Three Paradigms of Social Assistance

“Ideas,” which are defined as the normative and cognitive beliefs of actors, are fundamental to a full understanding of the welfare state and, in particular, of social assistance. However, policy ideas have been neglected in most typologies of social assistance regimes. Based on a selective review of the literature, this article proposes a brief but systematic analysis of policy paradigms in the field of social assistance. Three ideal types that emphasize the ideational dimension of social assistance are analyzed, namely, the entitlement, workfare, and activation paradigms. The value of the typology lies in its utility for characterizing the ideational orientation of social assistance regimes. Specifically, the typology provides a yardstick for measuring the ideas of policy actors with respect to social assistance and can facilitate the conduct of case studies, comparative research, and causal analyses on this policy sector.


Introduction: Welfare Reform in Ideational Perspective
Social assistance (welfare) is a fundamental component of the contemporary welfare state. In a very concrete way, social assistance is the "last safety net," that is, a last-resort income support program to which citizens can turn when they have exhausted all other options (Bahle, Hubl, & Pfeifer, 2011). Since the early 1980s, social assistance has been the object of substantial restructuring under the impulse of powerful structural forces such as the relative growth of the service sector, massive increase in female employment, globalization, population aging, and sustained budget deficits, which have led to an environment of "permanent austerity" (Esping-Andersen, 2002;Pierson, 2002;Taylor-Gooby, 2004). Although these structural forces are important to understand welfare reform, they only tell part of the story. Indeed, politics and the "ideas" held by decision-makers, in particular, have contributed to exacerbate the perceived need for reform and to shape its specific content (Huo, 2009). Another reason why welfare reform has occupied such a high position on the political agenda of many governments is that clients have been negatively "constructed" as a target group (Schneider & Ingram, 1993). In fact, deeply entrenched ideas about what the government should do, how, and why, prevail in the field of social policy (Deacon, 2002;Rice & Prince, 2013;Somers & Block, 2005;Stryker & Wald, 2009;van Oorschot, 2006): If we really wish to understand the dynamics of welfare reform, we need to identify the cognitive and normative orientations of policy actors (Scharpf, 1997b) [ . . . ] Policy reform is not merely "a contest for power" but depends also on "playing with ideas," the extent to which policy actors agree over the cognitive definition of policy problems, and the normative benchmarks for effective policy solutions. (Hemerijck, 2002, pp. 187-188) Ideas-the normative and cognitive beliefs of actors-are thus fundamental to a full understanding of policy dynamics (Béland, 2005(Béland, , 2009). However, studying welfare reform from an ideational perspective is challenging. First, few typologies have focused on social assistance regimes, let alone on the ideas that underlie them. Even the typologies specifically geared toward social assistance (e.g., Boychuk, 1998;Deacon, 2002;Gough, Bradshaw, Ditch, Eardley, & Whiteford, 1997) have emphasized institutional features over ideas. Indeed, most of the literature has taken the welfare state as the unit of analysis, by opposition to specific domains or programs such as social assistance, and when it has, the welfare state's institutional characteristics have been emphasized to the detriment of ideas such as values, causal beliefs, and assumptions (Abrahamson, 1999). Other typologies, which can be qualified as more "idea-friendly," have focused on social policy in general or larger constructsalbeit related-such as activation, Third Way political ideology, social exclusion, or social investment, but not specifically on social assistance (e.g., Barbier & Ludwig-Mayerhofer, 2004;Dean, 2004;Levitas, 2005;Morel, Palier, & Palme, 2012;Weishaupt, 2013). Thus, there are few, if any, typologies specifically centered on social assistance ideas. This leaves scholars interested in characterizing and comparing social assistance regimes with few options at their disposal but to proceed inductively or to develop their own framework.

Thinking Paradigmatically About Social Assistance
The concept of policy paradigm can serve as a useful starting point for building a typology of social assistance ideas, in particular because many scholars have characterized the result of recent social policy reforms as a "paradigm shift" (e.g., Barbier & Ludwig-Mayerhofer, 2004;Cox, 1998;Huo, 2009;Jenson & Saint-Martin, 2006;Weishaupt, 2013). Policy paradigm refers to . . . a framework of ideas and standards that specifies not only the goals of policy and the kind of instruments that can be used to attain them but also the very nature of the problems they are meant to be addressing. (Hall, 1993, p. 279) Paradigm shifts, which occur when there is fundamental change in policy goals in addition to policy instruments, are characterized by "radical changes in the overall terms of policy discourse" (Hall, 1993, p. 279). Despite their heuristic value, however, these concepts have not been defined or operationalized with sufficient precision (Daigneault, 2013;Kern, Kuzemko, & Mitchell, 2013). Moreover, a frequent confusion pertains to what precisely changes when there is a paradigm shift, namely ideas or policy. Using policy content as an indicator of a paradigm shift introduces circularity problems pertaining to the assessment of the impact of ideas on public policy (Daigneault, 2013). Conducting rigorous causal analyses on social assistance is especially important given the recognized need to move from descriptive to causal analyses on this topic (Marx & Nelson, 2013). Thus, although the concept of policy paradigm is useful to understand welfare reform, its successful application within a comparative or case study of social assistance is premised on a rigorous operationalization that makes its measurement both reliable and valid.

Purpose
This article proposes a selective review of the social policy literature that aims to develop a typology exclusively centered on policy ideas on social assistance, in contrast with expenditure levels or other institutional features. The value of this typology lies in its potential use to assess the alignment of policymakers' and citizens' ideas with one (or more) social assistance paradigms, as well as their evolution in time, including claims that a paradigm shift has occurred. Moreover, by establishing a tighter distinction between ideas and institutional features of social assistance, this classification scheme can contribute to strengthening causal analyses that examine the influence of ideas on policy change.

Conceptual Framework
The conceptual framework that serves to structure the literature review contains two main components. The first component is Daigneault's (2013) operationalization of the concept of policy paradigm, which contains several propositions intended to improve the rigor of ideational research on policy dynamics. Although it is beyond the scope of this article to discuss these propositions in detail, a couple of points must be noted. On one hand, "policy paradigms are normative and cognitive ideas, intersubjectively held by policy actors" (Daigneault, 2013, p. 461). In other words, policy paradigms are "made of" ideas, and the latter must display internal coherence. On the other, the concept possesses four fundamental dimensions, along which social assistance ideas are organized: i. Values, assumptions, and principles about the nature of reality, social justice, and the appropriate role of the state; ii. Comprehension of the problem that requires public intervention; iii. Ideas about which policy ends and objectives should be pursued; iv. Ideas about appropriate policy "means" to achieve those ends (i.e., implementation principles, type of instruments, and their settings).
A paradigm shift is operationally defined as a significant change on each of these four dimensions. Finally, from a methodological standpoint, this framework affirms that "descriptive studies of policy paradigms should primarily rely on direct evidence of policy actors' ideas and beliefs, which can be supplemented with, but not replaced by, indirect evidence of policy change in the expected direction" (Daigneault, 2013, p. 463). In other words, to establish that a paradigm characterizes social assistance in a given jurisdiction, it is necessary to examine the ideas of relevant policy actors within that jurisdiction.
The second component of the conceptual framework follows Levitas' (2005) three discourses of social exclusionredistributive (RED), moral underclass (MUD), and social integrationist (SID)-and develops them into three comprehensive paradigms along the four dimensions outlined above. According to this framework, [P]aid work [is] a major factor in social integration; all of [the discourses] have a moral content. But they differ in what the excluded are seen as lacking. To oversimplify, in RED they have no money, in SID they have no work, in MUD they have no morals [. . .] the discourses emphasize different elements-and posit different causal relationships between them. (Levitas, 2005, p. 27) Using Levitas' (2005) characterization of social exclusion discourse as a starting point is easily justified, as her concept of "political discourse" is very similar to the concept of "policy paradigm" coined by Hall (1993) and further developed in Daigneault (2013Daigneault ( , 2014. Indeed, like policy paradigms, (political) discourses are interpretive frameworks and promote certain values over others; they propose a conception of the problem to be solved by government as well as a set of policy prescriptions: It means that sets of interrelated concepts act together as a matrix through which we understand the social world. As this matrix structures our understanding, so it in turn governs the paths of action which appear to be open to us. A discourse constitutes ways of acting in the world, as well as a description of it. (Levitas, 2005, p. 3) Furthermore, a threefold typology allows for more finegrained distinction than twofold models (see S. Morel, 2002), while remaining relatively simple and user friendly.
The specification of the three paradigms of social assistance draws on a selective literature review that includes the works of Bane and Ellwood (1994); Barbier (2004); Dostal (2008); Gilbert (2002); Huo (2009); Lødemel and Trickey (2001); Morel (2002); Serrano Pascual (2007); van Berkel and Møller (2002b); and Weishaupt (2013), among others. Although the discussion of the various conceptions of social assistance borrows heavily from existing research, it is novel by the way in which it combines them using a systematic operationalization of the concept of policy paradigm (Daigneault, 2013; see Table 1 for a comparison of the three paradigms). It must finally be noted that the three paradigms are ideal types in that they deliberately accentuate various attributes of social assistance and that no perfect correspondence with reality is to be expected. High: segmentation of assistance between "deserving" and "undeserving" clients High: income supplements are restricted to clients who comply with work-related conditions

Three Paradigms of Social Assistance
The Entitlement Paradigm The first paradigm, which is grounded in the redistributive discourse (Levitas, 2005), takes its roots in social democratic thinking. Within this conception, social assistance is seen as a matter of right that is guaranteed and should be upheld by the state to all its citizens. A fundamental aspect regarding this right to income is that it "does not depend upon the economic value of the individual claimant" (Marshall, 2009, p. 36). Social rights are thus seen as the key to the social inclusion of citizens. In terms of social justice, this paradigm prioritizes the values of solidarity and egalitarianism (Huo, 2009). In terms of the welfare triangle composed of the state, market, and family (Boychuk, 1998;Esping-Andersen, 1990), the state clearly has the most important role. The rationale for public intervention in this conception of welfare is economic insecurity, poverty, and inequality; in short, public authorities must address the fact that some people "lack material resources." These problems are seen as being caused by structural forces linked to the economy and society; as a result, individuals cannot be held accountable for their situation. The policy objective that characterizes the entitlement paradigm is to reduce poverty by guaranteeing a decent level of income. Moreover, the aim is to attain a certain level of "decommodification" for individuals with respect to welfare provision; or, in other words, allowing individuals to make a living without depending on the labor market (Esping-Andersen, 1990). The ideas for achieving the policy ends in the entitlement paradigm are centered on "passive assistance," that is, unconditional cash transfers to all eligible individuals (Barbier, 2004; van Berkel & Møller, 2002a). It also holds that income transfers must be sufficiently generous to sustain the social rights of individuals. Moreover, while transfers could be universal under the entitlement paradigm, depending on whether or not a guaranteed basic income is favored (see Steensland, 2008), eligibility is defined broadly and few distinctions are drawn between types of claimants.

The Workfare Paradigm
The second paradigm, which is structured on the moral underclass discourse (Levitas, 2005), is aligned with a conservative political ideology. In this approach, individual liberty-defined negatively à la Isaiah Berlin as the absence of interference from state intervention-and individual responsibility are considered paramount values for an individual's welfare. Although the state must help those who cannot fend for themselves (i.e., "deserving poor," such as severely handicapped people), its role with respect to those who can (i.e., "undeserving poor" such as single employable individuals) is severely limited; markets and families thus have a primary role in supporting individuals. Proponents of the workfare paradigm endorse inequality as a natural fact of life. Although poverty and material deprivation are considered policy problems under all three paradigms, the main focus within the workfare paradigm is the "culture of dependency" (see Bane & Ellwood, 1994;Lødemel & Trickey, 2001) and the poor work ethic of individual claimants: Economic dependence on "welfare" was construed as "dependency," a pathological moral and psychological condition created by the benefit system itself-and fostered by the libertarianism of the 1960s-in which the state was seen as a universal provider, sapping personal initiative, independence and self-respect. Benefits were bad for, rather than good for, their recipients. (Levitas, 2005, p. 15) Although dependency has its origins in the welfare system, those who rely on welfare are blamed for it (on the "perversity thesis," see Somers & Block, 2005). The policy ends that are pursued under the workfare paradigm aim to discourage reliance on social assistance and to improve the work ethic, attitudes, and self-esteem of welfare clients. Ideas related to the policy means within the workfare paradigm include the principle of segmenting assistance between various categories of "deserving" and "undeserving" claimants. Segmenting assistance based on a claimant's perceived level of deservedness increases the stigma borne by recipients. Another element is the principle of "less eligibility" defined as the "practice of making welfare so odious that it was less attractive (eligible) than even the most demeaning life without it" (Somers & Block, 2005, p. 267). This principle entails, especially so for "undeserving claimants," substantially lower welfare rates and the introduction of various control measures (e.g., having to pick up one's check in person, frequent and intrusive verifications of one's eligibility) that have been characterized as "rituals of degradation" (Herd, Mitchell, & Lightman, 2005). Moreover, the receipt of social assistance is made conditional on compliance with various work-related requirements (Dufour, Boismenu, & Noël, 2003;Hibbert, 2007). The most important of these conditions is undoubtedly "workfare," defined as the requirement to work, for instance, in subsidized jobs in the public, private, or nonprofit sectors, in exchange for social assistance benefits. In short, the concept of workfare possesses three dimensions: It is compulsory, primarily about work (rather than training), and linked to social assistance or last-resort income support (Lødemel & Trickey, 2001).

The Activation Paradigm
The activation paradigm was coined after the active society ideal that has been promoted for the last 20 years by the Organization for Economic Cooperation and Development (OECD; Carcillo & Grubb, 2006;Martin & Pearson, 2005;Walters, 1997). Activation corresponds to the social integrationist discourse in Levitas' (2005) typology. Moreover, the values, principles, and assumptions of this paradigm have been associated with those of the "Third Way" ideology as it applies to social policy (Eichhorst & Konle-Seidl, 2008;Huo, 2009;Levitas, 2005). First, the value of "reciprocity" is paramount; social assistance is perceived as a contract between the state and individuals and the conception of social rights has moved toward individual responsibility (Bahle et al., 2011;Cox, 1998;White, 2004). As such, the new balance of individual rights and responsibilities in the activation paradigm fits in between those of the entitlement and the workfare paradigms (Etzioni, 2000;Lindsay & Mailand, 2004). As Anthony Giddens (1998), who coined the term Third Way, argued, One might suggest a prime motto for the new politics, no rights without responsibilities. Government has a whole cluster of responsibilities for its citizens and others, including the protection of vulnerable people. Old-style social democracy, however, was inclined to treat rights as unconditional claims. (p.

65, emphasis in original)
In the activation paradigm, the role of the state and markets with respect to welfare is predominant. This ideological orientation is consistent with the Third Way, which justifies state intervention provided that it is strategic and targeted; the state should supplement rather than replace the market (Giddens, 1998). In addition, the values of "prioritarian egalitarianism," "equality of opportunity," and "meritocracy" are very important within the activation paradigm (Huo, 2009;White, 2004). On one hand, the role of the state is to protect the worst-off; that is, to "shield" them from the more serious risks, including poverty and economic insecurity. On the other, activation ideas are geared to the notion that individuals are responsible for their well-being beyond a "decent floor of opportunity" (White, 2004, p. 36). This conception also emphasizes economic productivity, which signals closer ties between economic and social policies (Huo, 2009). Fiscally prudent financial management frees up the resources required for an ambitious social policy, which, in turn, contributes to the performance of the economy. The first policy problem emphasized within this paradigm is the lack of incentives for work, the so-called "welfare wall" (also known as the "welfare trap" or "poverty trap"), by which the earnings social assistance clients can earn, combined with the taxes they must pay and the subsidies they lose, render work unattractive (Battle & Mendelson, 2001). A second problem that social assistance addresses under the activation paradigm is insufficient human capital; social assistance clients cannot find work because they do not possess the required knowledge and skills to make them attractive to potential employers.
On the policy-ends dimension, activation favors enabling welfare recipients to work, boosting the economic activity rate and reducing poverty for low-income workers (Barbier & Ludwig-Mayerhofer, 2004;Huo, 2009). Although there has always been a significant connection between welfare and the labor market, the ideas of the activation paradigm makes this connection more explicit and systematic. "Enabling to work" means that welfare recipients must be trained, formally or informally, to increase their human capital and thus their attractiveness for potential employers. This policy end is a facet of the larger social investment approach to social policy (European Commission, 2013;Jenson & Saint-Martin, 2006;Morel et al., 2012). As a result of this diversity in ends, the activation paradigm commands the use of various types of policy instruments. First, because of its unintended effects in terms of work disincentives, the belief is that cash benefits under "passive assistance" should be significantly lower than those under the entitlement paradigm. In addition, rather generous income support, both in cash and in kind (e.g., extended health benefits), should be provided for low-income workers, especially single parents. The incentive structure should be altered by using both enabling (income support to top-up earnings) and demanding elements (conditions and requirements; Eichhorst & Konle-Seidl, 2008;Kenworthy, 2010;Lødemel & Trickey, 2001;Nybom, 2013) or, in Weishaupt's (2013) terminology, "push" and "pull" factors. Various human capital development measures such as formal education, job training, and employability measures (career counseling, job search assistance, etc.) should also be provided.

Conclusion
Based on a selective review of the social policy literature, this article proposed a typology of social assistance that is centered on ideas. The typology is structured around the concept of policy paradigm as put forward by Peter Hall (1993) and operationalized by Daigneault (2013). Three ideal types of social assistance have been outlined, namely the entitlement, workfare, and activation paradigms. Although this typology is only a first-cut that will need to be refined, it will facilitate the characterization of social assistance from an ideational perspective.
At this stage, an empirical application of the typology is warranted. It could be used to study the "fit" between the ideas of important policy actors and the three social assistance paradigms, keeping in mind that there is rarely (if ever) a perfect fit between ideal types and real-world cases (Bernard & Saint-Arnaud, 2004). Notably, it could be used in comparative studies to assess whether the significant institutional variations found in national/subnational social protection regimes (Bahle et al., 2011;Boychuk, 1998;Marx & Nelson, 2013) are associated with variations in the ideas of policymakers and/or citizens. Because the typology has potential for analyzing the influence of ideas on public policy, it would be interesting to examine whether the erosion of social assistance benefits in the last several decades in Europe (van Mechelen & Marchal, 2013) is the result of a shift from the entitlement paradigm to workfare or activation.
Significant methodological challenges related to the application of this typology lie ahead, however. First, directly measuring policy ideas implies having access to the relevant policy actors to interview or observe them. If policy actors are not accessible, then their ideas could be identified through official documents such as policy statements, discussion papers, or reports of political debates. Second, policy actors can have memory lapses and cognitive biases, and can purposefully distort reality to their advantage (Campbell, 2004). Third, assessing the internal coherence of policy ideas and their alignment with all four dimensions of the concept of a policy paradigm can be tedious, and even more so if the ideas of different periods are studied.
Still, these challenges should not be exaggerated. Whereas getting access to high-level policy actors is not always easy, people are usually willing to share their perspective on a given policy issue. Moreover, triangulating respondents, data sources, and methods of analysis, as well as "confronting" respondents with contradictory evidence, will improve the credibility of the analysis. Similarly, assessing the alignment of policy ideas with a given paradigm requires judgment calls but is nevertheless feasible, provided that quality evidence has been assembled.
Even more than other policy sectors, social assistance is characterized by deeply entrenched ideas about what the government should do, why, and how. If we neglect to consider these ideas, only a partial understanding of welfare reform can result. The three policy paradigms presented in this article provide a stepping stone toward a more systematic study and evaluation of the values, beliefs, and cognitive filters of actors with respect to social assistance.

Acknowledgment
I wish to thank Daniel Béland and Guy Fréchet for their comments on an earlier version of this article as well as Sara McPhee-Knowles and Elizabeth Winokur for linguistic revision.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research and/or authorship of this article: Postdoctoral funding from the Fonds de recherche du Québec-Société et culture, from the Canada Research Chair in Public Policy, Johnson-Shoyama Graduate School of Public Policy, University of Saskatchewan Campus, and from the Chaire de recherche du Canada sur l'évaluation des actions publiques à l'égard des jeunes et des populations vulnérables (CRÉVAJ), École nationale d'administration publique, is acknowledged.