Introduction

The idea that scientific knowledge holds the key to solving social problems has long been an article of faith in American liberalism. Nowhere is this more apparent than when it comes to solving the "poverty problem." For well over a century, liberal social investigators have scrutinized poor people in the hopes of creating a knowledge base for informed social action. Their studies have generated massive amounts of data and a widening array of research techniques, from the community-based social surveys of the Progressive Era, to the ethnographic neighborhood studies conducted by Chicago-school social scientists in the 1920s, to the technically sophisticated econometric analysis that forms the basis of the poverty research industry today. Although its origins can be traced to what historian Daniel Rodgers calls the transatlantic "borrowings" of late nineteenth- and early twentieth-century progressives, contemporary poverty research is very much an American invention, with a degree of specialization and an institutional apparatus that is unmatched in other parts of the world.¹ And yet, poverty remains a fact of life for millions in the world's most prosperous economy, stubbornly resistant to all that social scientists have learned about its "causes, consequences, and cures."²

Frustrated by what they routinely refer to as the "paradox" of "poverty amidst plenty," liberal social scientists often charge that politics and ideology are to blame. We know what to do about poverty, they believe, but ideologically motivated policy makers from both sidesofthe aisle lack the political will to do the right, scientifically informed thing. A powerful expression of such frustration came in response to the "end of welfare as we know it" in 1996, when three highly respected Department of Health and Human Services Department officials resigned in protest over President Clinton's decision to sign the harsh, Republican-sponsored Personal Responsibility and Work Opportunity Reconciliation Act—now widely referred to as welfare repeal. "The passage of this new law tells us what we already knew," wrote HHS Assistant Secretary Peter Edelman in explaining his actions. "[P]oliticians make decisions that are not based on research and experience." Welfare reform was a triumph of politics and ideology over knowledge, that is, and a defeat for the policy analysts who had mustered an enormous amount of scientific data showing that the bill would send millions more children into poverty—very much in the hope of preventing politicians from doing the wrong thing.³

Accurate though it may be in its characterization of recent welfare reform, this explanation for what happened in 1996 has one overriding problem: it fails to acknowledge the role that scientific poverty expertise played in bringing welfare as we knew it to an end. Following a well-established pattern in post-Great Society policy analysis, the Clinton administration's poverty experts had already embraced and defined the parameters of a sweeping welfare reform featuring proposals that promised to change the behavior of poor people while paying little more than rhetorical attention to the problems of low-wage work, rising income inequality, or structural economic change, and none at all to the steadily mounting political disenfranchisement of the postindustrial working class. Approaching the poverty problem within the narrow conceptual frame of individual failings rather than structural inequality, of cultural and skill "deficits" rather than the unequal distribution of power and wealth, the social scientific architects of President Clinton's original, comparatively less punitive welfare reform proposal made "dependency" their principal target and then stood by helpless as congressional conservatives took their logic to its radical extreme. Their helplessness in the matter was not just a matter of "bad" politics laying "good" scientific knowledge to waste. It was also a failure of the knowledge itself.

Taken on its own, the recent "end of welfare" offers evidence for one of the central arguments of this book: that building an antipoverty agenda will require a basic change in the way we as a society think collectively about "the poverty problem," a change that begins with a redirection in contemporary social scientific poverty knowledge. Here I am referring to the body of knowledge that, very much as a legacy of Lyndon Johnson's War on Poverty, has attained a kind of quasi-official status in defining "the poverty problem" and assessing how social programs affect the poor. Besides being social scientific, this knowledge is based principally on quantitative, national-level data. It is produced by a network of public agencies, think tanks, university-based and privately operated research institutes that traffics in the shared language and recognized methods of applied economics and policy analysis. Although liberal in origins, poverty knowledge rests on an ethos of political and ideological neutrality that has sustained it through a period of vast political change. Very much for this reason, it can also be distinguished by what it is not: contemporary poverty knowledge does not define itself as an inquiry into the political economy and culture of late twentieth-century capitalism; it is knowledge about the characteristics and behavior and, especially in recent years, about the welfare status of the poor. Nor does it much countenance knowledge honed in direct action or everyday experience, whether generated from activism, program implementation, or, especially, from living poor in the United States. Historically devalued as "impressionistic," "feminized," or "ideological," this kind of knowledge simply does not translate into the measurable variables that are the common currency of "objective," "scientific," and hence authoritative poverty research.

Certainly I am not the first to make the argument that poverty knowledge, as currently constituted, needs to change. On occasion such an argument has been sounded by recognized poverty experts, exasperated, for example, by how their colleagues have allowed the political obsession with welfare dependency to overshadow the problems of wage decline, labor market failure, and rising inequality that continually get shunted off to the side in the poverty/ welfare debate.⁴ More often, though, the argument for change finds expression in the not-always-articulated frustration of people on the periphery of the poverty research industry—the program administrators, advocates, legislators, community activists, or, as in my own case, the foundation program officers—who since the 1980s have grown increasingly dissatisfied with the narrow, individualized focus of poverty research, who feel cut off from its technical language and decontextualized, rational choice models of human behavior, and who rankle at its refusal to acknowledge the value judgments underlying measures of welfare "dependency" that have come to play such a prominent role in recent policy. To be sure, thanks to poverty knowledge we now have a more accurate statistical portrait of who suffers from substandard incomes, housing, nutrition, and medical care—a far more diversified and shifting population than lingering stereotypes of the "other America" would allow. So, too, has poverty knowledge provided an indispensable picture of actual program spending and benefit levels that contradicts popular notions of welfare mothers living off the fat of the state. Poverty experts have also amassed convincing evidence about the links between poverty and macroeconomic performance, and about the extraordinary effectiveness of Social Security in reducing poverty among the elderly. And yet, however impressive its data or sophisticated its models, poverty knowledge has proved unable to provide an analysis or, equally important, a convincing narrative to counter the powerful, albeit simplistic story of welfare state failure and moral decline—a narrative that, with the help of well-organized conservative analysts, has come to inform policy discourse to a degree hardly imaginable twenty years ago.

I got my first introduction to poverty knowledge as a new assistant program officer at the Ford Foundation in the mid-1980s, when the liberal research establishment was still reeling from the impact of Charles Murray's just-released missive, Losing Ground. In that book, Murray used data and techniques earlier honed in predominantly liberal think tanks to argue that the liberal welfare state was to blame for a whole host of social problems, including poverty, family breakup, and crime.⁵ From an empirical standpoint, Murray's argument proved easy to demolish, and a number of poverty experts rose convincingly to the task. But their careful empirical analyses were no match at all for Losing Ground as an ideological manifesto: couched, as they were, in the language and conventions of ideologically neutral objectivity, these critiques alone were inadequate as a response to Murray's attack on both the value premises and the performance record of the welfare state. Nor were poverty experts organized to counter the network of explicitly ideological conservative and libertarian think tanks that had managed, through their own organizing and publicity, to gain control of the terms of the poverty debate. Along with many others at the time, then, I welcomed what has since become a perennial conversation about how liberal and progressive philanthropy can use knowledge more effectively to shape rather than react to public debate. At the same time, I was struck by what is still a pervasive assumption in the network of research institutes that make up the core of the poverty research industry: that knowledge, in order to meet the standards of empirical testing and rigorous scientific scrutiny, must—indeed that it ever is or can be—apolitical if not entirely value and ideology-free.

In my job as assistant director of the Ford Foundation's Project on Social Welfare and the American Future, I was responsible for managing a portfolio of research grants that purposely ranged across the broad spectrum of social welfare policies, but that inevitably concentrated on the hotly contested issues of poverty, welfare, and what was becoming widely known as the "underclass." This proved my first introduction to the enormous influence of foundations and government agencies in setting social scientific research agendas, through control not just over what and who gets funding, but also over what, at any given time, constitutes policy expertise. To the extent that poverty research agendas were driven by "the science" (a standard foundation question: "what do we know and what do we need to know?"), it was always with an eye to making social science more "policy-relevant"—a virtual guarantee, during an era of rising deficits and antiliberal, antigovernment backlash, that poverty research would confine itself to an ever-shrinking realm of political possibility. The parameters of research were similarly narrow, as captured in what at the time was repeatedly characterized as the central fault line in the social scientific debate: whether poverty was "structural," and hence "caused" by an absence of human capital, or "cultural," as measured through various indicators of bad behavior, including whether dependency and single parenthood were somehow passed along as intergenerational character traits. In neither case was poverty defined as anything other than an individual condition, nor was it seen as susceptible to any other than individual-level reform. Most striking to me, though, was how rarely anyone acknowledged that this constricted, strangely either-or debate was not at all new, and not one that had ever been definitively settled through recourse to empirical data and social scientific models alone. Here again I agreed with the still-current assessment that poverty knowledge needed to be more interdisciplinary, qualitative as well as quantitative, and much broader in scope—and that it could use a good deal more of what we on the Social Welfare Project took to calling "blue sky" thinking in analyzing the possibilities for reform.

A few years later, as a staff associate assigned to the Social Science Research Council's Program for Research on the Urban Underclass, I had an opportunity to work more directly with social scientists to attempt such a broad, interdisciplinary approach, for the purposes of understanding at least one dimension of contemporary poverty—the dramatic economic decline of racially segregated neighborhoods in the nation's postindustrial urban core. That project, which I joined in the early 1990s, used the work of sociologist William Julius Wilson as the starting point for what was to be a more structural as well as interdisciplinary understanding of the roots of ghetto poverty, one that, in the eyes of its sponsor, the Rockefeller Foundation, could inform the design of community-based programs as well as national policy debates. These aims, as I soon discovered, were more easily stated than achieved. On the one hand were the methodological, conceptual, even linguistic barriers between disciplines, all exaggerated by our effort to broaden the conversation to disciplines that had been sidelined within established poverty research networks in recent years. On the other there were the divisions separating "academic" from "applied" policy research, and social scientists from neighborhood residents and practitioners—divisions rooted as much in professional cultures as in conflicting ideas about what constituted "usable" knowledge for purposes of policy and program. Especially telling, though, was that the efforts to "operationalize" and test the underclass concept continued to rest far more heavily on indicators of individual and group behavior than on comparable measures of structural economic and/or institutional decline in urban neighborhoods—reinforcing the notion that some form of behavioral "pathology" was what caused and sustained the underclass. When the SSRC Underclass Program was ending, in late 1993 and early 1994, it had just barely begun to broach the institutionalized barriers to collaboration and to address the limitations of conventional measures for documenting structural and community-level change. By then, too, poverty researchers had started to pay more attention to the growth in inequality and the decline of wages as defining, structurally rooted conditions of late twentieth-century poverty. Still, social science was a long way from realizing a genuinely different kind of poverty knowledge, one that revolved more around the problems of political economy than the behavioral problems of the poor.

Coming, as it did, from a planning group led by prominent poverty experts, the Clinton administration's 1994 proposal to "end welfare as we know it" seemed more a step backwards than a reflection of the powerful evidence emerging from recent research—particularly in the administration's willingness to make dependency the issue without adequately addressing the more pressing issues of declining wages and available work. Protest though they might once conservative Republicans took over, it was difficult to deny that welfare reform drew its logic from a so-called "consensus" on dependency that the administration's poverty experts had helped to construct—or that welfare, especially in recent memory, was simply not an issue that would be decided on the basis of high-minded, nonideological debate. And yet, the end of welfare has decidedly not spurred efforts to rethink the premises, the organization, or the overwhelmingly individualized focus of poverty research. If anything, it has been the occasion for growth and expansion in the existing research industry—in response to the well-warranted concern for keeping track of what actually happens to people under the new rules.

For me, then, the role of liberal social science in ending welfare confirmed the need to reexamine, and ultimately to reconstruct, the foundations of contemporary poverty knowledge. But while this view is informed by my experience as a funder and a kind of participant observer, it has been more deeply informed by historical research. Through historical analysis I have come to appreciate why poverty knowledge is so loaded with meaning: why "knowing" poverty generates such controversy and so much attention; why what is recognized as expertise can be so consequential—though rarely in ways the experts anticipate—for the poor; and why as a body of knowledge that has been historically constructed, it must be assessed as a part of historical trends in ideology, politics, institutions, culture, and political economy far more than as a disembodied store of learning about poverty's "causes" and "cures." By way of introduction, then, and in the chapters that follow, I highlight several insights that can be drawn from historical understanding of poverty knowledge, and that inform my conclusion that reconstructing poverty knowledge is more than simply a matter of generating new research questions for social scientists to pursue.

First and foremost among these insights is that poverty knowledge is fundamentally ideological in nature: It is above all a project of twentieth-century liberalism, dating most immediately from the 1960s and the Great Society, but more deeply rooted in the rise of the "new liberalism" that emerged in late nineteenth-century Euro-American political culture as an alternative to the lais-sez-faire individualism of the industrial age.⁶ Originating, as it did, in this formative period of twentieth-century liberalism, poverty knowledge rests on several characteristic commitments and beliefs: a commitment to using rational empirical investigation for the purposes of statecraft and social reform; a belief that the state, in varying degrees of cooperation with organized civil society, is a necessary protection against the hazards of industrial capitalism and extreme concentrations of poverty and wealth; a commitment, nevertheless, to maintaining a capitalist economy based on private ownership and market principles, however much it need be tamed or managed by public intervention; and, finally, a distinctly secular faith in human progress, not just through the accumulation of knowledge, but through the capacity to apply it for the common good. These core beliefs, to be sure, have been subject to varying interpretations, to internal conflict, and to revision over time. Nevertheless, in one form or another they have defined poverty knowledge as a liberal as well as a scientific enterprise, starting with the efforts by Progressive-era social investigators to depauperize thinking about poverty—to make it a matter of social rather than individual morality—by turning attention from the "dependent" to the wage-earning poor.

As a historically liberal enterprise, poverty knowledge also reflects the diversity and internal tensions within twentieth-century liberal social thought: differences between labor/ left and corporate/centrist liberals over how to manage the economy; between "top-down," elite-driven and "bottom-up," politically empowering approaches to reform; and even between class-based vs. cultural or "identity" politics, as can be seen in a long-enduring debate pitting "race" against "class" as alternative frameworks for explaining and responding to poverty among African Americans. Most fundamentally though, poverty knowledge reflects a central tension within liberal thought about the nature of inequality—not so much over whether inequality is innate or environmental in origin, but whether it is best understood and addressed at the level of individual experience or as a matter of structural and institutional reform. That this tension has more often been resolved in favor of the individualist interpretation can be seen in several oft-noted features in poverty research. One is the virtual absence of class as an analytic category, at least as compared with more individualized measures of status such as family background and human capital. A similar individualizing tendency can be seen in the reduction of race and gender to little more than demographic, rather than structurally constituted, categories. Poverty research treats the market and the two-parent, male-headed family in much the same way, as inevitable, naturally occurring ways of ordering human relations rather than as institutions that are socially created and maintained. The point is that these have not always been prevailing characteristics in poverty knowledge; nor are they simply a reflection of a shift towards economics as its disciplinary base. They grew just as much out of ongoing struggles within liberalism over the ideological boundaries of reform—the outcomes of which, in the name of remaining realistic or "relevant" for political purposes, have repeatedly eclipsed an alternative, more institutionalist and social democratic research tradition, that has challenged liberalism's individualistic assumptions from within. Nor, for this very reason, should we see the outcome in terms of some self-generating, inevitable ideological trajectory, or in terms of an irreversible end to an expansive, social democratic, or participatory vision of liberal reform. Indeed, the ideological boundaries of poverty knowledge have been drawn and redrawn amidst changing political and economic circumstances, and in an ongoing process of negotiation and debate.

Thus, by paying attention to the history of poverty knowledge, we can see that its very development as a science has been closely tied to the shifting preoccupations, to the political fortunes, and certainly to the major crusades of twentieth-century liberalism. Not all of these crusades were tied so obviously to the expansion of social welfare provision: World War II and the Cold War underwrote the anthropological studies in developing countries that fostered Oscar Lewis's infamous theory of the "culture of poverty." So, too, did they provide the occasion for the use of systems analysis in a burgeoning postwar defense industry—a weapon, so to speak, that federal research administrators imported directly from the Pentagon when it came time to fight the War on Poverty. Poverty knowledge was also shaped by domestic social welfare considerations, and none more powerfully than the experience of postwar economic affluence. Eager to push the expansive economy to its "full growth potential," Keynesian economists in the Kennedy administration cultivated an analysis that linked poverty to sluggish growth and less-than-full employment, and its solution to what skeptics considered the unlikely device of a growth-stimulating tax cut. And it was amidst the great African American migrations of the two post-world war periods that poverty knowledge began gradually to exhibit an assimilationist racial egalitarianism, brilliantly synthesized in Gunnar Myrdal's An American Dilemma during the 1940s, that explained differences of race and class in terms of culture rather than biology while implicating cultural exclusion and pathology in the persistence of black poverty. Here in particular poverty knowledge proved capable of accommodating, and to some degree anticipating, the social movements and world transformations that were reshaping liberalism at the time, and that made it sufficiently flexible as an ideology to sustain a loose consensus within a diverse constituency during the decades after World War II.

Nevertheless, as with liberalism, the capaciousness of poverty knowledge could only extend so far before bearing the brunt of internal conflict and battering from without. Thus, by the late 1960s both the culture of poverty and racial assimilationism had generated deeply divisive debates within a social scientific community that was itself being transformed by civil rights and women's movement politics. Similarly, with the end of postwar affluence and the collapse of the "Keynesian consensus," poverty knowledge lost both its link to macroeconomic policy and its central organizing idea. More recently, poverty knowledge has been profoundly shaken by the rise of the political Right, with its ideological, not-always secularist approach to knowledge and its extraordinary success in keeping the locus of discourse away from the economics of rising inequality and centered squarely on issues framed as "family values," "big government," and the decline of personal responsibility. It is in this context that the direction poverty knowledge has taken in the past two decades reflects the fragmentation of liberalism, and its subsequent efforts to reinvent itself on a more limited social base—this time in the guise of the "third way," "new Democrat," or market-oriented neoliberalism that ushered in the end of welfare while wholeheartedly embracing the private market as the ultimate arbiter of individual well-being and the common good. With the turn to dependency as its central concept, the contemporary neoliberal drift in poverty research marks an important break with the earlier "new liberal" past, for it in effect repauperizes the poverty issue while emphasizing individual, rather than social, morality.

A second major insight from historical analysis is that poverty knowledge is highly political in nature, in ways that go beyond its close association with the trajectory of liberalism, and that have led to the emergence of professional social science as the dominant source of expertise on poverty and welfare policy. To some degree this can be understood as part of the politics of knowledge—the ability of well-placed research entrepreneurs to act as advocates for particular approaches, theoretical frameworks, and for the necessity of social scientific expertise as the basis of enlightened policy. It is thanks to such efforts that poverty knowledge bears the markers of professionalization—specialization, standardized data, experimental methods, a body of theory, or at least a series of "testable hypotheses"—along with the mechanisms for training future generations to uphold established standards of scientific expertise. But the triumph of social science as a way of knowing poverty can also be understood as part of the politics of class, race, and gender in determining who qualifies and can participate as an authority—and who not—in the broader public sphere. Seen in this light, poverty knowledge can be characterized as the project of an increasingly credentialed, formally educated segment of the middle class—one that, despite important contributions from prominent female and nonwhite social scientists, has for most of its history been predominantly white and male. Moreover, the claim to scientific objectivity rests on technical skills, methods, information, and professional networks that historically have excluded those groups most vulnerable to poverty: minorities, women, and especially the relatively less-educated working class, putting poverty knowledge in a position not just to reflect but to replicate the social inequalities it means to investigate.

This is not to say that poverty knowledge can be reduced to a playing out of material class interest (populist and conservative critics to the contrary, there really is not much money or professional glamour to be had from studying the poor), nor to deny that individual social scientists have been capable of transcending their class, race, and gender-bound identities. It is to recognize, though, that not only despite but because of its quest for a particular scientific standard, poverty knowledge has been filtered, not just through the experiences and cultural biases of the privileged, but through the social position of "the professors" in relation to "the poor." It is in this regard that recent changes in political economy take on a special significance for poverty knowledge, not just as they affect the demographic "composition" of poverty, but as they pit the more- against the less-educated in the distribution of economic punishments and rewards. In the "new," information-hungry, postindustrial economy, poverty experts are in a position to benefit from the transformations that have destabilized the industrial working class; in economists' language, it is an economy that brings ever-greater "returns" to education while devaluing industrial skills. And yet, poverty experts show little inclination to question whether their own stake in the "new economy" might affect their interpretation that its disparities can be explained primarily as differences in education and skill—suggesting, in a way reminiscent of earlier cultural criticism, that the poor should simply strive to be more like us.

It is this disparity of status and interest that make poverty research an inescapably political act: It is an exercise of power, in this instance of an educated elite to categorize, stigmatize, but above all to neutralize the poor and disadvantaged through analysis that obscures the political nature of social and economic inequality. By the same token, it is the power to construct and give scientific weight to ideas of what is natural, "functional," or socially desirable, in terms that are exclusive of, if not in direct opposition to, the poor. Finally, it is the power to constitute or at least to influence the categories of social policy in ways that are of material consequence to the poor, whether those categories have to do with determining the particulars of who is eligible (or "deserving") of public assistance or with establishing the broader parameters of the welfare state.

The question of categorization in turn highlights a third important insight from historical analysis, and that is the degree to which poverty knowledge has been influenced by social welfare institutions and the categories they establish for channeling (or denying) aid to people who are poor. For just as social scientists and social research have played a part in shaping policy, so, too, has the structure of U.S. social welfare policy played a central role in designating what constitutes poverty knowledge, and in distinguishing it from labor, or economic, or other bureaucratically influenced categories of research. It was not until the War on Poverty in the 1960s that the state officially recognized poverty as a category for investigation, launching a research operation within the newly created Office of Economic Opportunity, adding poverty statistics to the federal census, and adopting an official "poverty line." Before then, the study of poverty had been segmented according to the categories and constituencies of social policy, acknowledged within the bureaucracy as an aspect of maternal and child welfare, old age, or unemployment but not privileged as a problem worthy of an elaborate investigatory apparatus in its own right. Even when infused with the crusading spirit of the Great Society, poverty could hardly be considered a truly "privileged" research category. Ever aware of its negative connotations, research bureaucrats continually struggled with ways to keep the word "poverty" out of their initiatives, while the institutionalized stigma assigned to "poor people's" or "welfare" programs created an incentive for agencies to sharpen, rather than break down, the distinctions between their own constituencies and the poor.

Poverty knowledge reflects the influence of institutional arrangements in other ways as well, and in particular the uneasy, and changing, relations between the state, civil society, and the private market economy that have characterized the twentieth-century American polity. Made possible from the start by the frequently cooperative ventures of state /federal research bureaus and corporate philanthropy, poverty knowledge has been cultivated primarily within a changing array of nonprofit research organizations and social policy "intermediaries" which, though established to be nonpartisan sources of knowledge, presumably independent of politics or the state, have themselves been affected by three major developments in the public/private "mix."

The first is what was, at least up until the 1980s, a fairly steady expansion of the state in the production of social scientific knowledge, resulting in a proliferation of agency research bureaus, along with opportunities for social scientists to move in and out of official government posts. It was not until after World War II, however, that the prospect of often large federal government contracts became a mainstay, as well as a source of legitimacy, in social scientific research. Like other large-scale government undertakings, the War on Poverty played a pivotal role in this regard, generating the impetus for the elaboration of a whole new set of specialized research institutions designed specifically to meet federal demands for policy research. Thus constituted, poverty research developed what by contemporary welfare criteria would have to be considered an unhealthy, long-term "dependence" on the state—certainly a tendency to follow, rather than to set, the parameters of policy debate. A second development, somewhat paradoxically, was the simultaneous tendency to embrace the values of the private market in the organization and production of knowledge—a competitive approach to procuring, and using, research in a federal social research "market" that was constructed to meet political as well as administrative needs. Nowhere was the competitive principle more operative than in the rise, expansion, and ultimate survival of the poverty research industry, due largely to its entrepreneurial capacity to win government contracts even after successive administrations began to dismantle the apparatus of the Great Society welfare state. Indeed, the dramatic devolution of federal welfare responsibility since the 1980s has actually sped the competitive pulse, as state and local think tanks proliferate and state agencies become increasingly important sources of funding once federally controlled.⁷ Contrary to stated expectations, however, the embrace of competitive principles has hardly been a guarantee of independent thought; if anything, it has tied poverty knowledge more closely than ever to a contract market defined by agency needs, and to a narrowly construed policy agenda that has given far higher priority to reforming welfare than to improving living standards for the working class. Thus, the most recent historical development is perhaps most paradoxical of all, and that is the rise of an alternative network of conservative and libertarian knowledge-producing institutions that have managed to exert far greater policy influence by eschewing government contracts, while embracing competitive market principles as the basis for policy as well as for aggressively publicizing their wares.

A fourth set of insights from historical analysis has to do with the nature of poverty knowledge as science: to some degree in the enlightenment sense of progressively accumulated knowledge, but more fundamentally as a product and shaping force in culture—a source of language, interpretive frameworks, even of the stylized rituals of investigation that give expression to broader social understandings of the human condition and of social change. Judged according to the norms of rational enlightenment, poverty knowledge can indeed be credited with certain achievements, even if they have more to do with documenting unequal or substandard conditions than with explaining why they persist. With the help, for example, of longitudinal data, social experimentation, and a wide array of evaluation studies, social scientists have systematically challenged the stubborn mythology that poor people are lazy, nonworking, or for that matter that poor people are all alike. As welfare debates never cease to remind us, however, very little in this body of presumably established learning is uncontested—scientists arrive at very different conclusions even when they use the same data and methodologies—nor has it, as learning, extended much beyond an expert elite. In contrast, scientific poverty knowledge has had a far more lasting impact on the American cultural and social policy vocabulary, albeit with ambiguous, at times contradictory results.

On the one hand is the notion, put forward initially by nineteenth-century social investigators, that poverty is an objective, quantifiable condition—measurable against a scientifically calculated standard of need known as the poverty line. This measure of poverty has since been absorbed into bureaucratic, political, and to a more limited degree popular culture—a way of determining program eligibility as well as an indictment of society's neglect. Equally important, at least in the eyes of its original proponents, is the social conviction the measure implies: poverty is not a mystery of nature; it can be explained, reduced, or eliminated by rational means. On the other hand, and far more ubiquitous in political and popular culture, have been the many social scientific variations on precisely the opposite theme: the notion, variously expressed in concepts such as social "disorganization," "deviance," or "dysfunction"; in metaphors such as the "vicious circle" or the self-perpetuating "tangle of pathology"; and in totalizing theories of the "culture of poverty," or, most recently, the "underclass," that poverty is deeply ingrained in "intractable" psychological and cultural processes that may very well be beyond rehabilitation or reform. Despite its current association with conservative politics, the culture of poverty and its variants gained the imprimatur of scientific objectivity within a liberal research tradition. As can be seen in recent efforts to measure the underclass according to behavioral indicators, they have since achieved the status of quantifiable fact—a status that at least some poverty experts, unable to control the politics of "blaming the victim," have subsequently come to regret. In this sense, at least, poverty experts have proved to be rather ineffective cultural brokers: even when offered in the name of social criticism or as a call to action, their formulations of cultural deviance have been used far more readily and regularly to stigmatize, isolate, and deny assistance to the poor.

Alongside the language that has been absorbed into popular and political culture, over the past three decades poverty knowledge has also cultivated an increasingly technical jargon as the common, if not exclusive, language of poverty expertise. More than simply a question of quantification—the "amateur" researchers of the social survey movement were every bit as quantitative as current-day econometricians—the technical jargon of recent decades has taken poverty knowledge to a level of abstraction and exclusivity that it had not known before. It is a language laced with acronyms that themselves speak of particular data sets, policies, and analytic techniques (PSID, NLSY, TRIM, FAP, PBJI, EITC, and, albeit without a detectable sense of irony, Five Year Plans and a model known as the KGB). It also speaks of a self-contained system of reasoning that is largely devoid of political or historical context, in which individuals are the units of analysis and markets the principal arbiters of human exchange. The effect has been to put entire questions and categories of inquiry outside the boundaries of critical scientific discourse—capitalism, for example, like the institutionalized systems of race and gender relations, does not translate into variables that can be scrutinized within these models of cause and effect.

On the whole, though, poverty knowledge has been perhaps most effective as a form of cultural affirmation: a powerful reassurance that poverty occurs outside or in spite of core American values and practices, whether those are defined in terms of capitalist markets, political democracy, self-reliance, and/ or a two-parent, white, middle-class family ideal. Although present in much of the social scientific literature before then, it was not until the 1960s that this theme became virtually institutionalized in research. That, after all, was when federal officials, designating "poverty" as a distinct social, policy, and analytic category, quite consciously detached it from the language of income distribution, class, and racial inequality. Poverty, to use the terminology of the day, occurs in some "other," separate America; as an aberration, an exception, a "paradox" of plenty rather than as an integral or necessary condition of the affluent society.⁸ Built on this premise, poverty knowledge continues to hold out a certain promise: doing something about, even eliminating, poverty will not require radical change; whether through social engineering, wage subsidies, economic growth, or the new/old-fashioned strategy of pushing people into the market, the paradox can be resolved without resorting to a massive redistribution of power and wealth. It also offers a substitute language, of deviance and deprivation, for the language of inequality. Most important from a policy perspective, it conceptually disenfranchises poor people from the larger political community—experts refer to the "working poor," not the "working class"—and in this way has helped to confine the reform conversation to the problem of welfare rather than the problems of political economy and work.

In addition to these insights about the nature of poverty knowledge, historical analysis provides the basis of a narrative that weaves the various dimensions of poverty knowledge together—ideological, political, institutional, cultural—while paying attention to the ever-changing contingencies of politics, social movements, and critical events. This narrative, as laid out in the next several chapters, is a story of transformation: over the course of the twentieth century, the study of poverty has changed. What started out as a series of reform-minded, sometimes "amateur" investigations into the political, or "social" economy of industrial capitalism has become an ostensibly more detached, highly professionalized and technically proficient inquiry that takes postindustrial capitalism as a given and focuses primarily on evaluating welfare programs, as well as on measuring and modeling the demographic and behavioral characteristics of the poor. This transformation did not occur as a smooth, one-directional, or cumulative progression, but more as a series of "turns," or paradigm shifts. Nor did it take place along a single, clear-cut political or ideological continuum so much as along liberalism's complicated twists and turns. Moreover, within this narrative of transformation are several continuities, in particular an enduring tension that has only recently become polarized along liberal/conservative lines, between a discourse that associates poverty with some form of cultural pathology or difference and one that points to structural barriers in society and political economy. As we shall see, the tension between "culture" and "structure," while long-standing, has not always been sharply drawn. For many, indeed, the existence of a poor, presumably pathological subculture has been both a product of and a reason for redressing structural inequities in the political economy. It is also the case that the tension has been bound up just as much in disciplinary rivalries as in prescriptions for policy. Nevertheless, as poverty knowledge became more and more about poor people and less and less about culture or political economy more broadly defined, the terms of the question became more oppositional: what differentiates poor people—money or culture—from everyone else? It is in this context that "culture" vs. "structure" has come to be regarded as an either/or choice.

Two other themes warrant special mention in this narrative of transformation, and indeed help to explain its twists and turns. One is just how deeply race has influenced the course of poverty knowledge, in the form of racial ideology and racial politics, as well as in the racialized nature of poverty and social policy. Thus, for example, it was at least in part the battle against pseudo-biological justifications for racism that, in the early decades of the twentieth century, helped to draw racially liberal social scientists to culture, both as a way of explaining racial differences and inequities and as a way of showing that they were neither natural nor inevitable. By the 1960s, though, culture was itself becoming a suspect category in poverty knowledge, largely in reaction to an unrelenting, heavily psychologized imagery of black cultural deviance and pathology that, many suspected, had come to replace biology as a basis for scientific racism. In other instances, race has exerted an equally powerful influence as an unacknowledged variable, in analyses that, for political and ideological as well as for scientific reasons, have diminished the importance of racially discriminatory institutions and social practices in explaining racialized patterns of poverty. In this context, poverty has been conceptualized as an alternative to rather than as a dimension of racial inequality—and itself a problem that can be addressed without explicit "race-targeted" policies. Nevertheless, the reality that poverty, and particularly welfare, have themselves become such racially charged political problems has consistently undermined the very possibility of "race-neutral" antipoverty policy.

A second theme running throughout the narrative is that poverty knowledge, especially in recent decades, has frequently assumed far different political meanings than what is envisioned by social scientists. Nowhere is this more apparent than in the transformation of the culture of poverty in the late 1960s and early 1970s from an argument for liberal intervention if not radical social revolution (as Oscar Lewis occasionally hinted) to an argument for conservative withdrawal from the welfare state—a transformation brought about as much by liberal and left critics who drew out the implications of the behavioristic cast of Lewis's theory as by an outright conservative embrace of the culture of poverty theory. While themselves avoiding thorny issues of culture, economists affiliated with the War on Poverty similarly saw their research used for unintended political purposes, when conservative policy analysts effectively appropriated their methods, findings, and to some degree their style of discourse to undermine support for the welfare state in the 1980s. It is not only the culture of poverty, then, that has been absorbed into conservative policy thinking—Charles Murray, indeed, insisted that unmarried mothers grown dependent on welfare were simply responding as any rational actor would to the perverse incentives of the liberal welfare state. It is more a matter of a knowledge base that, however unintentionally, has opened itself to conservative interpretation by locating the crux of the poverty problem in the characteristics of the poor. But the use of poverty knowledge for overtly conservative purposes also reveals an aspect of the relationship between knowledge and policy that liberal or purportedly "neutral" social scientists have continually underestimated—no matter how many times the best-laid plans of empirically informed policy intellectuals have gone either unattended or misconstrued. What matters in determining whether and how knowledge connects to policy is not only the classical enlightenment properties of rationality and verifiability; nor is it only the way knowledge is mobilized, packaged, and circulated; nor even whether the knowledge corresponds with (or effectively shatters) popularly held values and conventional wisdom. All of these things have, indeed, proved important in affecting the course of poverty and welfare policy. Even more important in determining the political meaning and policy consequences of poverty knowledge, though, has been the power to establish the terms of debate—to contest, gain, and ultimately to exercise ideological hegemony over the boundaries of political discourse. It is within this broader context, of ideological battle that for the past two decades has been dominated by the conservative right, that poverty knowledge has been used most effectively for politically conservative ends.

Part One of this book begins with a discussion of what was known during the Progressive Era as "social" economy and its efforts, most fully realized in the social survey movement, to recast public understanding of poverty by emphasizing its roots in unemployment, low wages, labor exploitation, political disfranchisement, and more generally in the social disruptions associated with large-scale urbanization and industrial capitalism. To be sure, Progressive social investigators wrote with conviction about what they considered to be the moral and cultural deficiencies of poor people. But they also used their investigations to frame a much different kind of critical discourse: in the case of the famed Hull House and Pittsburgh surveys, about the policies and institutional practices of corporate capital; about the history and political economy of racial discrimination in the case of W.E.B. DuBois; and even, in studies of women in industrial, agricultural, and household work, about the burdens created by the gendered division of labor. In this way, Progressive social investigators sought to extend the boundaries of antipoverty thinking to issues of industrial democracy, political reform, and trade union organizing as well as to the kind of community-based cultural uplift for which the settlement houses have become renowned.

In chapter 1 I trace the shift from this Progressive "social" economy to Chicago-school "social ecology" as the dominant paradigm in poverty research. With substantial funding from the Rockefeller Foundation, Chicago-school sociologists built a formidable research and training institution, with an emphasis on theory-based, "objectivist" research as the appropriate knowledge base for policy. Emulating the rigors and experimental techniques of the natural sciences, Robert E. Park, Ernest W. Burgess, and their students treated local neighborhoods more as laboratories for research and experimentation than as sites for political organizing, social uplift, or industrial reform. Their naturalistic models of urbanization, assimilation, and social "disorganization" explained poverty as an inevitable by-product of modernization, and looked to more limited attempts to achieve neighborhood and family "reorganization" in response. By the late 1920s, this model of community study and action had largely displaced Progressive-era reform investigation as a source of expertise, while reinforcing a growing professional and gender divide between academic social science and feminized or "amateur" reform research. Equally important, the Chicago-school turn in social investigation marked a shift away from political economy as a framework for understanding poverty, and an embrace of the newer, social psychological and cultural approaches of sociology and anthropology. The implications of these developments were profound: social disorganization and cultural lag, not industrial capitalism, were at the root of the poverty problem in the new social science, and cultural, not industrial, "reorganization" was the cure.

In chapter 2, I show how these themes and methods continued to frame social scientific understanding of poverty during the Great Depression, amidst renewed concern over unemployment, low wages, and class polarization. Shifting away from an initial, antistatist emphasis in social ecology, however, sociologists and social anthropologists used the techniques of cultural analysis, social psychology, and laboratory-like community study to reintroduce and invigorate the case for progressive-style political and economic reform. In landmark community studies by Robert S. Lynd, W. Lloyd Warner, and E. Wight Bakke, poverty was indeed a problem of corporate restructuring, and unregulated capitalist markets, but it was also a sign of the cultural "lags" of a society unable to adjust to the need for a welfare state. At the same time, according to these scholars, poverty also led to deep-seated, potentially self-perpetuating cultural and psychological disorders that stood as powerful evidence of the need for enhanced social engineering to accompany the project of relief and reform. So, too, according to regionalist sociologists at the University of North Carolina, had poverty hardened into cultural affliction in the backward, "colonial" political economy that had earned the South recognition as the nation's "number one economic problem." In chapter 3 I draw out the ambiguities of this turn to culture as manifest in the sociology of poverty and race, showing how racially liberal social scientists used the concept of culture as at once a challenge to the biological racism of earlier social science and as a powerfully stigmatizing way of explaining why such a large proportion of the African American population remained mired in poverty. Drawing alternately from Chicago-school social ecology and social anthropology, sociologists in the 1930s and 1940s arrived at competing explanations for the high rates of poverty among blacks. Those explanations came together, though, in treating poverty as a form of cultural deviance or pathology—whether a legacy of the cultural damage done by slavery, or an expression of the psychologically distorting influence of persistent white racism. It was this formulation of cultural pathology that would most heavily engage social scientific thinking about race and poverty for decades to come, and that, even when invoked as a rationale for greater social and economic inclusion, reinforced the imagery of a basically unassimilable black lower class.

The analytic emphasis on social psychology and culture redoubled in the postwar decades, fueled by a combination of widening prosperity, Cold War politics, and especially by the tremendous expansion in funding for research in the behavioral sciences by private foundations and federal government agencies. Turning away once again from political economy as the focus of investigation or intervention, postwar sociologists and anthropologists concentrated instead on the unique culture and psychology of what they regarded as an isolated class of poor people, sharply distinguished from the more respectable working class, in an otherwise affluent society. As discussed in chapter 4, this notion of cultural isolation also rested on an increasingly psychological understanding of the family, and specifically of gender relations within low-income, and particularly within black families, that drew heightened attention to poor women's reproductive behavior while ignoring their economic role. These ideas about gender relations, expressed powerfully in studies of the impact of the "mother-centered" or "matriarchal" family, were central to the culture of poverty theory developed by Oscar Lewis, and widely accepted in liberal social science by the late 1950s and early 1960s. As discussed in chapter 5, this and other theories of cultural deprivation and social disorganization became the basis of a whole series of sociologically informed, community-based, primarily urban social interventions sponsored by foundations and government agencies, that served as testing grounds for the War on Poverty. Depicting the poor as socially isolated, and culturally deprived, these experiments proved inadequate as a response to the ongoing problems of racial discrimination, suburbanization, and industrial decline that were then reshaping the urban United States. At least in their earliest stages, they also embraced an essentially apolitical vision of deliberative, rational, "top-down" change that the actual experience of community action in the 1960s would quite literally explode.

Part Two of this book focuses on a set of developments that pulled poverty knowledge in a somewhat different direction. It begins, in chapter 6, with the emergence of a new political economy of poverty in the decades following World War II, ushered in by the Keynesian and human capital "revolutions" in economic thought and by the growing influence of economists in the expanding policy apparatus of the federal government. Grounded in market-centered, neoclassical economics, the new political economy returned to the older categories of income, wages, and employment in its definition of the poverty problem, but explained it as an indicator of inadequate economic growth, high unemployment, and individual human capital deficiencies rather than relating it to the unequal distribution of wealth and opportunity. Like its predecessors, this formulation of the poverty problem reflected political and ideological concerns as much as new analytic approaches. Aware of the political hazards, administration economists made a conscious effort to avoid mention of redistribution or economic restructuring in their proposals, emphasizing instead the power of macroeconomic growth, high employment, and individual human capital investment to bring poverty to an end. They also presented their antipoverty initiative as essentially "race-neutral," confident once again that growth and tight labor markets would diminish the need for more overt, politically risky, antidiscrimination policies. In many ways, this approach shared more in common with the psychology and culture of modernist social science than with the political economy of Progressive reform: poverty stemmed not from the economic and institutional relationships of industrial capitalism, but from the individual—in this case skill—deficiencies of the poor. Thus, while uncomfortable with psychological renditions of a problem they sought to redress with economic measures, administration economists nevertheless incorporated the notion of a culture of poverty in their blueprint for the War on Poverty in 1964, and called for programs that would break the "vicious cycle" that had captured the poor.

The tensions within the Great Society idea of poverty soon started coming unraveled, however, when social scientists became embroiled in a series of disputes that left their tenuous "consensus" in disarray. In chapter 7 I discuss some of the less visible of these disputes, between economists and community action administrators in the Office of Economic Opportunity over the kind of social knowledge that was needed to fight poverty. This time, ironically, it was the new political economy, armed with "hard" quantitative data, econometric modeling techniques, and cost/benefit policy analysis, that laid claim to the mantle of objectivity and political neutrality—and that, with the swift political demise of community action, displaced sociology as the dominant framework for poverty knowledge in the OEO. Borrowing from the experience of postwar defense research, OEO economists led the way in creating the institutional infrastructure for a poverty research industry—an industry designed with the needs and interests of government policy makers in mind, principally reliant on federal agencies for funding, and thriving long after the War on Poverty had been abandoned. Chapter 8 then turns to the more public and visible of the poverty "wars" of the mid-to-late 1960s, tracing a series of highly polarized debates over the ideas about culture, race, and poverty underlying administration policy, most prominent among them the debate over the Moynihan Report on the Negro Family following the Watts riot in 1965.

Part Three of Poverty Knowledge follows the fortunes of the poverty research industry in the aftermath of the War on Poverty, when, in the face of growing inequality, wage deterioration, urban deindustrialization, and a profound ideological challenge to the liberal welfare state, the social scientific poverty discourse narrowed even further to focus principally on understanding the "dynamics" of welfare dependency, the skill deficits of the "working poor," and the size and characteristics of the urban "underclass." In chapter 9 I outline the political origins and institutional structure of the poverty research industry, in the form of an interlocking network of government agencies, private foundations, and nonprofit research institutes that operated together to define and contain the boundaries of scientific poverty research. Reflecting a research agenda that was substantially defined by the political obsession with welfare reform, poverty researchers acquiesced to the shrinking parameters of social policy by confining their sights to diagnoses and interventions targeting poor people and their behavior while avoiding the pressing issue of growing disparities in income and wealth. Nor could analysis offer anything more than limited, mostly descriptive explanations for why poverty was on the rise—explanations that, confined as they were to what was measurable in existing databases, invariably pointed to individual-level characteristics as the cause. In chapter 10 I show the poverty research industry faced with challenges that these individualized models could not explain—growing inequality, wage deterioration, deindustrialization, concentrated urban poverty—while grappling with an even more fundamental ideological challenge from the political right. Little wonder, then, that poverty analysts were overshadowed by the more explicitly ideological, heavily publicized explanations offered by Charles Murray and other conservative social scientists who, nominally using the same "neutral" analytic techniques perfected by economists at the OEO, blamed the rise of poverty on the liberal welfare state. In two ways, I conclude in chapter 11, liberal poverty knowledge contributed to the end of welfare in 1996—its acquiescence to a political agenda that had less to do with reducing poverty than with reducing the welfare rolls, and its failure to provide an explanatory knowledge base for an alternative agenda of political and economic reform.

It is with these failures in mind that I conclude by outlining what a reconstructed poverty knowledge might look like, a project that would draw upon the insights from historical analysis to take in the political, ideological, institutional, and cultural as well as the more immediate research agenda-setting dimensions of the task. I aim, with this outline, to start a conversation rather than to offer precise prescriptions for change. The first task is to redefine the conceptual basis for poverty knowledge, above all by shifting the analytic framework from its current narrow focus on explaining individual deprivation to a more systemic and structural focus on explaining—and addressing—inequalities in the distribution of power, wealth, and opportunity. A second is to broaden the empirical basis for poverty knowledge—recognizing that studying poverty is not the same thing as studying the poor—by turning empirical attention to political, economic, institutional and historical conditions, to the policy decisions that shape the distribution of power and wealth, and to interventions that seek to change the conditions of structural inequality rather than narrowly focusing on changing the poor. A third task is to change the way poverty knowledge is produced and organized, shifting away from the state-centered "research industry" model created during the War on Poverty in order to generate more independence and diversity in setting research agendas. A fourth is to challenge the distinctions that associate narrowly construed, hypothesis-testing models of inquiry with "objectivity" while denigrating more theoretical, historical, and structural analyses as "advocacy" or ideology. Above all, a reconstructed poverty knowledge would challenge two fallacies that, despite having been subject to frequent criticism, continue to inform the quest for more or better knowledge about the poor: one, that good social science is a necessarily apolitical, ideology- or "value-free" endeavor; the other, that rational, scientific knowledge about poverty will yield a rational, scientific "cure."

Justice Is Conflict

By Stuart Hampshire

Princeton University Press

Copyright © 1999 Princeton University Press. All rights reserved.