Table of Contents
2 Context of US Welfare Reform: 1996-2012
3 Literature Review
3.1 Discourse & Power
3.2 News Discourse
3.3 Marginalized Groups in the Press
3.5 Ideology & Welfare
3.6 Structural Explanations
3.10 News Discourse on Welfare
3.13 Research Questions
4.1 Content Analysis
4.2 Critical Discourse Analysis
5 Results & Discussion
On 22 August 1996, President Clinton signed into law the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), the most significant reform of the US welfare state in half a century. Considering that the law would drastically affect 12.6 million people on welfare (Moffitt & Gottschalk 2001), one might expect that the reforms arose from a widespread public understanding of recipients’ interests. In this dissertation, I critically analyze this assumption by looking at public discourse on welfare, particularly by examining the media’s illustration of recipients.
The press has generally claimed welfare reform to be a success despite complicated trends illustrating that for many in need of social assistance, life has gotten harder. Wealth and income gaps have been increasing between high- and low-income earners in the US for over a quarter-century (Piketty, Saez, & Zucman 2016), and although increased welfare spending has shown to decrease such inequality (Brooks & Manza 2007), public support for redistribution has stagnated (Ashok, Kuziemko, & Washington 2015). While public opinion opposed to social assistance is detrimental to those who need it, it is especially so for ethnic/racial1 minorities, who are disproportionately poorer than Whites (US Census Bureau 2016) despite elimination of legal racial discrimination. The racialization of poverty is not simply residue from the pre-Civil Rights era; as Brown (2013) posits, “the demise of de jure discrimination has done little to diminish the centrality of race in merican politics (͙) *and+ racial attitudes (͙) remain among the most powerful predictors of policy opinions and outcomes” (395). The complex entanglement of beliefs about inequality and race in determining public support for welfare begs a detailed understanding, so as to enable authentic discussions of poverty reduction in the US.
As minority populations on the welfare rolls increase and become more diverse, and redistribution becomes increasingly essential to low-income individuals, it is important to analyze the ideologies and values of the US welfare state in comparison with their context. To do this, I analyze the social construction of welfare recipients in US news discourse from 1996 until 2012. Specifically, I aim to uncover surreptitious biases that perpetuate stigmatizing, otherizing, or racialized associations with welfare recipients. To do so, I employ an integrated content and discourse analysis testing for presence of and patterns within these associations. Generally, my findings illustrate that stigmatization, otherization, and racialization remained prevalent within welfare discourse over the period of analysis; although functioning separately, they reinforced each other to ultimately portray welfare recipients, especially Blacks and Hispanics, as unlikely to warrant public support.
2 Context of US Welfare Reform: 1996-2012
The US welfare system is complex, and made up of many different social programs with varying regulations. The most funded programs are Temporary Assistance to Needy Families (TANF), food stamps, housing assistance, Medicaid, Supplemental Security Income (SSI), and the Earned Income Tax Credit (EITC) (Rector 2012). While the latter three programs are crucial to many low- income individuals, they are not included in my analysis as they target working, sick, disabled, and/or elderly people, who are classified as ‘deserving poor’ (Hunt & Bullock 2016, 107), whereas I focus on why able-bodied individuals are designated as ‘undeserving.’ The most notorious welfare program, TANF, provides cash assistance to families with the goals of promoting work and two-parent families (US Department of Health & Human Services 2017). Food stamps (or Supplemental Nutrition Assistance Program (SNAP)) provide a monthly allowance based on income (US Department of griculture 2017). The government’s primary social housing program, Section 8, consists of vouchers for eligible units in the local housing market (US Department of Housing & Urban Development 2017). The highest rates of participation in welfare programs occur for Native American (~1/4 of their population is on welfare), Black (~1/3), and Hispanic populations (~1/3), while Whites and Asians have the lowest rates (~1/10) (Moffitt & Gottschalk 2001). Since the 1996 welfare reforms, the proportions of Hispanics and Asians has increased relative to those of Whites and Blacks, which have remained relatively stable (Irving & Loveless 2015; Loveless & Tin 2006; National Center for Education Statistics 2011; Tin & Castro 2001).
PRWORA, signed into law in 1996, was based on a conservative-leaning criticism that welfare promoted government dependency and encouraged contemptible behaviors (Mead 1986); nonetheless, the bill had a degree of bipartisan and public support (Blank 2002). It replaced its predecessor, Aid to Families with Dependent Children (AFDC), with TANF, giving more financial responsibility to states, requiring work participation in exchange for benefits, and placing a five-year lifetime limit on benefit receipt (Moffitt & Gottschalk 2001). TANF was designed to be stricter on non-working, unmarried parents—particularly women with out-of- wedlock pregnancies (Moffitt & Gottschalk 2001). PRWORA also limited food stamp receipt to three months at a time and required recipients to work in exchange for assistance (Semuels 2016). The reforms denied TANF and food stamps to many non-citizens who had previously been receiving benefits (Brown 2013, 588).
In 2002, the policies set forth by PRWORA were set to expire and thus were up for reauthorization, with possible amendments. The Bush administration drafted reauthorization proposals based on ‘traditional’ family values (particularly, the male-breadwinner model) via reallocation of funds for ‘healthy’ marriage promotion (e.g. research, high-school abstinence education programs), and an increasingly punitive approach to welfare recipients (Daguerre 2008). These proposals helped shape reauthorization of PRWORA, which remained in this revised format until 2010 (Falk 2016).
Since then, a series of medium/short-term welfare extensions have been authorized, with only slight changes. One of the most important extensions which made changes to the welfare system was the Middle Class Tax Relief and Job Creation Act of 2012. Emphasizing prevention of fraud and enforcement of behavioral control, federal regulation of welfare was tightened. The act extended TANF funding for one year, required states to prevent cash assistance from being withdrawn at illicit locations, and was no longer usable on loans or debts (Falk 2016; US Congress 2012).
Welfare reform was considered a success by the political establishment, as TANF receipt declined 75% from 1996 to 2014 (Edin & Shaefer 2016). However, less than one-quarter of those in poverty now claim TANF benefits, and the reduction in poverty has been less than the reduction in welfare caseloads (Blank 2002; Edin & Shaefer 2016). While labor participation of single mothers increased by 10% for the first five years after reform, it ultimately declined from 2001-2014 (Blank 2002; Ehrenfreund 2016). Chang (2016, 4) argues that “there is no evidence yet that welfare reform has substantially improved the real lives of clients.” Rather, as states’ priorities shifted from comprehensive assistance to getting people into work as soon as possible (e.g. money allocated from long-term training to job search assistance) (Blank 2002), poverty deepened for those with routine difficulties finding jobs who also lost access to most benefits (Semuels 2016).
3 Literature Review
3.1 Discourse & Power
According to Foucault, discourse delimits our ways of thinking, structuring how we interact, both socially and practically (Hall 1997, 45). After all, discourse represents and delineates what can be said in society (I refer to this as a ‘discursive corridor’) and is often dictated by a select number of powerful interests (Foucault 1978). Thus, it is not neutral, passive, nor equitable, and should not be taken for granted. In fact, marginalized groups are often excluded from public discourse (Fraser 1990), not least because of the delineating function of elite media corporations. As media influence becomes more competitive, content is increasingly tailored to the demands of profitability, which are incongruent with the preferences of low-income individuals and non- White minorities of lower economic and political clout. This results in ‘consumer-oriented media’ with little interest in poverty, minorities, or especially the intersection of the two (Deane 2005, 184).
Discourse is not just responsive, but also productive. In adapting social realities to the interests of the powerful, language can be a crucial tool (Bourdieu 1990), especially when employed in strategic fashion through an institutionalized channel such as the media. For Chomsky (1997), elite media in particular create the frameworks which other news organizations use when choosing and describing topics and to which the public and policymakers refer for information. Especially as media ownership becomes more concentrated, media elites have significant influence in defining the discursive corridors of public debate, including on ethnic/racial topics (Chomsky 1997; van Dijk 1993a). This also means they have substantial influence on public opinion and subsequently public action regarding electoral and policy decisions (Wodak & Meyer 2001).
Patterns of discursive corridors constitute dominant frames and ideologies, which are often obscure, hidden in patterns of concepts and syntax (Wodak & Meyer 2001, 8). They often appear neutral, and their assumptions remain largely uncontested (Gramsci 2014). Thus, the reproduction of discursive patterns over time necessitates the participation of both those in power and the public, even if the objectives of such discourse are primarily in the interests of the former. Although counterideological resistance can result in societal change over time (Foucault 1978), this dissertation focuses on the macro-level reproduction of inequalities and therefore emphasizes those ideologies which are most societally influential.
3.2 News Discourse
The news media are meant to inform on issues of public affairs. As these issues are often a vague and distant reality for many people, our knowledge about the world is primarily drawn from the media. Specifically, the media attributes relative importance to certain issues over others (agenda-setting) and emphasizes certain attributes about them (McCombs 2000). Agendasetting is about what is presented to the audience, what isn’t, and the political consequences which arise from that (McCombs & Shaw 1972).
Framing, on the other hand, concerns how issues are presented. The media focus on certain aspects of topics; the way in which that is done influences how people process and act upon that information (Avery & Peffley 2003; Goffman 1974). As Nelson, Oxley, and Clawson (1997, 569) put it, “frames influence opinions by stressing specific values, facts, and other considerations, endowing them with greater apparent relevance to the issue than they might appear to have under an alternative frame.” For example, welfare can be framed either as an unearned handout or as a needed safety net. Additionally, frames tend to be reproduced, as they provide templates to other journalists to quickly package information for their audiences (Scheufele 1999). Assuming that public opinion has an impact on welfare policy (Brooks & Manza 2007), the way the media set the agenda of and frame welfare is crucial to how the social safety net evolves over time. That is not to say, however, that the effect of media discourse relies on a homogeneous audience. While elite news outlets contribute to the (re)production of a dialectically-formed dominant ideology, discourse is ultimately shaped by existing cultural frames and political climates, which are in return shaped by discourse (Hall 1978; van Dijk 1993a, 1993b, 2001, 2012; Wodak & Meyer 2001).
3.3 Marginalized Groups in the Press
Within the elite press, marginalized groups are unlikely to be selected as newsworthy unless they fit stereotypes as dictated by dominant ideology (Harley 1982), or are discursively differentiated from dominant groups (Hall 1978). This is either out of elites’ lack of interest in or awareness of the realities of those with less political/economic agency, or is a strategy of power reproduction of elites to ensure hegemonic control over resources (van Dijk 1993a). The status quo power structure is supported by discursive news patterns, particularly via stigmatization of minority characteristics and/or emphasis of their cultural differences from majority groups (van Dijk 1990, 1993a). The in-group/out-group distinction is not just a strategy that serves the interests of elites; it is also used to achieve positive identification between journalists and readers, who are more likely to be of the majority group (van Dijk 1985). These dynamics within the press are potentially detrimental to public perceptions of minorities, as significant proportions of the white majority have little or no alternative sources of information than the news for racial/ethnic issues (van Dijk 1993a). The limiting of topics regarding the marginalized, particularly non-Whites and low-income individuals, means that their struggles are less likely to be framed accurately (if at all) due to the absence of their perspectives and situations in the media, putting effective public support and policies in their interests at stake.
Before delving into discussions of welfare, I first define ‘racism,’ as it is an integral concept to understanding the dynamics of welfare discourse in the US. Using van Dijk’s (1993a) classification, racism is a relation between racial/ethnic groups based on historically rooted patterns of dominance, particularly of Whites over ‘others.’ Racism has four key features, namely: 1) the use of stereotypes to emphasize ‘them’ and ‘us’; 2) one racial/ethnic group having preferential access to material and symbolic resources (and is attempting to maintain the exclusivity of that access); 3) it is socially constructed, especially through otherization and stigmatization; and 4) it is historically acquired. Racism is often informed by systemic ideologies that help (re)produce racial inequalities, and thus may be either intentional or unintentional. As overt racism has become unacceptable in public discourse, racist terminology is generally avoided and has been replaced by indirectly racist code-words, such as ‘welfare mother’ (van Dijk 2000, 39). Racism is not innate, but can be naturalized through discursive repetition over time. Therefore, although discourse appears to be race-neutral, underlying strategies of power or lingering patterns of racial dominance aid in reproducing racial inequalities.
3.5 Ideology & Welfare
The US welfare state is a two-tier system which prioritizes the elderly, disabled, and sick over able-bodied adults, leaving the latter more vulnerable to stigmatization (Daguerre 2008; Hunt & Bullock 2016). This circumstance can be explained largely by the increasing dominance of both neoliberal and conservative ideologies of welfare. After the onset of globalization and technological developments in the 1970s/80s, American elites largely decreased their support for the welfare state, as they were becoming less reliant on and accountable to US workers; additionally, employers were concerned that welfare was becoming more attractive than the lowest-paid jobs ( bramovitz 2006; Schram 2006). PRWOR “represented the culmination of a neoliberalism discourse that emphasizes free and private markets” and personal responsibility for one’s well-being (Lens & Cary 2010, 1037). As a result, the welfare state started to be blamed for financial barriers to economic growth, specifically for encouraging people to claim benefits instead of work and undermining individual responsibility and entrepreneurial work ethic (Abramovitz 2006; de Goede 2016).
Neoliberal criticisms of welfare paired well with the moralized rhetoric of US conservative ideology regarding welfare. The Reagan administration increasingly associated recipients with ‘bad’ behavior and family structure, as opposed to the ‘admirable’ work ethic and family structure of middle- and upper-classes, thus linking recipients’ morals and behaviors to the breakdown of ‘traditional merican society’ (de Goede 2016). Influential scholars, such as Charles Murray and Lawrence Mead, helped legitimize this narrative. Murray argued that welfare ‘paid the poor’ to be poor, and that recipients became accustomed to the alleged laziness and irresponsibility of ‘dependency culture’; additionally, he claimed that welfare threatened the traditional, heterosexual, male-breadwinner family model (Abramovitz 2006). Similarly, Mead contended that the government’s lack of authority over welfare recipients led them to become greedy and entitled, therefore they needed enforcement to work in return for social support or be punished (Sparks 2003, 188). Despite the bias of these beliefs, Democrats increasingly endorsed them, mainly to pander to a changing voter base (Daguerre 2008, 376).
Consequently, conditional assistance became central to the US welfare system, disregarding most structural factors explaining welfare need and unemployment.
Beliefs based on blaming individuals for poverty are widely present among the American public. While Americans generally do not reject the notion of supporting those who need assistance, certain types of welfare recipients are systematically otherized and stigmatized, influencing public support for public assistance (Gilens 1999). In other words, Americans support welfare as long as they believe it targets those who deserve it. In fact, research has shown that when individuals’ beliefs are consistent with dominant neoliberal/conservative ideology, they are less likely to support higher levels of redistribution, while counterideological beliefs are correlated with higher support for welfare (Kluegel & Bobo 1993).
3.6 Structural Explanations
Individualized accounts of poverty and welfare need are countered by structural explanations, which focus on macro-level sources of disadvantage: firstly, they refer to discrimination in labor and housing markets against welfare recipients, particularly those of color (Holzer & Stoll 2003; Moffitt & Gottschalk 2001). Minorities are less likely to be referred to educational, transportation, and childcare support and more likely to be drug tested and be sanctioned by caseworkers than Whites (Gooden 2003; Moffitt & Gottschalk 2001). Additionally, changes in the labor market have led to the strengthening of capital’s power over labor, and particularly increased the precarity of low-wage jobs (Reisch 2006); the low-wage service sector is the largest-growing sector in the labor market, and does not tend to provide benefits, pensions, or healthcare (Morgen, Acker, Weigt, & Gonzales 2006). This impedes the overall well-being of recipients, along with their ability to find quality jobs and keep them. Although there are certainly more factors to be accounted for, this general outline of structural patterns within the welfare system highlights many issues that behavioral explanations of poverty cannot account for.
The ideological shifts relating to welfare in the 1980s alleged a difference in human nature between the working class and middle- and upper-classes; the former were increasingly framed as responding primarily to punishment, whereas the latter were framed as responding to rewards (e.g. profits) (Schram 2003, 218). Furthermore, recipients were characterized as threats to the ‘ merican values’ of entrepreneurial work ethic and White, middle-class culture (de Goede 2016, Lens 2002). Such an emphasis on welfare recipients as ‘others’ justified blaming the worse-off for their situations, and is “designed to make racism, sexism, poverty, and other forms of social injustice appear to be natural, normal, and inevitable parts of everyday life” (Collins 2009, 69). Because welfare recipients make up a relatively weak political constituency and are less likely than non-recipients to have an influence in resisting such biased depictions, their situations are vulnerable to misrepresentation, making welfare especially prone to ideological spin (Daguerre 2008, 376; Piven 2003).
One of the most prominent arguments for individualized welfare need relied on ‘bad family structure,’ claiming that welfare undermines the structure of the family, as single mothers avoid marriage in order to retain benefits (de Goede 2016). The ‘welfare queen’ or ‘welfare mother’ became a symbol of female welfare recipients (Collins 2000), painting them as irresponsible, greedy, and promiscuous (Sparks 2003) and decreasing positive identification among the general public with recipients overall (McCormack 2004, 356). The ‘deadbeat dad’ was another caricature of recipients, characterized by a male who has no authority over his household (his family can get by on welfare benefits), hence turning to violence, crime, and drugs/alcoholism (de Goede 2016, 326). Because about half of households in poverty are headed by single mothers, this association lends itself to assumptions of a causal relationship. However, there is no evidence that this is the case (see Abramovitz 2006; Badger 2014). Other common stereotypes include laziness, short-sightedness, and greed (Gilens 1999; Schram 2003). The continued usage of such biased exaggerations is problematic because it justifies a faulty association between welfare and choice.
Historically, race has been conceptually embedded in the US welfare system (Schram & Soss 2003). Racialized understandings of AFDC made racial stereotypes central to public opinion on and debate around the 1996 reforms (Lieberman, 2003; Reisch 2006), partially because many conservative arguments overtly implicated Blacks. Illustrating the prevalence of individualized beliefs of welfare, Whites tend to think that other Whites would prefer to be self-sufficient rather than live off welfare, but are slightly more likely to think that Asians would prefer welfare, and significantly more so for Hispanics and especially Blacks (Fox 2004). While welfare receipt by itself is a marker of stigma in the US, being a minority adds an extra layer of disadvantage (Brown 2013).
Much of the racialization, stigmatization, and otherization of welfare is reproduced passively, but it may also serve strategic purposes for Whites and/or the wealthy, particularly vis- à-vis social prestige, political power, and economic accumulation (Neubeck and Cazenave 2001). A large, low-wage labor pool has the potential to reinforce the current economic and political order by reducing the chances that the disadvantaged will collectively advocate for policies more suited to their interests (Piven 2003, 323). Thus, the momentum of these associations with welfare may be partially sustained by powerful groups who benefit from the status quo.
3.10 News Discourse on Welfare
Following PRWORA, the press focused on the decrease in welfare receipt. However, the consequences of the reforms as well as other structural factors contributing to poverty were sidelined (Hunt & Bullock 2016, 94; Kilty & Segal 2006, xix; Pardee 2006, 136), marking a shift further away from the interests of those in need of welfare.
When it comes to race, research has shown that the press plays an active role in (re)producing racialized associations with and stereotypes of welfare. Van Dijk (1990, 1993a, 1993b, 2000) illustrates that racism is often denied or unmentioned, but minorities themselves tend to be framed negatively, as ‘others,’ and/or without reputable agency.
1 Throughout this dissertation, I use this hybrid term due to the problematic nature of race, which has been shown to be a social, as opposed to biological, construct (American Sociological Association 2003)