Connections to the dimensions of working capacity are especially apparent. It will be interesting to analyse how the new scales relate to other issues of the PASS questionnaire, e.g. the responsibility of looking for a job, social participation or general attitudes. Some initial questions are: To what extent do the unemployed feel stigmatised and how many prejudices exist among others against the unemployed? Do long-term unemployed suffer more than short-term unemployed? Are groups with alternative role models resistant to prejudices? Do different social groups hold more or fewer prejudices against the unemployed?

Executive Summary

The paper develops scales to measure stigma consciousness in the unemployed and prejudices against them. Participants in a preceding qualitative survey reported various experiences of being a target of negative ascriptions regarding their inferior status. The panel study “Labour Market and Social Security (PASS)” seemed to be an appropriate place for gaining some new and substantial knowledge on questions like: What do unemployed think which kind of negative characteristics are attributed to them? What negative attributes towards the unemployed prevail in society? Therefore, we develop two scales whose subjects are on the one hand the perception of social exclusion and stigmatisation among unemployed, and on the other hand the (negative) evaluation of being unemployed among employees.

According to Goffman, stigma is an “attribute that is deeply discrediting” and that reduces the bearer of this attribute “from a whole and usual person to a tainted, discounted one” (Goffman 1963: 3). Stigmatisation takes place in situations where labelling, stereotyping, separation, devaluation and discrimination by others happen. Stigmatised people experience more negative outcomes in their working and personal lives, have lower self-esteem and suffer from negative effects on their educational attainment or health. Only a few studies have examined negative stereotypes towards the unemployed. The newly developed stigma consciousness scale reflects the extent to which individuals believe on their stereotyped status as unemployed and how much harmful impact they experience on their own life.

Discussing the origins of stigmatisation we among other things refer to theoretical considerations about group-based hierarchies and status differences that exist within societies. They are created on the basis of salient characteristics along specific affiliations like class, race, religion, or employment status as well. The social functions of stigma are manifold: militating against out-groups may enhance the personal self-esteem, help to maintain a positive social or collective identity and legitimate advantages of a group with higher status in order to preserve existing inequalities. As employment is the key mode of social integration missing access to a job can be connected to painful social experiences affecting the self-confidence and social interactions with others.

We constructed two scales to measure stigma consciousness and prejudices against unemployed. The starting point was a 10-item scale of stigma consciousness among women developed by Elizabeth Pinel (1999). Because these items turned out to be too vague, too complex or not sensitive enough after translation and adjusting them to our topic, we developed further items guided by the theoretical framework. For developing the items several techniques (cognitive interviews, mapping sentence technique) were applied. Additionally, for phrasing items we collected prejudices against unemployed from the corresponding literature.

The results reported are taken from 178 pretest interviews. The interviewers could take a note for each item in the case of a problem emerging and report what kind of problem this was. Items were selected on the basis of these reports and a factor analysis.

The factor analysis of the stigma consciousness scale resulted in four dimensions: social relations of unemployed people, avoidance of incriminatory situations, being under pressure to take action and awareness to be affected by prejudices. To determine convergent and discriminant validity of the scales, the questionnaire included scales of awareness of social exclusion and sense of coherence. The social exclusion scales both significantly correlate to the first and second dimension, the correlation to the third and fourth dimension is almost significant. The sense of coherence scale that was intended to derive discriminant validity has two dimensions in the pretest data: disappointment at others and internal resignation. It turned out that some dimensions of stigma consciousness and sense of coherence significantly correlated. We do not interpret this result as a refutation of our scale because the detailed results have some plausibility. Additionally, we analysed whether the scores of stigma consciousness scale vary between subgroups.

The prejudice against unemployed scale applies a per cent rating method. Respondents were asked about their perceptions how many people perform the disapproved behaviour. This gives respondents the impression of being asked about empirical facts but not about prejudices. Here the factor analysis resulted in a one-dimensional scale. Both scales now are included in the 7th wave of the panel study “Labour Market and Social Security (PASS)”. After data collection researchers may analyse whether and how stigma consciousness and prejudices are directly connected to conceptions of working capacity, own labour market experiences, or other issues of the PASS questionnaire, e.g. responsibility of looking for a job, social participation or other attitudes. It may be analysed, e.g. whether long-term unemployment is related to stigma or whether the stigma effect varies with the local unemployment rate.

Kurzdarstellung

In diesem Beitrag werden Skalen zur Messung von Stigmatisierungsbewusstsein bei Arbeitslosen und zu Vorurteilen gegenüber Arbeitslosen entwickelt. Die TeilnehmerInnen einer vorangegangenen qualitativen Studie berichteten von zahlreichen Erfahrungen, bei denen sie das Ziel von negativen Zuschreibungen bedingt durch ihren inferioren sozialen Status waren. Das Panel „Arbeitsmarkt und soziale Sicherung (PASS)“ erschien daher als eine geeignete Möglichkeit, um neues und substanzielles Wissen über verschiedene Fragen in diesem Zusammenhang zu erlangen: Was denken Arbeitslose welche negativen Charakteristika ihnen zugeschrieben werden? Wie verbreitet sind negative, stereotype Einstellungen gegenüber Arbeitslosen in der Gesellschaft? Zu diesem Zweck entwickeln wir zwei Skalen, deren Gegenstände zum einen die Perzeption von Ausgrenzung und Stigmatisierung unter Arbeitslosen und zum anderen die (negative) Bewertung des Arbeitslosenstatus durch Erwerbstätige sind.

Gemäß Goffman wird Stigma als eine Eigenschaft gebraucht, „die zutiefst diskreditierend ist“, die deren Träger in unseren Vorstellungen „von einer ganzen und gewöhnlichen Person zu einer befleckten, beeinträchtigten“ herabmindert (Goffman 1967: 10–11, im Orig. 1963: 3). Zu Stigmatisierungen kommt es in Situationen, in denen Etikettierung, Stereotypisierung, Ausgrenzung, Abwertung und Diskriminierung durch Andere vorkommen. Stigmatisierung und deren Wahrnehmung haben negative Auswirkungen auf das Selbstbewusstsein, die Leistungen im Bildungssystem oder die Gesundheit. Allerdings gibt es bisher nur wenige Studien, die negative Stereotype gegenüber Arbeitslosen untersucht haben. Die neu entwickelte Skala zum Stigma-Bewusstsein versucht das Ausmaß darzustellen, in dem Personen negative Stereotypisierungen als Arbeitslose empfinden und wie sehr sie sich dadurch in ihrem Leben beeinträchtigt fühlen.

Bei der Diskussion über die Ursachen von Stigmatisierungen beziehen wir uns unter anderem auf theoretische Überlegungen zu gruppenbasierten Hierarchien und Statusdifferenzen, die in Gesellschaften existieren und auf der Basis kennzeichnender Charakteristika entlang spezifischer Zugehörigkeiten – die Rede ist hier von Klasse, ethnischer Zugehörigkeit, religiöser Orientierung und eben auch Erwerbsstatus – gebildet werden. Die sozialen Funktionen von Stigma sind vielfältig: Sich gegen andere Gruppen auszusprechen steigert das persönliche Selbstwertgefühl, trägt dazu bei, eine positive soziale und kollektive Identität aufrechtzuerhalten und Vorteile der statushöheren Gruppe sowie existierende Ungleichheiten zu legitimieren. Da Erwerbsarbeit ein zentraler Modus von sozialer Integration ist, kann ein fehlender Zugang zu einem Job mit leidvollen sozialen Erfahrungen verbunden sein, die das Selbstbewusstsein und die sozialen Interaktionen mit anderen beeinträchtigen.

Wir konstruierten zwei Skalen, um Stigma-Bewusstsein und Vorurteile gegen Arbeitslose zu messen. Der Ausgangspunkt war eine 10-Item Skala zu Stigma-Bewusstsein unter Frauen, die von Elizabeth Pinel (1999) entwickelt worden ist. Da diese Items nach ihrer Übersetzung und Anpassung an unsere Fragestellung sich als zu vage, zu komplex oder nicht sensibel genug erwiesen, entwickelten wir weitere Items basierend auf dem theoretischen Rahmen. Dazu wurden verschiedene Techniken angewandt (kognitive Interviews, mapping sentence Technik). Zudem wurden für die Entwicklung der Vorurteilsskala die in der einschlägigen Literatur besprochenen Vorurteile gesammelt.

Die berichteten Ergebnisse stammen aus 178 Pretest-Interviews. Die Interviewer konnten bei jedem Item eine Notiz machen, wenn Probleme aufgetreten sind und berichten, welcher Art die Probleme waren. Die Items wurden auf der Basis dieser Berichte und mittels einer Faktorenanalyse ausgewählt. Die Faktorenanalyse führte zu vier Dimensionen für die Stigma-Bewusstseins-Skala: soziale Beziehungen von Arbeitslosen, Vermeidung von belastenden Situationen, unter Handlungsdruck stehen und das Bewusstsein, von Vorurteilen betroffen zu sein. Um konvergierende und diskriminierende Validität zu bestimmen, enthielt der Fragebogen auch Skalen zum Exklusionsbewusstsein und zum Kohärenzsinn. Die sozialen Exklusionsskalen korrelierten positiv mit der ersten und zweiten Dimension der Stigma-Bewusstseins-Skala, die Korrelation der dritten und vierten Dimension war nahezu signifikant. In der Kohärenzsinn-Skala, die dazu vorgesehen war, die diskriminierende Validität zu bestimmen, konnten mit den Pretest-Daten zwei Dimensionen identifiziert werden: Enttäuschung über Andere und innere Resignation. Es zeigte sich, dass einige Dimensionen von Stigma-Bewusstsein und Kohärenzsinn signifikant korrelierten. Wir interpretieren dieses Ergebnis aber nicht als Zurückweisung unserer Skala, da die detaillierten Ergebnisse inhaltlich plausibel sind. Zudem betrachteten wir, ob die Werte der Stigma-Bewusstseins-Skala zwischen Subgruppen variieren.

Die Vorurteilsskala wendet ein Prozent-Rating-Verfahren an. Auskunftspersonen wurden zu ihrer Wahrnehmung befragt, wie viele Personen ein zu missbilligendes Verhalten zeigen. Dies vermittelt den Befragten den Eindruck, dass sie zu empirischen Daten, aber nicht über Vorurteile befragt werden. Die Faktorenanalyse führte hier zu einer eindimensionalen Skala. Beide Skalen werden nun in der siebten Welle der Panelstudie „Arbeitsmarkt und soziale Sicherung“ eingesetzt. Nach der Datensammlung lässt sich untersuchen, ob und wie Stigma-Bewusstsein und Vorurteile direkt mit Konzepten des Arbeitsvermögens, eigener Arbeitsmarkterfahrungen oder anderen Themen des PASS Fragebogens, z.B. dem Verantwortungsgefühl für die Jobsuche, sozialer Partizipation oder anderen Einstellungen zusammenhängen. Es kann analysiert werden, ob Langzeitarbeitslosigkeit mit Stigma-Bewusstsein korreliert oder etwa ob der Stigmatisierungseffekt mit der regionalen Arbeitslosenrate variiert.