The New Paternalism

1. Two Accounts of Paternalism

Paternalism has justifiably been a key concern for autonomy theorists. It cuts at the very core of what autonomy refers to: the capacity to determine one’s own life plans. Paternalistic policy programs hinge on the claim that some individuals ought not to have the opportunity to exercise this capacity in given contexts, where such constraint has been deemed to be in accordance with the individuals’ “own good.” How, then, can such a claim be justified if autonomy is accepted as (at least) one primary value in our society? This is the question that political theorists have grappled with in their treatments of paternalism. In pursuing this problem, the question of state intervention into citizens’ lives has often been a proxy for the question of paternalism.

What I refer to here as “conventional” accounts of paternalism often begin from the work of John Stuart Mill, the great advocate of individual liberty (Mill and Rapaport 1978). Gerald Dworkin (1971, 108) defines paternalism as “the interference with a person’s liberty of action justified by reasons referring exclusively to the welfare, good, happiness, needs, interests or values of the person being coerced.” Explaining Mill’s opposition to paternalistic measures, Dworkin emphasizes that Mill views paternalism as an affront to the essence of what it means to be a human being, which is deeply tied to recognition as an autonomous agent. Dworkin writes, “It is the privilege and proper condition of a human being, arrived at the maturity of his faculties, to use and interpret experience in his own way” (p. 117).

Autonomy, this core human characteristic as Dworkin frames it, can be understood as both an ascribed status and a capacity, the two of which are closely related but distinct. ⁴ It is a status insofar as being recognized as an autonomous individual has particular social and political meanings. It is a capacity insofar as it refers to an individual’s ability to determine her or his own ends. On both accounts, I suggest, autonomy emerges out of social relations. Our capacities are developed or restricted in the context of enabling or constraining arrangements of social relations. Our status as autonomous agents is often constituted by larger social relations that configure the distribution of recognition and respect in our society: institutional, cultural, and market relations, among others. ⁵ Where forms of misrecognition restrict the latter notion of autonomy, the former notion may also be impeded since the development of the capacity for autonomy is closely tied to self-esteem or self-respect—qualities that misrecognition may quash.

The conventional critique of paternalism addresses infringements on each of these dimensions of autonomy in some respects. Insofar as autonomy is ascriptive, Dworkin and other liberal theorists’ critiques of paternalistic measures suggest that these measures fail to ascribe autonomy to their targets. Dworkin’s (1971, 117) exegesis of Mill proceeds as follows: “It is because coercing a person for his own good denies [his] status as an independent entity that Mill objects to it so strongly and in such absolute terms. To be able to choose is a good that is independent of the wisdom of what is chosen.” The target of paternalist measures is unjustly denied her or his status as a “chooser,” which is equivalent on these accounts to being denied her or his status as an autonomous individual. Moreover, on this account of the potentially deleterious effects of state intervention into citizen lives, the opportunity to develop and exercise autonomy as a capacity is also limited. Dworkin’s account demonstrates this when he refers to the most plausibly palatable forms of paternalism as couched in “a concern not just for the happiness or welfare, in some broad sense, of the individual but rather a concern for the autonomy and freedom of the person” (p. 125). He goes on to suggest that a potentially justifiable form of paternalism “preserves and enhances for the individual his ability to rationally carry out his own decisions” (p. 125).

This last statement by Dworkin points to the source of criticisms of the conventional account of paternalism raised by feminist and other theorists. Insofar as Dworkin suggests it is possible for paternalism to be autonomy fostering—or at least autonomy preserving—what is noteworthy is that such a configuration of relations is still termed “paternalism.” The Dworkin view, Marion Smiley (2004, 308) writes, leads us to view “all forms of government protection as paternalistic.” She explains, the accepted definitions “ignore the context of paternalistic choice-making—or in other words, the relationships of domination and inequality that exist between a paternalist and those subject to paternalistic treatment” (p. 308). The problem with paternalism, then, is not only or entirely its infringement on individual free choice but the fact that it “perpetuates (or at least expresses) relationships of domination and inequality among individual members of a community” (p. 309). With this problem acknowledged, Smiley suggests that we need not accept the assumption that all government protection is paternalistic, including that autonomy-enhancing form referred to by Dworkin above (p. 311). Smiley’s rejection of Dworkin’s version of paternalism points to one of the key insights of a relational understanding of autonomy: the necessity of distinguishing between different types of interventions in the lives of individuals (and the relationships on which such intervention hinges) rather than equating intervention with necessarily diminished autonomy.

A power differential between state and citizens or state agents and citizens in itself does not signal paternalism; it is specifically when such power is used in a coercive fashion that serves oppressive ends that paternalism can be seen. This is central to the distinction Smiley makes between paternalism and nonpaternalistic “protection.” Smiley (2004, 314) emphasizes how protective legislation can challenge systems of domination and inequality, whereas paternalistic legislation perpetuates them. She notes, “Protective legislation enables individuals to organize themselves collectively against powerful actors who, because of their institutional positions of strength, are able to lead other individuals to take serious physical risks.” Protective legislation gives marginalized and weak citizens the collective power that they may inherently lack given their societal positioning. In contrast, the conditionality that characterizes new paternalist services, which I turn to below, compounds domination by explicitly deeming service users incompetent and ignoring contextual details that contribute to their marginalization.

Like Smiley’s account of protective legislation, which stresses the importance of enabling collective action, Julie White’s (2000, 135-37) reconceptualization of paternalism highlights the importance of participatory politics in distinguishing it from other forms of intervention. White directs us to consider the nature of the process of intervention. As an alternative to paternalism, “a democratic politics of care” requires “a more participatory process of defining needs, where the discussion privileges the voice of those presently ‘in need’ in the course of defining ‘need’ and determining arrangements of resources to meet those needs” (p. 136). Thus, White too emphasizes the value of some forms of intervention in enabling social relations that foster collective action. But for White, it is equally important that this collective action be oriented toward making demands prefaced on needs that have themselves been collectively defined in a manner that foregrounds the often silenced voice of those in need. Because new paternalists assume that welfare recipients’ needs are a function of incompetence, their role in the politics of needs interpretation is discredited at the outset.

Smiley’s and White’s accounts of paternalism usefully clarify the distinction between different forms of intervention, some of which enable subjects to challenge oppressive social structures, while others compound them. I suggest that it is useful to supplement these accounts with the dual notion of autonomy as status and capacity cited above. When cast in this light, the differences between the conventional notion of paternalism and the revised accounts can provide greater explanatory power. As the quote above suggests, Dworkin claims (along with Mill) that paternalism limits autonomy as status insofar as autonomy is understood to be an individual’s status as “an independent entity.” Immediately, the conflation of independence and autonomy suggests a problem. Where paternalism is understood to be government interference in individuals’ lives that perpetuates relations of domination and inequality, the status is denied not because it signifies a lack of independence but because of the power relations that, in many cases, are tied to the assumption of incompatibility between autonomy and dependence. When the state imposes certain restrictive conditions on a woman on welfare, it acts paternalistically not simply because it fails to view her as independent (we know she is not, nor any of us) but because her dependence is stigmatized. It is stigmatized in this context in such a way that it is inconceivable that her status as welfare dependent (as well as her racialized and/or gendered identity, which may connote “dependence” within given contexts) could be consistent with a status as an autonomous agent, or one who deserves to be treated as such.

The distinction between the conventional and revised understandings of paternalism can also be seen in light of an understanding of autonomy as a capacity. Again here it is useful to turn to Dworkin’s potentially acceptable cases of paternalism. Dworkin (1971, 119) suggests that an acceptable form of paternalism might allow for the enhanced ability to carry out one’s his own “rational decisions.” Putting aside the fact that we might want a more expansive notion of autonomy than one that requires rationality as a primary marker, the difference between the revised understanding of paternalism and Dworkin’s is that the former situates the capacity within a given social context. Although some interventions by the state can be seen on Dworkin’s account as examples of acceptable paternalism—insofar as they limit the ability to act independently to further the aim of enhancing the capacity for autonomy overall—Smiley might suggest that such interventions resist rather than perpetuate relations of domination and equality, and therefore need not be thought of as paternalistic. On Julie Anne White’s (2000, 136) account, if the intervention opens up space for democratic participation rather than generating an instance of “speaking for others,” we may also reject the label of “paternalistic.” The latter approaches to state intervention might also draw our attention to cases in which the intervention is indeed paternalistic—not because it is an intervention into the lives of service users but because it exists in a context that perpetuates relations of domination and equality. Dworkin’s account might not be able to make such a distinction since it does not highlight the contextual details that contribute to power relations in such a situation.

2. New Paternalism

With this revised account of paternalism in mind, I now shift to the “new paternalism.” New paternalism is a philosophy of social service delivery, coined as such by Lawrence Mead, that seeks to further the trend of more paternalistic social welfare policies. It is, I argue, marked by the relations of domination and inequality that characterize our revised notion of paternalism. While new paternalists acknowledge the centrality of coercion to their mode of service delivery, they suggest that these programs ultimately foster autonomy, even if they temporarily restrict it to attain this end. I contest the theoretical bases of this claim here, and in doing so demonstrate this philosophy’s incompatibility with a theory of an autonomy-fostering state.

According to Mead (1997a, 2), the new paternalism involves “social policies aimed at the poor that attempt to reduce poverty and other social problems by directive and supervisory means.” Such means impose penalties or restrict benefits when recipients fail to conform to certain behavioral requirements—work, mandatory attendance at various programs, abstinence from drugs, and so on. According to Mead, “These measures assume that the people concerned need assistance but that they also need direction if they are to live constructively” (p. 2). The relations of power involved in new paternalist measures are configured, in part, by this claim regarding the need for direction: Mead suggests that the poor are specifically lacking in some capacities, including that of autonomy. This is because, on his account, it is misguided to assume (with regard to the welfare dependent) that “behavior is consistent with intention” (p. 5). He argues the poor do not have the capacity to live according to their life plans, even when they have generated these plans and express a desire to pursue them. New paternalist policy is therefore not coercive in the sense that it dictates what values individuals ought to have; rather, “the clients of paternalism commonly do accept the values being enforced. . . . However, they commonly fail to conform to these values in practice. Paternalism seeks to close that gap” (p. 5).

Mead offers an explanation regarding why his “new paternalism” is indeed “new.” New paternalism diverges from its forebears in several ways: first, new paternalism is not instantiated primarily by restriction of benefits through the narrowing of eligibility rules. Second, Mead (1997a, 7-8) notes, “today’s paternalism is mostly government led.” A third respect, which Mead mentions briefly, is that new paternalist programs are not custodial: “The current paternalism often involves supervision within society” (p. 9).

The first and third elements cited by Mead are true to some extent, but primarily at a rhetorical level that masks the practical effects of these policies. This is not to say that there is nothing new about new paternalism, however; it may be that this gap between the rhetoric and practice itself directs our attention to the more meaningful differences. Mead’s claims that new paternalism is not primarily about restriction nor generally custodial can be understood as reflecting a shift to the exercise of what Foucault refers to as “biopower.” Anna Marie Smith (2007, 38) describes biopower, quoting Foucault, as “largely concentrated in ‘positive’ functions; it mainly works to ‘incite, reinforce, control, monitor, optimize, and organize the forces under it.’ Instead of operating as an exclusionary and external force, it is ‘a power bent on generating forces, making them grow, and ordering them, rather than one dedicated to impeding them, making them submit or destroying them.’” Thus, insofar as new paternalists aim to “optimize” poor people’s behavior—aligning intention and action—such programs use paternalism not in a restrictive manner but in a manner that reforms “the self.”

As it suggests that poverty ought to be seen as rooted in individual pathology, new paternalism denies its targets’ status as agents capable of autonomy. The programs described by Mead and his cohorts in a collection of essays on new paternalism rely heavily on the claim that poverty can be only minimally explained by structural conditions—social, economic, or political. Rather, as Sanford Schram (1999, 671) points out in his review of the collection of essays, the emphasis laid in particular on mental health highlights the medicalization and infantilization of poverty and the poor, respectively. It may be the case that such pathologization is necessary to justify new paternalist policy (671). Mead (1997a, 26) acknowledges that the assumptions of paternalism can be seen as demeaning, specifically because of the lack of recognition of autonomy: “By assuming that recipients cannot be trusted to pursue their own interests, paternalism in effect treats adults like children.” Although he initially refers to paternalist policy as “postracial social policy,” he notes that the demeaning nature of such policy “is especially egregious in the case of black Americans,” ostensibly because of the historical injustices inflicted on them. Nonetheless, “[t]he assumptions of paternalism no doubt are demeaning, but the problems the poor have with working and other civilities are far more damaging to them” (Mead 1997a, 27).

As Schram notes, these claims to “reforming” the flawed individual are used rhetorically to justify new paternalist policy. However, at the level of practice, this justification may not reflect the actual aims of the new paternalist welfare programs (Schram, Fording, and Soss 2008). While a Foucauldian analysis of the rhetoric of new paternalism would suggest that programs exercise power through therapeutic dialogue (“confessional discourses”) or rehabilitative frameworks, in reality they often do not pursue such aims. A. M. Smith (2007, 55) explains, many post-PRWORA paternalist programs have “instead contributed to a regime that prioritizes blocking needy mothers from entering the program, imposing financial sanctions upon existing recipients, pushing already working poor mothers into humiliating workfare positions, and trimming the rolls through diversions and expulsions.” While the rhetoric of new paternalism paints a picture of a “less offensive” form of paternalism, in fact it masks what are often more direct forms of domination, not necessarily motivated by the paternalist aim of doing what is in the best interest of the subject.

A further example of the gap between new paternalist rhetoric and practice can be found in the second element of novelty emphasized by Mead: what he refers to as the government-led nature of new paternalism. Mead’s intention is to distinguish new paternalists from those who wish to “roll back” the state; appealing to both liberals and conservatives, Mead suggests that the former should approve of this “new” aspect of paternalism. Yet while Mead suggests that “old” paternalism (pre-1960) was initiated primarily by charities or religious groups, in the United States as early as 1935, Aid for Dependent Children programs offered federal assistance to single mothers. Such assistance was surely paternalistic insofar as it primarily sought to shield white widows from the need to work so that they could act as full-time caregivers to their children—a goal that reflects a claim to the nature of ideal motherhood. Moreover, this account fails to acknowledge the increasingly privatized nature of new paternalist programs, which often outsource elements of service delivery to nongovernmental organizations. Schram, Fording, and Soss (2008, 19) describe “the new poverty governance” as marked by a state that “has not only become more directive and supervisory in the sense described by Mead; it has also become more hollow.” Devolution to lower jurisdictions and private providers, they argue, has created a governance “that is, at once more muscular in its normative enforcement and diffuse and diverse in its organization” (p. 19).

What are we to make of these gaps between the rhetoric of new paternalism and its practice on the ground? It is perhaps these gaps that make new paternalism new, to some extent. New paternalists often claim that their preferred programs do important “work” on the self, reforming, reorienting, and remotivating the poor. They make this claim at the same time that they make assumptions about the nature of the self they seek to modify. Yet as my analysis below suggests, these programs often fail, primarily as a result of the faulty assumptions that fuel misguided programs. The relationship among the assumptions, intentions, and ensuing policies that new paternalists lay claim to may in fact reflect a level of disingenuousness. Since it is difficult to ascertain the “true” intentions of policy makers who espouse the type of rhetoric that Mead uses to describe the new paternalism, my analysis of new paternalist programs focuses on the rhetoric and the intentions it implies, trying to make sense of the ideological foundations of new paternalist policies and the justifications offered in support of these policies.

3. Conditionality

To respond to the supposed incompetence of the poor on which new paternalism is premised, many new paternalist programs employ a strategy of conditionality, making vital services conditional on conformity to behavioral requirements. The power relations that follow from the demeaning assumptions of new paternalism in combination with the use of this strategy of conditionality affect not only recognition of autonomy status but also the development and exercise of the capacity itself. While conditionality on its own does not necessarily imply a violation of autonomy—under the right circumstances it could be a sign of respect rooted in expectations that individuals can be responsible agents—because paternalist conditionality is so closely tied to a lack of respect, it becomes autonomy constraining.

Nevertheless, new paternalists suggest that although autonomy constraining in the immediate application, these policies ultimately enable the poor to become more autonomous. In the sense that parents may place limitations on their children’s autonomy to enable them to become autonomous in the future, new paternalists suggest that programs like workfare make the same sacrifices in the interests of long-term gain. Mead refers to the strategy of service delivery employed under a philosophy of “help and hassle” (Mead 2004, 158). Case managers under an entitlement system once acted “as advocates for the poor who helped them get all the benefits to which they were entitled” (p. 158). In contrast, paternalist case managers “do this but they are authority figures as well as helpmates” (p. 158). Such “authority,” in the context of welfare, however, creates relations of domination; authority in the context of parenting has fundamentally different implications and does not necessarily constitute domination.

To make more sense of this distinction and further consider the implications of the conception of autonomy underlying paternalists’ use of conditionality, we might say that it is plausible that behavioral expectations that follow from the authority expressed by these caseworkers are a marker of respect. That is, the existence of expectations indicates that the authority figure understands the service user to be an autonomous agent capable of conforming to these expectations; such respect, one could say, is thus autonomy fostering. Indeed, we might understand expressions of parental authority this way. However, in the context of new paternalism, the assumption is not that the service user is an agent but rather that she or he is incompetent. Rather than offering a type of enabling respect, the constraining social relations that emerge hinder the development of the capacity for autonomy. Moreover, conditionality isolates and stigmatizes service users by virtue of the regulations imposed on some, rather than all, citizens, therefore hindering recognition. While other citizens are subject to reciprocal obligations to access certain goods, unlike service users they have much greater latitude in negotiating and consenting to the terms of such agreements.

One example of a way such “authority” is expressed in the context of new paternalism highlights why such intervention is distinct from what I refer to as autonomy fostering. Mead (2004, 158) quotes a caseworker in a paternalist program, John Gardner:

I’ll do anything to help you [get] a job. But if you disappoint the employer—if you make me look bad—if you screw me over—you better watch out. I’m coming after you. I’m in your face. You’ll wish you’d never been born.

According to Mead, although to the sensibilities of middle-class people such a statement may seem harsh, on new paternalist accounts the poor are in fact enabled by such an approach (p. 158). On such an account, this could be seen as an example of an attempt to enable relational autonomy: the relationship between caseworker and service user enhances the user’s ability to pursue her or his own ends. But in fact, even in Mead’s assertion about the differences between communication styles of middle-class and poor people, the lack of respect that premises such an approach is evident. Though I believe caseworkers must play an important role in a genuinely autonomy-fostering state, here the required recognition and respect are absent and reinforced as such by the language of the caseworker. It is hard to see how the relations of domination evident in such an approach foster autonomy, even in the long run; instead, a culture of intimidation pervades—hardly conditions that enable an individual to ultimately define her or his own life plans.

1. Mistaken Assumptions I: Incompetence or Incoherence?

New paternalist workfare policy and pregnancy prevention policies are premised on two types of assumptions: first, assumptions about the competence of the service user and, second, assumptions about the value of paid employment and care work. Women who become pregnant at a young age or while living in poverty are seen to do so as a result of incompetence, not autonomous decision making. In a similar vein, the primary reason for nonwork is assumed to be incompetence—not lack of opportunity, not the market, and not discrimination. These claims follow in part from new paternalist assumptions about values. In keeping with widely held beliefs in many Western societies, “work” is valorized as a good in itself, one that is somehow constitutive of our identities as citizens and human beings. Moreover, the boundaries of what is considered work are relatively narrow and inflexible. Work refers to paid labor outside of the conventionally understood “private” sphere. In turn, since the value—at least in terms of meaning making or “fulfillment”—of work is primary, child rearing or other domestic care work is seen as both less desirable and a less legitimate life choice. Delaying childbirth and focusing on work as a source of dignity and full citizenship status are presented as an obviously superior life plan. Below, I look at the assumptions of incompetence framing pregnancy prevention and workfare policy. I return to the set of assumptions around values in the subsection following.

Teenage pregnancy, as well as multiple fertility in poor, unwed women more generally, has long been a focus for poverty researchers, who worry about not only the material implications of additional mouths to feed but also the effects on a variety of other outcomes typically measured in children. Moreover, they warn of the possibility of a generational “cycle” of unwed pregnancy and welfare dependence. New paternalist thinkers have turned their attention to these issues, citing teenage pregnancy in particular as a classic example of the claims made by paternalists regarding the rift between values and actions of poor people. Rebecca Maynard (1997, 91) claims, “As a group, those who unintentionally get pregnant and begin parenthood at a young age signal their inability to make decisions that are in their own best interests, the best interests of their children, and the best interests of society.”

The unemployed also signal their incompetence by failing to act in accordance with their own interests. According to new paternalists, nonwork is a problem not of lack of opportunity or structural constraints but rather of competence on the part of the job seeker. Deterred by previous experiences of failure, preoccupied with other concerns, or simply lacking in motivation, the poor fail to find work because, without the coercive force of the state, they cannot organize themselves to do so. Mead (1997b, 64) presents a psychosocial explanation for this failure. The poor share with the rest of society the value placed on work; indeed “[n]ot working . . . causes shame and discouragement, since they are not living by their own values.” However, the “gap between intention and behavior makes work enforcement necessary” (p. 64). This enforcement is facilitated by the fragmented yet existing “work ethic” that already exists among the poor; “[m]andatory work programs do not ask people to do something alien to them. . . . They now have to do what they always wanted to do” (p. 64). Despite these shared values, Mead argues, the poor are “different” psychologically from more successful members of society. It is in the realm of intentionality where this difference is seen:

Better-off people generally behave according to their own intentions. If they do not do something, it is because they do not want to. They will resist anyone telling them to do otherwise. Middle-class analysts too readily assume poor people are equally consistent. (Mead 1997b, 64)

In our blurry middle-class analysis, we mistakenly believe that those who do not work choose not to, according to Mead.

Mead emphasizes an individualized explanation for nonwork. Though some explanations of joblessness refer to what William Julius Wilson (1996, 39) calls “spatial mismatch”—“a growing mismatch between the suburban location of employment and minorities’ residence in the inner city”—Mead rejects such structural explanations of poverty. Although he acknowledges that the mismatch between skills and jobs that resulted from deindustrialization in major cities in the United States, Mead dismisses this as relatively insignificant. A study conducted in Chicago by Wilson, he notes, “found . . . that low-skilled immigrants worked at high levels in the same ghetto areas where poor blacks and Puerto Ricans worked at low levels” (Mead 1997b, 49). Therefore, if some groups are able to find work while others are not, the problem is with the intention–behavior relationship and not the system.

However, other research suggests that Mead’s analysis is overly simplistic and fails to take into account complexities that deeply affect employment possibilities. One important factor is the structure of social relationships that shape the ways in which individuals both view and participate in the job search process. Some research suggests that as a result of chronic conditions of racism and poverty, black job seekers may adopt an approach to job search that is particularly individualistic, failing to seek out the support (and when sought, to successfully receive this support) necessary to secure employment (S. Smith 2007). ⁸ Sandra Smith (2007, 22) describes a phenomenon of “defensive individualism.” Defensive individualists do not reach out to the community for fear of failing to live up to the expectations of those around them; they justify this behavior in individualistic terms. Smith explains, “Within the context of poverty, friends, relatives, acquaintances, and institutions in their social milieu blamed the black poor and jobless for their persistent joblessness, deploying discourses of joblessness that privileged individuals’ moral shortcomings and stressed personal responsibility and self-sufficiency as a panacea” (p. 22). Defensive individualism is a reaction to the experience of being viewed in this manner. The individualizing message of new paternalism exacerbates this phenomenon of defensive individualism; the discourses that Smith refers to may stem in part from the internalization of the messages espoused by privileged actors who deploy such strategies as workfare to respond to supposed incompetence. This argument calls into question Mead’s claims about the availability of jobs based on the success of some groups rather than others in finding employment.

While S. Smith’s (2007, 30) arguments seem to counter some earlier studies of poor black communities, wherein the importance of connectedness through kinship relations in particular is stressed, she notes that the individualistic reaction to joblessness does not necessarily prevail in all arenas, particularly in the “private sphere.” Given the individualism—motivated by distrust—that may characterize some arenas of poor black women’s lives, it might make sense to suggest that seeking relational support in other arenas, that is, through childbearing, follows logically. This points to the flaws in logic underpinning pregnancy prevention programs, too.

Pregnancy prevention policy puts in place “directive” programs to respond to the supposed failure of young poor women (largely of color) to implement family planning strategies that are in their interest. But as with the case of work above, ethnographic data demonstrate a much more complex picture of the intentionality of poor young women who become pregnant than the one presented by Maynard. Pregnancy in poor women and teenagers may have much greater personal and symbolic implications than the explanations provided by Maynard (that of irresponsibility and incompetence) suggest; these implications are directly tied to conditions of injustice and oppression that many of these women face. Even Maynard’s own ethnographic data seem to suggest that the assumption of incompetence is misleadingly straightforward. Maynard quotes teenage mothers who signal their ambiguous intentionality. Says one mother, “‘I didn’t plan it, and then again, I kind of knew that it was going to happen because I wasn’t really taking the pills like I was supposed to. I couldn’t remember every day to take the pill. And, I still don’t’” (Maynard 1997, 91). Although the pregnancy was not planned per se, it is also unclear that the interviewee specifically believed that she should avoid pregnancy yet failed to act in accordance with this view, nor that her actions entirely oppose her intentions.

Going beyond this ambiguous intentionality, other research more explicitly sheds light on alternative explanations of teenage pregnancy. In a study of motherhood and marriage among low-income women, Kathryn Edin and Maria Kefalas (2005) highlight the central role a child can play in a context where relational support is limited and the ability to define oneself is constrained by material and relational conditions. Like Smith, they point to the “relational poverty” that emerges from “the social isolation that is the common experience of those who live in poverty [which] is heightened for adolescents, whose relationships with parents are strained by the developmental need to forge an independent identity” (p. 34). ⁹ These limitations of relational support “can create a compelling desire to give and receive love” (p. 34). The question of intentionality in teen pregnancy, then, is in part colored by the very real, and arguably logical, reasons why these women may be motivated to become pregnant, or at least to not actively prevent pregnancy. “[P]regnancy offers the promise of relational intimacy at a time few other emotional resources are available,” the study’s respondents suggest (p. 24).

In this light, pregnancy and child rearing can be seen as signaling the competence of the young women who make such choices to further their interests. This is not to say that it is ideal to live in a society where one’s primary (if not only) chance for relational support is to be found in care work, with few other options available. Yet given the circumstances that exist on the ground, it is hardly incompetence that leads to such a conclusion. Maynard’s assumption that the young women she studies could not have “competently” chosen to be become pregnant parallels the narrow nature of the interpretation of work offered by advocates of workfare: when they claim that nonwork reflects a disjuncture between a value system that favors work and actions that help to access work, they refer to work only as defined by paid employment. As Carole Pateman (2005) and other feminists point out, this overlooks the unremunerated caretaking labor that many women are engaged in and may rely on for important sources of relational support. ¹⁰

Women engaged in such care work are counted as “able but unwilling to work” and therefore as undeserving. Their domestic work, the narrow definition of work ultimately implies, does not “earn” them any benefits. Mead argues that one of the demands of citizenship is the civility of “work.” However, Pateman (2005, 36) points out that such a view ignores another important role of social reproduction, which refers not only to motherhood but also to “the maintenance and future of the public or common weal and the care of citizens.” If work is a condition of citizenship because it is something “owed” by one citizen to another, social reproduction, too, ought to be included in this notion of obligation. Although welfare reform rhetoric links dependency to a “failure to perform a duty owed to fellow citizens,” it does so only by ignoring the other duties citizens perform outside of paid employment (Pateman, 42).

Mead rejects feminists’ claims that care work ought to be acknowledged as falling under the umbrella of status-granting “work.” Rather, he suggests that “civic labor” must fulfill certain conditions: it must be accountable to society and must be done “well enough to serve a public interest” (Mead 2005b, 188). With regard to the first condition, Mead suggests that the “self-chosen” character of bearing children renders caregiving an inappropriate means of fulfilling the conditions of accountable work; paid employment in public or private sectors is assigned by others and is therefore legitimate. To claim support for raising one’s child is equivalent to making a claim to societal support for “writing this essay” (p. 188). Caring for someone else’s child or a sick relative, since these tasks are externally assigned and/or unchosen, may be seen as legitimate (p. 188). Yet it is ironic that here Mead attributes agency to poor women who have children and make welfare claims. As I have suggested, the very premise of new paternalism is the incompetence of the welfare recipient, her or his inability to link intention and action. It is convenient, then, that when evaluated in terms of what counts as “work,” childbearing becomes intentional and agentic. This contradiction aside, even if poor women intend to bear children, it is unclear why “self-chosen” tasks rather than assigned tasks ought to count as less legitimate means to meeting the requirements of civic labor. In fact, social, political, and economic privilege grant increased levels of occupational autonomy, and such autonomy is supported by not only private means but also public funds (e.g., education funding, tax benefits, etc.).

Though Mead points to studies regarding the poor outcomes of children whose mothers are on welfare, the disadvantages that he points to emerge within the context of a system that fails to provide the material and social supports that may be necessary to enable poor women to best care for their children. Under different circumstances, these women may indeed be better able to “serve” societal interests. If poor women were indeed such incapable caregivers, the suggestion that caring for other people’s children would be a more legitimate form of civic labor is perplexing. Moreover, the lack of checks in place on mothers’ performance is hardly unique to this form of work, as opposed to the forms of work Mead claims meet the requirements of civic labor. Furthermore, wealthy women who rely, by virtue of privilege not desert, on private sources of support such as spouses or extended family to care for their children are not seen as failing to contribute to society. As I discuss in the following section, this class-based stratified valuation of women’s caregiving is evident not only with regard to civic contribution but also on a more individualized level, with regard to the expected meaning women can derive from this activity rather than paid labor market participation. The devaluation of poor women’s ability to mother is evident in extant policies in some states that actively encourage welfare recipients to give their children up for adoption (A. M. Smith 2001, 177-84).

Ultimately, Mead (2005b, 189) claims, “what counts as civic labor is finally a political question.” That is, what matters is that American society does not recognize caregiving as an activity comparable to work. This claim, however, fails to account for the fact that not only does politics influence policies; the reverse is also true. Policies that reject caregiving as a form of civic labor, discourage poor women from having children, and encourage poor women to give up their children surely have an effect on the politics of civic labor.

Not only do Mead’s and other new paternalists’ accounts overlook the societal value of care work, they also overlook the personal fulfillment derived from care work—a point that complicates the picture painted by paternalists of paid employment as the sole route to dignity and fulfillment, as described below.

2. Mistaken Assumptions II: Whose Values?

The second set of assumptions underlying new paternalist policy hinges on the notion that work, as opposed to nonwork or care work, is a primary source of value in the vast majority of citizens’ lives and is therefore a conduit to fulfillment and meaning making in our lives. New paternalists claim that, for most able-bodied individuals, work is a marker of citizenship and a gateway to all the rights that citizenship bestows on us. ¹¹ Paralleling this claim is the one highlighted in Maynard’s work: young unwed mothers, she argues, appear to share in the mainstream value system that suggests waiting until one is older to bear children. (This claim, of course, serves to justify the claims regarding incompetence discussed above.) Given the value assigned to paid work, it makes sense to seek out work prior to bearing children. Maynard’s claims about pregnancy represent a larger ideology found in dominant discourse and new paternalist thought, which presumes an individualistic view of the ideal self. This is accompanied by an overarching devaluation of relationality and interdependence that is paradoxically accompanied by an unrealistic view of the ideal, altruistic mother. As the discussion above regarding the potential for relational support via childbearing indicates, this view obviates the importance of a range of other values associated with childbearing in young, poor women. ¹²

Edin and Kefalas’s (2005, 182) work underlines the ways in which not only masculinist conceptions of the self but also many appropriations of feminism are uncomfortable with the positioning of childbearing as a primary source of meaning making. They note, “The idea of a woman viewing her offspring as a resource violates powerful social norms about how a mother should behave. Altruism, not need, ought to govern her relationship to her children.” Though feminists have long endeavored to destabilize it, the ideal of the mother as self-sacrificing and entirely other-regarding remains a strong one in our society. Yet despite the feminist critique of this self-sacrificing ideal of motherhood, a typical feminist alternative conception of motherhood still does not fit comfortably with the empirical evidence Edin and Kefalas offer. Even if we reject the notion of motherhood as necessarily only altruistic, the idea of women deriving meaning primarily from their role as child bearers might be seen as constraining from the vantage point of white, middle-class feminism. Are these women simply falling victim to normative conceptions of the sexual division of labor or idealized notions of “maternal yearning”?

Consider Aliya and Pamela, two of the women Edin and Kefalas interviewed. Aliya says, “‘Some people may say it was for the wrong reasons, but it was like too much around me going on. . . . I guess that was my way out of all these situations. [But] I wanted a child because it was mine. It was [for] love’” (Edin and Kefalas 2005, 34). Pamela, in turn, contends, “‘I just knew, growing up, “Oh, you’re gonna have your kids . . . the kids are gonna love you. They’re yours”’” (34). Although Aliya and Pamela express different sentiments here, it is noteworthy that for both women motherhood is not only or even mostly about the needs of the child but also about the fundamental needs of the mother for love, for affirmation, and for support. Moreover, as Edin and Kefalas note, “the stronger preference for children among the poor can be seen in the propensity of the women we interviewed to put children, rather than marriage, education, or career, at the center of their meaning-making activity” (p. 206).

I interpret the comments of Edin and Kefalas’s interviewees in two ways. First, where some women may fall prey to the dominant ideal of white, middle-class motherhood, poor women who see their children as providing them with much-needed self-affirmation subvert the dominant paradigm, and within this resistance we find a kernel of autonomy. Second, viewing childbearing as the primary source of women’s meaning making seems antiquated, a result of the oppressive socialization attributed, for example, to the “self-abnegating mother.” ¹³ The women’s movement has struggled to open up a far greater range of opportunities that can contribute to a meaningful life; these women, on this interpretation, are limited in their autonomy insofar as they fail to access or take advantage of such broader opportunities. However, in reality, their assessment of the opportunities available to them is quite accurate, complicating the oppressive socialization explanation.

On the first account, consideration of the contextual variables at play suggests that situating childbearing as an avenue to the types of self-affirmation and support described by the interviewees can be seen as resistant, and thus, perhaps, as autonomous. Kaplan’s theory of the poverty of relationships helps to explain why viewing childbearing as a relational resource makes sense and may be considered to be a marker of autonomous agency. The poor, black, young women she interviews “describe being disconnected from primary family relations, abandoned by their schools and by the men in their lives, and isolated from relations with other teenagers at the time of adolescence, when it is most important that they experience positive relationships” (Kaplan 1997, 11). Motherhood is a strategy used to cope with the conditions under which these young women are operating (p. 181). This does not mean that childbearing at a young age is an ideal autonomous choice under the given conditions. In fact, that it is not ideal is exactly the point; to question paternalist pregnancy prevention strategies is not to endorse teenage pregnancy or (for the most part) multiple pregnancies of poor women. It suggests instead something about what type of interventions are necessary, shifting the terms of the “diagnosis” offered by the paternalists and in turn the “cure” (to use the paternalist language of pathology).

On the second interpretation, the structural conditions that affect these women suggest instead that poor women are not necessarily succumbing to sexist norms or values that limit women’s opportunities. Rather, they make a fairly accurate assessment of the limitations that exist on their potential resources for fulfillment. As Edin and Kefalas (2005, 205) write, while middle-class women face significant opportunity costs when they have children at an earlier age, the same cannot be said for poor women. ¹⁴ Rather, “[d]isadvantaged girls who bear children have about the same long-term earnings trajectories as similarly disadvantaged youth who wait until their mid or late twenties to have a child.” Other outcomes are similarly minimally affected by early child rearing. Situating childbearing as a primary source of meaning not only disrupts the assumptions made by new paternalists about the centrality of work as a meaning making device but also draws our attention to the lack of opportunities that exist for poor women (p. 206).

3. Conditionality in New Paternalism: A Failure of Reciprocity

If child rearing is a greater—and more legitimate—source of meaning making in citizens’ lives than new paternalists make it out to be, the value of the sort of work associated with workfare is in fact a much less significant source of meaning making than this theory suggests. I do not wish to dismiss the possibility that work is indeed an important source of meaning making in people’s lives and, in the context of our social and economic systems, a primary means of gaining recognition. Nevertheless, most workfare jobs, as well as those jobs that welfare recipients who leave the rolls take, are low-wage, low-flexibility jobs that tend to be associated with low levels of personal fulfillment. This is not to say that these jobs are necessarily burdensome and never enabling but rather that conditions are hardly ideal. ¹⁵

Advocates of workfare lay claim to the technique of conditionality as an important key to enhancing recognition of service users as citizens or as potentially autonomous agents. Charles Beem and Mead (2005) claim that moving welfare recipients into the workplace renders their dependency acceptable and allows them to integrate into broader society. Even though income may not increase significantly for these recipients, “gains to equal citizenship, however, were significant” (Mead 2005b, 177). Though they are dependent on the state, like social security recipients, ¹⁶ welfare recipients who work are accepted because “[i]n citizenship terms, fulfilling the demand to function is far more important than minimizing one’s demands on the society” (p. 177). Work serves to “rebuild ties between the poor and the rest of society” (p. 178).

Mead (2005b) suggests that the public’s acceptance of the radical changes made under PRWORA is evidence that average citizens now regard welfare recipients engaged in “work activities” as full citizens. However, if participation in the labor market actually garners recognition that was once denied to dependents in our society, the conditional nature of benefits that is the hallmark of workfare programs undermines this recognition by flagging service users as lacking in the qualities dominant culture ascribes to citizens. While I do not think conditionality is necessarily unjust, in the context of our current social and political conditions it fails to foster autonomy, which I regard as one element of a just welfare policy. In this context, as Desmond King (2005, 74) charges, conditionality in the form of “workfare might well produce a deepening sense of alienation and exclusion among those it claims to help.” Mead either neglects or minimizes the deleterious effects of the assumptions that underlie the policy and the stigma that follows from the singling out of certain individuals to be the targets of coercive policy. King argues, “Participants in workfare are treated differently, and not in a positive sense, from participants in other state-administered benefit programs” (p. 74).

In fact, post–welfare reform research indicates that Mead’s predictions regarding eroding stigma on welfare recipients as a result of workfare have not panned out. Soss and Schram (2007, 111) evaluate the policy feedback mechanisms at work in the case of the 1996 welfare reforms, considering the claim that “new policies create a new politics.” ¹⁷ Do work requirements minimize public opposition to welfare assistance and in turn, as Mead argues, shift politics to the left and reduce racism (cited in Soss and Schram 2007, 112-13)? According to Soss and Schram, though the public views work requirements and time limits in a positive light, these changes “did not generate more positive images of poor people, welfare recipients, or welfare itself” (p. 120). Supporting Soss and Schram’s findings, Dyck and Hussey (2008) have shown that following welfare reform Americans continue to associate welfare with negative perceptions of African Americans’ work ethic.

These findings may suggest that the stigma associated with being compelled to work under workfare inevitably shapes the perception citizens have of workfare participants. Contrary to the claim that work will enhance citizens’ views of welfare claimants, King (2005, 74) argues, “we may instead view them as incompetent, hence lacking the qualities need for membership in the polity as equal citizens.” Mapped onto the racialized nature of welfare stigma, workfare only further differentiates recipients from their fellow citizens. Indeed, the risk of viewing workfare participants as incompetent is high given that incompetence is the very foundation of paternalist policy. Consider James Q. Wilson’s (1997, 341) claims about paternalism. He suggests that government should extend paternalism to people “who have by their behavior indicated that they do not display the minimal level of self-control expected of decent citizens.” Those lacking self-control include “the homeless, criminals, drug addicts, deadbeat dads, unmarried teenage mothers, and single mothers claiming welfare-benefits” (p. 340). Clearly, there is a difference between viewing an individual as reciprocating for the benefits she or he receives and viewing her or him as duty bound to obey because she or he is personally flawed. Moreover, the flaws that are attributed to welfare recipients, as Wilson’s argument suggests, are specifically those that cast the recipients as individuals incapable of autonomy; therefore, whatever recognition labor market participation may have garnered is ultimately obscured by the underlying reasons for the coercive conditions under which they work.

Nevertheless, as Stuart White (2000) notes, the argument for reciprocity in the realm of welfare benefits is not easily dismissed. White argues that conditionality in social welfare benefits is not necessarily unjust; under some conditions the imposition of conditional welfare benefits may be acceptable and indeed necessary to uphold our egalitarian institutions. That is, “free-riding, or accepting benefits without social contribution, generates a clear risk that the egalitarian institutions in question will provoke feelings of alienation and resentment and so undercut the very spirit of solidarity on which they depend” (p. 515). The obligation to reciprocity, he writes, can be enforceable only under certain required background distribution conditions; “to assert otherwise is to assert that significantly disadvantaged individuals in a highly inegalitarian society may have an enforceable moral obligation to co-operate in their own exploitation” (p. 515). Instead, he suggests four intuitive conditions that must be in place for fair reciprocity to be enforceable: guarantee of a decent share of the social product for those meeting minimum participation standards, decent opportunities for productive participation, equitable treatment of different forms of participation, and universal enforcement of the minimum standard of participation (p. 515-16). Indeed, such requirements make conditionality appear more palatable when compared to paternalist arguments hinging on incompetence claims. However, in practice they are not likely to be met in our current social and political context, where both stigma and material circumstances make the fulfillment of such requirements appear lofty.

When examined within the actual structural context in which it is enacted, the conditionality applied in the realm of pregnancy prevention suggests a deep misunderstanding of the problems advocates of this strategy claim to address. One paternalistic approach to pregnancy prevention makes benefits conditional on teenage parents (under eighteen) residing with their own parents. If they do not do so, they will receive less income support from the state. Another program described in Maynard’s piece offers young women “clear” moral messages discouraging further pregnancy; again, benefits are conditional on participation. The perversity of such programs is vast. Given the explanations proffered by the women Edin and Kefalas and Kaplan have interviewed, we see that it is often the case that parents fail to provide the support girls and young women need. This is partly the motivation for bearing children. Here we can see that a failure to grapple with the intentions of pregnant teenagers leads to a faulty policy prescription. If teenage pregnancy were truly the result of incompetence alone, then perhaps parental supervision would help to rein in such behavior. Yet since it seems this is not the case, rather than reinforcing the conditions that contribute to a lack of relational, an autonomy-fostering state would seek to provide options for alternative sources of meaning making and relational connections. This requirement and the program of moral indoctrination referenced above individualize the behavior of teenage parents, failing to recognize the ways in which teen pregnancy can be understood as a response to institutional oppression, wherein childbearing appears as a reasonable and viable strategy (Kaplan 1997, 184).

The strategies of conditionality employed in both the pregnancy prevention and workfare instantiations of paternalist policy point us back to the distinction made in early sections between paternalist and interventionist policy. Paternalism is best understood as referring to forms of intervention that serve to perpetuate oppressive social relations by way of their coercive and stigmatizing tactics. As Kaplan (1997, 190) notes, the ways in which a lack of relational support contributes to teen pregnancy suggest that service delivery personnel ought to be “retrained to see themselves as supporters, to be empathetic, to offer real job training, and to seek economic and emotional support for the entire family unit.” How this is enacted depends on context and further research, but what is important here is that such intervention need not be paternalistic; rather, it can be autonomy fostering when the faulty exercise of unequal power relations entailed by new paternalist policy is curbed.

The value of such a revisioning of social service delivery extends, too, to our democratic commitments. As Julie White’s account of paternalism highlights, nonpaternalist interventions must be premised in some way on the service user’s involvement in the definition of her or his needs and the determination of how these needs must be met. Social service delivery that is founded on such a participatory model, rather than marked by “speaking for others in the process of defining their needs,” moves us toward affording service users equal citizenship rights (White, p. 136). As Mead suggests, the question of what constitutes civic labor is a political one; an autonomy-fostering model of service delivery would, unlike new paternalist models, respect the perspectives of service users and afford them opportunities to participate in the definition of, among other things, needs and social requirements. Because paternalist policy intervenes in marginalized citizens’ lives on the basis of a form of misrecognition—the misattribution of incompetence—it ultimately restricts the possibilities for autonomous agency and undermines democratic participation.