Contrasts and classtalk: A critical discourse analysis of welfare-to-work program managers

Abstract

Historically, discourse about welfare in the United States has changed from a language of ‘need’ to a culture of ‘dependency’. Adopting ‘dependency’ as a frame to construct the public opinion of welfare, workers help maintain the current punitive welfare state. In this study, we use critical discourse analysis to examine how county program managers in Ohio, USA (N = 69) use several discursive techniques to legitimate their identities as good workers while neutralizing negative connotations associated with administering welfare policy. Further, we find managers use discursive techniques of contrast to heighten boundaries between three pairings: (1) ‘generational’ and ‘situational’ clients, (2) clients and non-clients, and (3) welfare workers and clients. Managers engage in ‘classtalk’ in their contrasts in a way that blames the poor through contrasts with others with ‘middle-class values’.

Keywords

Classist discourse classtalk critical discourse analysis discourse discourse analysis gender prejudice racism racist discourse social class TANF welfare welfare discourse

Introduction

Throughout much of United States history, poverty was thought of as an issue of personal talents and motivations without considering the structural inequalities which confine and define the opportunities available. Public assistance was shameful to seek and minimal in its effect. By the 1960s, the public and political rhetoric of welfare policies of the USA shifted from a language of ‘need’ to a culture of ‘dependency’ (Kingfisher, 1996; Marston, 2008; Misra et al., 2003). This type of language escalated to fever pitch during Reagan’s administration. While campaigning for President in the 1970s, he coined the term ‘welfare queen’, fueling raced and gendered images of the single mother with numerous children she could not financially support (Gilens, 1999; Neubeck and Cazenave, 2001).

Armed with ‘welfare queen’ stereotypes, the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA) passed under the Clinton administration (Gilens, 1999; Hancock, 2004). PRWORA effectively abolished Aid to Families with Dependent Children (AFDC) and replaced it with Temporary Assistance to Needy Children (TANF). The name change was telling as this new program would indeed be temporary, with state and federal time-limited assistance. The new program was also framed as ‘assistance’ rather than ‘aid’ and mandated participants to fulfill a work requirement. Finally, the children had to be ‘needy’, which is perhaps reflective of ‘family caps’ policies that prevent additional cash assistance if the program participant has another child while receiving assistance. This program is designed and implemented largely under the discretion of counties within states and the effectiveness of the programs is measured by reported caseload numbers and program participants’ involvement in work-related activities.

While more is known about welfare discourse in the media (Misra et al., 2003) and among policymakers (Hancock, 2004), and something is known about the welfare discourse of service providers in other countries (e.g. Van De Mieroop, 2011), we are aware of little research that examines the welfare discourse of service providers in the USA. We attempt to fill this gap in the literature using rich interview data from 69 welfare-to-work managers in Ohio, USA. We use critical discourse analysis (CDA) (Fairclough, 1985, 2010; Van Dijk, 1987, 1993) to examine how county program managers use discursive techniques to legitimate their identities as good workers, while neutralizing the negative connotations associated with administering welfare policy. Further, we find managers use discursive techniques of contrast to draw on ‘discourses of difference’ (Wodak, 1997) that heighten boundaries between three pairings: (1) ‘generational’ and ‘situational’ clients, (2) clients and non-clients, and (3) welfare workers and clients. Managers often engage in what we call ‘classtalk’ in which they celebrate ‘middle-class values’ while blaming the poor for supposedly lacking these values.

Literature review

Welfare discourse in the USA

In the USA, the word ‘welfare’ is inseparable from gendered and racialized images (Gilens, 1999; Neubeck and Cazenave, 2001). The welfare queen image has been carefully constructed and reinforced through political rhetoric and discourse couched in the language of ‘self-sufficiency’. Furthermore, the topic of structural racism within the welfare system is labeled ‘politically’ and ‘foundationally’ incorrect (Gilens, 1999; Neubeck and Cazenave, 2001; Schram, 1993). The effect of these discursive strategies is the continuation of the racialized and gendered stereotypical image that allows policymakers to pass punitive welfare policies (Marston, 2008). Additionally, welfare discourse implicitly creates both symbolic and social boundaries (Lamont and Molnar, 2002) along race, class, and gender lines. These techniques also lend themselves to generalized depictions of welfare dependency in media discourse.

Media discourse is a constituent of public discourse and, as such, draws on a specialized vocabulary to create frames of understanding for issues of public concern (Gilens, 1999; Misra et al., 2003). These frames are powerful conduits of understanding and knowledge which, when backed by influential individuals and media outlets, serve to replicate existing stereotypes and inequalities. The practice of using ‘dependency’ as a frame to construct the public opinion of welfare allows the system to maintain the status quo (Kingfisher, 1996; Marston, 2008). By couching welfare in a language of dependency and personal choice, welfare regulations and policy can be implemented and enforced with punitive and unsympathetic efficiency (Schram, 1993).

Unfortunately, the lived experiences that exist in this discourse are those of suffering, frustration, and social exclusion. The treatment of low-income women by welfare caseworkers threatens not only their livelihoods, but also their citizenship, their status as contributing members of society (Cook and Marjoribanks, 2005; Van De Mieroop, 2011). They are seen as dependent on the welfare bureaucracy to correct their moral failings. A dichotomy arises from this status; a quagmire in which welfare recipients are seen as personally responsible for needing welfare assistance (Schram, 1994), but are simultaneously robbed of their means for individual empowerment (Marston, 2008). This discourse leaves welfare mothers feeling frustrated, nihilistic, and powerless against the towering leviathan of the system (Kingfisher, 1996; Pollack and Caragata, 2010; Van De Mieroop, 2011). The lack of agency on the part of welfare recipients is reflected in the proliferation of paternalism by welfare offices, and welfare discourse serves to legitimate the welfare regime and the trend of neoliberalism (Connor, 2010; Lorenzo-Dus and Marsh, 2012; Wiggan, 2012).

In talking about welfare discourse, it is important to contextualize service delivery within bureaucratic human services agencies, which are in turn embedded in a capitalistic economic system (Ferguson, 1984; Swift, 1995). Consequently, service delivery is also shaped by bureaucratic discourse, which is informed by and maintains the social order of inequality created by capitalism (Ferguson, 1984). Bureaucratic discourse is characterized by regulation, depersonalization, and mechanisms of dominance and social reproduction (Ferguson, 1984). These features and processes serve to produce certain types of workers and clients (Ferguson, 1984). The highly regulative and depersonalized workplace constrains and isolates workers, who have little room to resist policy lest they be penalized or replaced (Ferguson, 1984). This discourse also enables service agencies to ‘produce knowledge’ about clients by determining client eligibility criteria, program requirements, and ‘treatments’ for clients, which tend to simultaneously reinforce norms about work (Ferguson, 1984; Swift, 1995). In these institutions, workers and managers are viewed as ‘experts’ on the poor, which legitimates this production of clients (Ferguson, 1984; Swift, 1995). Furthermore, clients lack a voice in this process, as it is constructed in the ‘foreign’ language of bureaucratic discourse (Ferguson, 1984; Swift, 1995).

Integrating critical discourse analysis

In much of society, mechanisms of domination are concealed and naturalized. Dominant discourse ideologies fall under this umbrella and they frequently go unquestioned as a result. As such, the interests and power of the dominant group tend to be protected and maintained (Fairclough, 1985, 2010; Van Dijk, 1993, 1996). Critical discourse analysis seeks to denaturalize those aspects of dominant discourse that get taken for granted (Fairclough, 1985). It further connects local interaction and speech acts to social structures by recognizing that social structures shape and are shaped by micro-level processes (Fairclough, 1985, 2010). In this light, discourse can be a tool for both social reproduction and social change (Fairclough, 1985, 2010; Van Dijk, 1993, 1996).

One way in which critical discourse analysis denaturalizes dominant discourse is by unveiling strategies that social actors use when speaking. In Van Dijk’s (1987) work he discusses semantic moves, which are discursive strategies that social actors use in order to manage the content of their speech. Often times, social actors use these strategies to positively present themselves while negatively presenting others. While Van Dijk (1987) specifically talks about these strategies in the context of racially prejudiced talk, managers in our data used a similar form of prejudice talk that was more class-based (though we have discussed the gendered and racialized discourse of welfare). We draw from his framework, specifically focusing on semantic moves of contrast, examples and explanations, generalizations, apparent concessions, and mitigation.

Contrast is a semantic move in which the speaker focuses on the differences between two groups. This facilitates an ‘us versus them’ mentality through which the in-group is often portrayed as a victim. By contrasting groups, speakers both draw on and contribute to ‘discourses of difference’. Explanations are used to excuse or validate a speaker’s stance on a potentially delicate topic. Similarly, examples offer narratives or specific instances in order to illustrate the validity of the speaker’s stance. Generalizations are when someone applies an experience/story/example to an entire group. That is, rather than understanding the account as an individual instance, generalizations suggest that it is typical of all group members. Apparent concessions describe when a speaker makes a positive or neutral acknowledgment about a group or group member, and are often used in tandem with negative presentation of the group or group member. Finally, a speaker uses mitigation to reduce the negativity of what they just said, or were trying to say.

Methods

Data and data collection

To examine managers’ welfare discourse techniques, we use semi-structured telephone interviews with Ohio Works First (OWF) managers in 69 of 88 Ohio counties (N = 69; nearly 75% of the population). Ohio Works First managers are responsible for administering the OWF program. Their duties include (but are not limited to) supervising and integrating the work experience program, ensuring program requirements are fulfilled, and coordinating with other local service providers. Additionally, some OWF managers also perform caseworker duties. The telephone interviews that we conducted lasted from under an hour to nearly three hours, averaging just over an hour. The interview questions covered topics such as manager perceptions of welfare participants (i.e. describe ideal participant, describe typical participant); what nonprofits the agency works with to serve families; how successful different methods of service delivery are perceived to be; the ways in which agency personnel work with community nonprofits and for-profits to assist families in need; and barriers and challenges to helping families reach self-sufficiency. All interviews were audio recorded and transcribed, resulting in 581 pages of single-spaced transcripts.

While it is somewhat unusual to collect qualitative data from telephone interviews, there are several advantages to using this method of data collection. The first advantage is a practical one, in that it is far less expensive to conduct telephone interviews than face-to-face interviews. More importantly, telephone interviews can yield higher response rates and reduce coverage bias. We wanted good coverage of Ohio counties to capture the stories of managers working in a wide variety of conditions (Czaja and Blair, 1996). A third advantage of this method of data collection is that telephone interviews are advantageous when dealing with a population for which face-to-face interviews are difficult to schedule and where respondents may be reluctant to participate (Fenig et al., 1993). Welfare-to-work managers are extraordinarily busy and time is always a scarce resource. In most instances, scheduling a phone interview took some coordination, but we would not have had the county coverage we obtained if we had conducted face-to-face interviews.

When we designed our study, we had some concern that telephone interviews would not yield the same rich data as face-to-face interviews. Telephone interviews have long been regarded as suitable in quantitative research, however, and recently scholars (Novick, 2008; Sturges and Hanrahan, 2004) have shown that telephone interviews are also appropriate for qualitative research. For instance, Sturges and Hanrahan (2004) find comparable results in their telephone and face-to-face interviews of public service employees. This suggests that using telephone interviews is not significantly different from using face-to-face interviews, and, reassuringly, the authors’ results are from a similar population to the individuals we study in this research. It is sometimes suggested that the lack of nonverbal cues in telephone interviews is a disadvantage; however, little to no evidence supports this claim. Telephone interviewers can gauge respondent tone, pauses, and sighs in much the same way a face-to-face interviewer will ‘read’ these in person (Novick, 2008). Finally, our experience has been that welfare-to-work managers spend the majority of their days on the phone. This is a method of communication in which they are comfortable. Our data are rich with stories about managers’ work lives and managers were especially candid in their illuminating discussions of clients. In sum, our findings are indicative of this level of comfort with telephone communication and the rapport interviewers gained with the managers in the interviews.

Coding and analysis

The uniqueness of our data allows us to examine understudied processes of program managers’ welfare discourse. In our analysis we follow a modified grounded theory approach (Charmaz, 2001, 2006), in which we open and then focus code data, allowing themes to emerge from the data for further coding and analysis. All authors assisted with early phases of coding, and it was during the open coding phase of our analysis that we noticed the ‘rhetorical gymnastics’ that program managers perform when speaking about clients (Mallinson and Brewster, 2005: 789). This rhetorical gymnastics, combined with our knowledge of the power differential between program managers and clients, informed our decision to use critical discourse analysis (Van De Mieroop, 2011; Van Dijk, 1992, 1993) in subsequent coding phases. Once we knew we would employ CDA, we developed a detailed code guide and the first author then coded for the concepts discussed in the following paragraph.

Following Van Dijk (1987), we coded for the following concepts: contrast, examples and explanations, generalizations, apparent concessions, and mitigation. Contrast was coded when a program manager highlighted differences between groups. We combined instances of two of the semantic moves – explanation and example – into a single category since the concepts overlap as a discourse strategy intended to offer evidence when talking about sensitive issues. We coded for generalizations when managers implied that an entire group possesses specific characteristics, attitudes, or behavior based on their impression of some of its members. In many cases, generalizations also oversimplified characteristics, problems, and solutions within the group. Apparent concessions describe when a program manager makes a positive or neutral acknowledgment about a group or group member. Finally, we coded for mitigation when program managers attempted to reduce negativity by negotiating their phrasing.

We note that managers used a variety of semantic moves (Van Dijk, 1987) singularly, but also in combination. Coding and analysis was an interactive process consistent with our modified grounded theory approach. As we began to focus-code the data using CDA, we generated analytic memos to help elucidate the sociolinguistic processes we were observing in our data. From these memos, several themes emerged from the data. We report those themes in this article. To protect the identity of managers and counties, when we quote managers we use unique county numbers (like pseudonyms) that were randomly assigned.

Findings

We find that program managers make three major contrasts between: (1) generational and situational clients, (2) clients and non-clients, and (3) clients and welfare officials. We also find that program managers use generalizations, explanations and examples, and – to a lesser extent – mitigations and apparent concessions when talking about clients. We will discuss each of these paired contrasts and explain how program managers make use of semantic moves in order to clearly and strategically positively present one group, while casting the other group in a negative light. While it is possible that the race and gender of program managers is influential, and we offer demographic characteristics of each manager, it is important to note that Ohio Work First managers were overwhelmingly white women. Therefore, very little variation occurs by race and gender since the respondents are homogeneous. Further, we focus on managers’ ‘classtalk’, though the discursive racialization and gendering of clients is also important. Consistent with our prior research in North Carolina (Taylor et al., 2011), we find Ohio managers also avoid discussing race. Managers take gender for granted, but construct strong images of ‘good’ and ‘bad’ mothers. However, this is a research project in its own right and is the focus of another article.

Generational clients versus situational clients

Managers use semantic moves to differentiate between generational and situational clients. The term ‘generational’ refers to clients who have been on welfare long term, who mostly grew up in families who received public assistance. The resulting connotation behind the term ‘generational’ is that poor families have a ‘culture of poverty’, whereby certain values and practices keep them poor rather than facilitate mobility out of poverty (Kingfisher, 1996). Part of this image is that poor parents do not teach their children the value of work and instead teach them to continue to live on public assistance (Seale et al., 2012). Conversely, ‘situational’ clients are those who receive public assistance temporarily, often as a result of the economic recession. These clients are thought to possess more ‘middle-class values’ which allow them to only need to receive assistance temporarily. While managers use the terms ‘situational’ and ‘generational’, the distinction they really seem to be making is based on the ‘cultural values’ that managers assume the two groups of clients possess. Thus, we see that program managers use these terms to make thinly disguised class comparisons:

(1) [A white male program manager when asked to describe a typical OWF client]

So I think (sigh) once again the simplified answer is, it’s somebody that lacks the very basic skills that are necessary to succeed um, in-in employment. And also, typically it’s somebody that has come from generational upbringing, whatever. We’ve had a big list of circumstances lately, we see a lot more of our OWF participants come from people that are in situational poverty rather than generational poverty, people who have lost a job or that type of thing. (Ohio_27)

This manager explicitly uses and frames generational and situational as two different types of poverty. The manager generalizes that typical clients come from a generational upbringing and lack the very basic skills that are necessary to succeed. Yet, after making these generalizations, the manager goes on to say that many current clients are situational and are only on OWF circumstantially. Thus, this manager describes typical OWF clients in a way that presents generational clients in a negative light and excludes situational clients from the negativity.

Another male program manager (who declined to specify his race) further defines the differences between generational and situational clients when asked about the types of clients who are able to reach self-sufficiency:

(2)

Like I think I told you before there’s a – there’s a uh, percentage of people that use this for what it’s for – temporary assistance for the needy. TANF or whatever it used to be called.

They come on, they need it for two months or a year and they get another job and they’re off assistance. It happens all the time. The problem we have is like you said before, we have generational people that, you know, 3rd, 4th, 5th generation, living off assistance. I think it has a lot to how they’re brought up. It, it, the type of person they are. Uh, a lot of times you have people, you know, that doesn’t have the ambition to do that, you know. (Ohio_18)

In this quote, the program manager uses an example/explanation and generalizations to contrast generational and situational clients. He uses example/explanation to offer explanatory evidence of why some clients are more likely to become self-sufficient than others. He explains that certain clients do not reach self-sufficiency because of how they’re brought up . . . the type of person they are, and they are people that do[n’t] have the ambition to do that. The program manager also uses generalizations to assume that ‘ambition’ is based on clients’ upbringing and personal characteristics. By doing so, he makes a distinction between clients who reach self-sufficiency and those who do not.

Program managers also use discourse strategies to differentiate between successful and unsuccessful clients. A white female program manager said the following when asked about why some clients reach time limits, while others do not:

(3)

Okay. So we’re talking, and keeping that in mind now, and then we’ve got the working poor, you know? Who, you know, may be able to get a job but they lose the job very quickly. So they all become very dependent and very scared um, to be without that cash assistance. Um, the bulk of the new person that coming in, the middle income, they’re seeing this as temporary mandate. They don’t want to be on this very long, they want to be getting a job. But the ones who do max out are the ones that have a lot of issues.

They’ve got domestic violence, they’ve got health issues, mental health issues. Um, they’ve got turbulence in the household, the kids are involved in juvenile court systems or the children have learning disabilities, they’re on IUT so they’re ADHD. You know, they just have so, and they have never developed those skills themselves to manage, to life manage I guess is a better way to put it. They just don’t life manage. (Ohio_82)

This program manager uses generalization, example/explanation, and mitigation to talk about clients. She uses generalizations to suggest that clients who reach time limits have a lot of issues. In the sentences that follow this generalization, the manager lays out a series of further generalizations about what those issues are, ranging from health issues to domestic violence, and ending with a generalization about clients’ inability to life manage. Yet the program manager generalizes that clients who do not reach time limits are middle income and they’re seeing this as a temporary mandate. She further generalizes that situational clients don’t want to be on this very long, they want to be getting a job. Through these generalizations, this program manager creates a clear distinction between the circumstances, skills, and desires of generational and situational clients.

The explanation/example in this quote consists of the manager explaining that clients who reach time limits have a lot of personal issues and barriers that prevent them from becoming self-sufficient. By listing out these barriers, the program manager uses them as evidence of why certain clients are unable to become self-sufficient. Yet this explanation fails to mention the more structural aspects of poverty that prevent the poor from rising out of it (Schram, 1994; Sherman, 2009). Finally, the program manager uses mitigation in order to reduce negativity by using the term ‘life manage’ when talking about her negative opinion that certain clients are incapable of running their own lives.

In some instances, program managers would specifically reference the ‘culture of poverty’ when talking about differences between types of clients. A white female program manager did so when asked if there were some clients that were more likely to become self-sufficient than others:

(4)

(Pause) I, yes. But I still think a lot of times it depends on their family culture, the culture of poverty. If they have been in, if they have lived their whole life, um (pause) in that culture it is a stretch for them. It is a stretch for them. It is very different for someone who has worked their whole life and then their job ends or their illness comes or something comes.

Those folks are much more like to get back into employment quickly. Than those people who have um (pause) used it as their means of survival, (pause) rather than just a temporary stretch. (Ohio_47)

This program manager uses the semantic moves example/explanation and generalization to talk about clients and, consistent with findings in other research (Kingfisher, 1996; Pollack and Caragata, 2010), finds service providers focus on dependency and culture of poverty rhetoric. She uses example/explanation to explain that clients who have lived their whole life, um (pause) in that culture [of poverty] have a more difficult time reaching self-sufficiency than clients who have worked their whole life and then their job ends or their illness comes or something comes. This explanation offers evidence as to why certain clients cannot overcome poverty. The program manager also uses the semantic move generalization to generalize that a culture of poverty exists and that people who grew up in poor/non-working families have a more difficult time finding a job and becoming self-sufficient than people who did not grow up in this culture and have worked for most of their lives. She further generalizes that generational clients view public assistance as a means of survival, while situational clients see it as just a temporary stretch. It is worth noting that OWF is a time-limited program and therefore temporary. Ultimately, by contrasting generational and situational clients, program managers account for and justify differences in expectations, treatment, and success between these two groups.

Clients versus non-clients

Another pair of groups that program managers contrasted via semantic moves is clients and non-clients. As in the previous section, ‘clients’ refer to individuals on the OWF cash assistance program. The term ‘non-clients’ refers to the general public, or an average person who is not on welfare. In contrasting these groups, program managers suggest that welfare clients have deficiencies in certain areas that the rest of the world would see as ‘common sense’ (see also Cook and Majoribanks, 2005). They use semantic moves in order to show that clients lack knowledge, skills, or a routine that presumably everyone else has. In contrasting clients and non-clients, program managers use these discourse strategies to negatively present clients as incompetent and lazy, while positively presenting an ‘average’ – or middle-class – person as competent and having a good work ethic because they are ‘in the know’ and have work experience.

Some program managers deconstructed the source of the differences between clients and non-clients, which was often along class lines. When asked about challenges to becoming self-sufficient, a white female program manager elaborated on these differences:

(5)

Um well, if you’ve done any training with bridges out of poverty, that is very, very true that the poverty class thinks about their priorities are very different and the working employment mentality is very middle class so it’s very hard for them to cross that line. Um an example of that is that um, you know, somebody in the poverty class would very willingly not go to work and babysit somebody else’s kids that needed help, not thinking what it might do to them because their most important thing is I might need help someday and that person could help me then. You know, so helping people out is very important and is not so, you know, in the middle class it’s very, you go to work, no matter what, you go to work. That’s the mentality. And that is, you know, the poverty class says you need to help other people out because you might need them one of these days. (Ohio_78)

This program manager begins by equating ‘employment mentality’ and ‘very middle class’ to draw the distinction between the values that poor people possess and those that workers and middle-class people possess. Yet this implies that clients need only learn middle-class values to succeed in the workplace and rise out of poverty (Seale et al., 2012). She uses the semantic moves generalization and example/explanation when contrasting clients and non-clients in this quote. In the remainder of the quote, the program manager uses the example of how a member of the poverty class and a member of the middle class would respond to a situation in order to illustrate the difference in priorities between these two groups. She also uses generalizations within the framework of this example to generalize that the poor would very willingly not go to work and babysit somebody else’s kids that needed help, not thinking what it might do to them, while middle-class people go to work, no matter what. These generalizations attribute certain priorities to the poor, that are not conducive to the values associated with work, which the manager suggests are middle class. Thus, this manager uses semantic moves to illustrate how differences between the priorities of the poor (clients) and the middle class (non-clients) contribute to differences in success in the workplace. Consequently, this mentality oversimplifies poverty and ignores the structural barriers that perpetuate it (Sherman, 2009), placing all the blame on clients for their lack of success (Kingfisher, 1996; Pollack and Caragata, 2010).

Many program managers said that clients do not understand what it takes to get a job and lack the knowledge and skills to keep a job. The following was a white female program manager’s response to a question about why some clients are more likely to get sanctioned than others:

(6)

And then there are others that they think um, do have things that come up, people in poverty are in reactive mode and they don’t necessarily have the same work ethic or things like that. A perfect example is that we had a lady last week who didn’t show up at her Work Experience Program because she took her neighbor to the hospital and then stayed with her all day. I’m not saying that she shouldn’t have taken her neighbor to the hospital but in the real world-in work, would you have been allowed to stay all day and miss your job? (Ohio_24)

Here, the program manager again employs an example/explanation and generalization when talking about differences in how clients and non-clients work. She generalizes that clients don’t necessarily have the same work ethic, presumably comparing them to non-clients. Furthermore, the program manager generalizes that the work clients do is substantively different from the real world of work. To illustrate the difference in work ethic between clients and non-clients, the program manager uses an example of a woman taking her neighbor to the hospital rather than going to work. Since the client chose to miss work to help her neighbor, the client’s priorities are also called into question. This reflects the sentiment of a previous quote in which the program manager suggested this difference in priorities is due to the poor subscribing to the ‘culture of poverty’, while middle-class values suggest you go to work, regardless of the situation (Pollack and Caragata, 2010; Seale et al., 2012).

The contrasts program managers made between clients and non-clients also extended out of the realm of work and into clients’ households. In this quote, a program manager specifically uses the term ‘common sense’ when talking about making decisions about whether or not to have children. This white female program manager offered the following response when asked to describe a typical OWF client:

(7)

I think one of the things that I observe here daily is how many little children that they have and yet they’re unmarried and they may have two little kids and they come in and they’re pregnant and they’re still not married, they’re already in a bad a situation but they’re having another child. We see that a lot. It’s heartbreaking. It’s almost like you know you would think common sense would tell you this is not that, you know, you

don’t need another child. You can’t afford the ones that you have and which limits, and again it goes back generational. (Ohio_13)

In this quote, the program manager uses generalizations and examples/explanations. She generalizes that clients have many little children, are unmarried, are pregnant and they’re still not married and are already in a bad situation but they’re having another child. These generalizations not only make assumptions about clients’ situations, but they also impose value judgments on clients’ family situations. By doing so, the program manager segues into an explanation of why clients should not be having more children. She explains that clients should not be having children because they cannot afford them, nor do they have a stable family structure in which to raise them. The program manager relates this back to ‘common sense’, suggesting that non-clients would not consider having children if they could not afford them or did not have a stable family structure. Again, we see negative presentations of a specific kind of client since she suggests that it’s ‘generational’, which implies that only clients who grew up in the ‘culture of poverty’ would violate ‘common sense’ in such a way (Cook and Majoribanks, 2005). Thus, we see that discourse strategies are used not only to negatively present clients in the context of work, but also in regards to topics like clients’ family structure.

Clients versus welfare officials

The final group that program managers contrast is clients and welfare officials. Since welfare officials are also non-clients, there was some overlap between this group comparison and that of the previous group. The welfare officials they typically mention are either themselves or caseworkers. These comparisons used semantic moves in order to negatively present clients and positively present themselves. These comparisons were often made to draw fault away from welfare officials for the clients who failed, and thus drew on generalizations about ‘bad’ clients. Managers presented themselves and their caseworkers as doing everything within their power to help clients attain self-sufficiency (Seale et al., 2012), while they presented clients as unwilling to do anything to help themselves. Thus, they contrast their own work ethic with those of clients. This seemed to be the main way program managers contrasted themselves with clients.

The following quote highlights how program managers use semantic moves to positively present themselves as helpful and to negatively present their clients in a way that emphasizes that failure was on the client’s part. In the first quote, a white male program manager responds to being asked how effective OWF is in helping clients become self-sufficient:

(8)

Huh. I really don’t know but I’ve always hit the wrong time with people with initiative to want to work. We’ll work with you. There’s not one person here that won’t help you apply for a job or I have even personally given referrals out to consumers. We have consumers that just don’t want to work. (Ohio_07)

The program manager in this quote uses generalization and example/explanation in order to negatively present the clients (‘consumers’) and positively present the welfare officials. He generalizes that some clients just don’t want to work. Yet he generalizes positively about welfare officials, saying that there’s not one person here that won’t help you. The program manager also uses an explanation by using this differential input of effort by welfare officials and clients as evidence that clients are personally responsible if they do not achieve self-sufficiency (Schram, 1994).

This second quote similarly describes a program manager talking about welfare officials as helpful and willing to invest effort in getting clients to become self-sufficient. The program manager, a white female, was asked about why some clients are more likely to receive sanctions than others:

(9)

Some- some of it is generational. Um, when their parents were on the program they didn’t have to do anything like this so they don’t feel that they should have to. Um, some of it is mental health issues. Even though we try really hard, sometimes you can’t help someone who doesn’t want to be helped. Um-you know, so that’s probably the majority. (Ohio_08)

In this quote, the program manager uses generalization and example/explanation in a similar way to the previous program manager. She generalizes that clients don’t feel that they should have to [do anything], may have mental health issues, and do[n’t] want to be helped. These generalizations present clients as lazy and stubborn, since they don’t feel like they should have to work and are unwilling to accept help. She also positively generalizes about welfare officials, saying we try really hard. Thus, the generalizations suggest that welfare officials do everything they can to help clients, while clients are lazy and stubborn. This again segues into the program manager’s explanation that places personal responsibility for clients’ failures on clients, since managers try their very best to help them (see also Seale et al., 2012).

Also present in these contrasts between clients and welfare officials is language that reflects the ‘mutual responsibility agreement’ between them. This agreement theoretically holds both welfare officials and clients responsible for making an effort to reach self-sufficiency for the client. Thus, some welfare officials speak critically about clients’ inability to uphold their end of the agreement. One manager – a white female – seems to do this when responding to a question about barriers to clients reaching self-sufficiency:

(10)

Um, their attitude. Um, uh, lack of work ethic um, just kind of an attitude that, you know, there’s nothing out there so I shouldn’t have to do anything. Um, you know, not having the consideration to do the proper call off procedures. Um, them thinking that it’s okay to call off for every little thing that goes on in their lives. Um, not really taking the program seriously and for what it’s intended to help them become um, better workers and gain and grow with their work ethic. Um, so I think that’s the biggest obstacle we have at this point. (Ohio_86)

In this quote, the program manager uses generalization and example/explanation while talking about clients. She generalizes that clients have an attitude, feel like they shouldn’t have to do anything, [do] not hav[e] the consideration to do the proper call off procedures, think … it’s okay to call off for every little thing that goes on in their lives, and do not tak[e] the program seriously. Each of these generalizations calls out a fault with clients and negatively presents them in a way that blames them for not reaching self-sufficiency. That is, she suggests that clients have not been playing their part and upholding their end of the ‘mutual responsibility agreement’. By placing this blame on clients, she appears to be excusing herself from blame and implying that she (and other welfare officials) has upheld her end of the agreement. The program manager also uses an example/explanation in order to explain that clients have a poor work ethic and attitude because they don’t feel like they should have to do anything, and call off work when anything is going on in their lives. Thus, consistent with other research (Kingfisher, 1996; Marston, 2008; Misra et al., 2003), this program manager suggests that clients’ falling through on their end of the ‘mutual responsibility agreement’ is the greatest barrier for clients in reaching self-sufficiency.

Discussion and conclusion

Our findings add to the literature on welfare, welfare discourse, and critical discourse analysis. Program managers use semantic moves to contrast generational and situational clients, clients and non-clients, and welfare officials and clients. These contrasts are significant as they highlight the ‘good’ individuals whom managers associate with ‘middle-class values’, while denigrating the ‘bad’ individuals whom managers argue lack these values. These contrasts, then, serve as ‘classtalk’ that invokes a ‘culture of poverty’ ideology and ignores structural conditions and causes of poverty (Kingfisher, 1996; Marston, 2008; Schram, 1993).

The most prominent discourse strategy managers used in our data was generalization. Managers generalize that generational clients are ‘bad’ clients who grow up on welfare, lack a work ethic, and do not understand how the working world functions. Situational clients, however, are ‘good’ clients with a work history and skills. When contrasting clients with non-clients, managers generalized all clients as having characteristics of ‘generational’ clients and described them as incapable of using ‘common sense’. In other contrasts, managers generalized themselves as helpful and non-clients as having ‘common sense’ and ‘middle-class values’.

Generalizations alone do not support the various contrasts managers made. Rather, they were used in tandem with examples/explanations of clients’ thoughts, behaviors, and upbringings to solidify group differences. Managers commonly used examples/explanations to illustrate reasons why clients were failing to become self-sufficient. The examples/explanations for situational clients, non-clients, and welfare officials were markedly more positive. Additionally, program managers used mitigation to strategically frame their speech to appear less prejudiced. Apparent concessions were made to illustrate that some clients are able to succeed, and suggest that if some clients are able to reach self-sufficiency, then all clients should be able to.

We argue that managers use contrasts as a form of ‘classtalk’ that parallels ‘racetalk’ (Mallinson and Brewster, 2005). Managers incorporate symbolic and social boundaries into their discourse (Lamont and Molnar, 2002) which separate welfare recipients from non-recipients. By using ‘classtalk’, program managers both draw on and contribute to these symbolic and social boundaries and ‘discourses of difference’ along class lines (Lamont and Molnar, 2002; Wodak, 1997). However, ‘classtalk’ is also rooted in the bureaucratic production of clients and the gendered and racialized image of the welfare queen (Ferguson, 1984; Gilens, 1999; Hancock, 2004). Additionally, ‘classtalk’ is reflective of US meritocracy and neoliberalism (Connor, 2010; Marston, 2008; Wiggan, 2012), in which inferences are made about good clients and bad clients, often with assumptions about the amount of effort that clients make. Thus, we see that the ‘classtalk’ of these program managers reflects broader welfare discourses of difference, dependency, personal responsibility, and the ‘culture of poverty’ (Ferguson, 1984; Kingfisher, 1996; Marston, 2008; Schram, 1994; Seale et al., 2012; Sherman, 2009; Wodak, 1997).

Footnotes

Acknowledgements

The authors would like to thank their colleagues on the research project, especially Christi Gross, for their assistance with this research. The authors also thank Christine Mallinson, and the reviewers and editor of Discourse & Society for their helpful feedback, which greatly assisted in the development of this article.

Funding

This research was generously supported by the Kent State University Department of Sociology and the American Sociological Association/National Science Foundation Fund for the Advancement of the Discipline.

Author biographies

Brianna Turgeon is a graduate student at Kent State University, USA. Her current areas of interest include mothering, gender, inequality, social psychology, and adolescence. She is currently working on her Master’s thesis, in which she uses qualitative data to analyze discourses about poor mothers on the welfare-to-work program in Ohio.

Tiffany Taylor is an assistant professor of sociology at Kent State University, USA. In her research, she examines a number of topics related to inequality. Her recent research focuses on policy implementation of programs for poor mothers in North Carolina and Ohio.

Laura M Niehaus is a second-year graduate student at the University of Akron, USA. Her current areas of interest include deviance, gender, discourse, media, and hoarding. Laura has just finished collecting qualitative data for her Master’s thesis and her current work on adult children of hoarders.

References

Charmaz

(2001) Grounded theory. In: Emerson

(ed.) Contemporary Field Research: Perspective and Formulations. Prospect Heights, IL: Waveland Press, pp. 335–352.

Charmaz

(2006) Constructing Grounded Theory: A Practical Guide through Qualitative Analysis. Thousand Oaks, CA: SAGE.

Connor

(2010) Promoting ‘employ ability’: The changing subject of welfare reform in the UK. Critical Discourse Studies 7(1): 41–54.

Cook

Majoribanks

(2005) Low-income women’s experiences of social citizenship and social exclusion. Just Policy 38: 12–19.

Czaja

Blair

(1996) Designing Surveys: A Guide to Decisions and Procedures. Pine Forge, CA: SAGE.

Fairclough

(1985) Critical and descriptive goals in discourse analysis. Journal of Pragmatics 9: 739–763.

Fairclough

(2010) Critical Discourse Analysis: The Critical Study of Language, 2nd edn. Edinburgh: Pearson Education.

Fenig

Levav

Kohn

. (1993) Telephone versus face-to-face interviewing in a community psychiatric survey. American Journal of Public Health 83(6): 896–898.

Ferguson

(1984) The Feminist Case against Bureaucracy. Philadelphia, PA: Temple University Press.

10.

Gilens

(1999) Why Americans Hate Welfare: Race, Media and the Politics of Anti-poverty Policy. Chicago, IL: University of Chicago Press.

11.

Hancock

(2004) The Politics of Disgust: The Public Identity of the Welfare Queen. New York: New York University Press.

12.

Kingfisher

(1996) Women on welfare: Conversational sites of acquiescence and dissent. Discourse & Society 7: 531–557.

13.

Lamont

Molnar

(2002) The study of boundaries in the social sciences. Annual Review of Sociology 28: 167–195.

14.

Lorenzo-Dus

Marsh

(2012) Bridging the gap: Interdisciplinary insights into the securitization of poverty. Discourse & Society 23(3): 274–296.

15.

Mallinson

Brewster

(2005) ‘Blacks and bubbas’: Stereotypes, ideology, and categorization processes in restaurant servers’ discourses. Discourse & Society 16(6): 787–807.

16.

Marston

(2008) A war on the poor: Constructing welfare and work in the twenty-first century. Critical Discourse Studies 5(4): 359–370.

17.

Misra

Moller

Karides

(2003) Envisioning dependency: Changing media depictions of welfare in the 20th century. Social Problems 50(4): 482–504.

18.

Neubeck

Cazenave

(2001) Welfare Racism: Playing the Race Card against America’s Poor. New York: Routledge.

19.

Novick

(2008) Is there a bias against telephone interviewing in qualitative research? Research in Nursing & Health 31(4): 391–398.

20.

Pollack

Caragata

(2010) Contestation and accommodation: Constructions of lone mothers’ subjectivity through workfare discourse and practice. Journal of Women and Social Work 25(3): 264–277.

21.

Schram

(1993) Postmodern policy analysis: Discourse and identity in welfare policy. Policy Sciences 26: 249–270.

22.

Schram

(1994) Postindustrial welfare policy: Just say no to women and children. Review of Radical Political Economics 26(1): 56–84.

23.

Seale

Buck

Parrotta

(2012) Who’s to blame? The identity talk of welfare-to-work program managers. Sociological Perspectives 55(3): 501–527.

24.

Sherman

(2009) Those Who Work, Those Who Don’t: Poverty, Morality and Family in Rural America. Minneapolis, MN: University of Minnesota Press.

25.

Sturges

Hanrahan

(2004) Comparing face-to-face and telephone qualitative interviewing: A research note. Qualitative Research 4(1): 107–118.

26.

Swift

(1995) Manufacturing ‘Bad Mothers’: A Critical Perspective on Child Neglect. Toronto, ON: University of Toronto Press.

27.

Taylor

Samblanet

Seale

(2011) The new ‘lazy’: Structure and agency in managers’ discussions of welfare clients’ motivation. Race, Gender and Class 18: 171–188.

28.

Van De Mieroop

(2011) Identity negotiations in narrative accounts about poverty. Discourse & Society 22(5): 565–591.

29.

Van Dijk

(1987) Communicating Racism: Ethnic Prejudice in Thought and Talk. Newbury Park, CA: SAGE.

30.

Van Dijk

(1992) Discourse and the denial of racism. Discourse & Society 3: 87–118.

31.

Van Dijk

(1993) Principles of critical discourse analysis. Discourse & Society 4: 249–283.

32.

Van Dijk

(1996) Discourse, power and access. In: Caldas-Coulthard

Coulthard

(eds) Texts and Practices: Readings in Critical Discourse Analysis. New York: Routledge, pp. 84–104.

33.

Wiggan

(2012) Telling stories of 21st century welfare: The UK coalition government and the neo-liberal discourse of worklessness and dependency. Critical Social Policy 32(3): 383–405.

34.

Wodak

(1997) Das Ausland and anti-Semitic discourse: The discursive construction of the other. In: Riggins

(ed.) The Language and Politics of Exclusion. Thousand Oaks, CA: SAGE, pp. 65–87.