If President Clinton’s health care reform proposal becomes law in anything like the form he has proposed, it will be the first genuinely universal system of social insurance in American history, and a striking departure from all previous efforts to build and expand the welfare state. For until now, virtually all American social welfare policies have taken one of two quite different forms, separate and unequal.

One consists of programs designed for people who have, in theory, earned their benefits, either because of special service to society, as with military veterans, or because they (or their employers) have contributed to the funds from which they draw when they grow old or lose their jobs. The other form consists of programs designed for people whose major claim to benefits is their apparent need for them, people who seem in some special position of dependency that requires a social response. Those programs are what most Americans call “welfare”: aid to families with dependent children, food stamps, and other forms of public assistance to the poor.

During much of the twentieth century at least, programs for those who have “earned” their benefits have been relatively generous and politically unassailable. Programs for those who simply “need” their benefits have been penurious and politically vulnerable. An unacknowledged result of this distinction is that the largest and most reliable benefits of the welfare state go primarily to men; the paltriest and least reliable go disproportionately to women.

How do we explain a welfare system in which those who most need assistance receive it grudgingly and inadequately while those who least need it receive it freely and relatively lavishly? No other major industrial nation is so stingy or so punitive in providing benefits to its citizens. Why has the United States, to use a term long popular among scholars in this field, been such a “laggard” nation?

Theda Skocpol, a Harvard sociologist long involved in trying to explain why states evolve in the way they do, offers a challenging new answer to this old question in an impressive and intelligent study of the history of American social policy from the end of the Civil War to the eve of the Great Depression. (A second volume will continue the story into the New Deal and beyond.) On the one hand, Skocpol challenges the conventional view that the United States was consistently behind Europe in developing social welfare programs, and that it took the New Deal to introduce any important social security system. There were, in fact, extensive programs of public assistance and important constituencies fighting to expand them throughout the half-century she chronicles. On the other hand, she makes clear, the result of this fifty years of experimentation was a feeble, inadequate welfare system, one in which damaging and invidious distinctions were already deeply embedded and ready to afflict future generations of reformers.

Skocpol is one of a group of social scientists who, for the last decade or so, have been trying to reconceive the role of the state in contemporary society. Scholars from the 1950s through the 1970s, influenced by Marxist or other structuralist theories, tended to treat the American state as secondary even in the creation of its own programs. Some of these analysts argued that government policies reflected the influence of broad social forces (industrialization, urbanization, immigration, aging) or of powerful interest groups (workers, farmers, women, the elderly, organized minorities). Other scholars described the state as an agent of the capitalist power structure; institutions of government responded directly (“instrumentally”) or indirectly (“functionally”) to the needs of corporate elites. American social policy, according to such scholars, has been designed by capitalists to dominate or control the working class or other less privileged groups who might threaten the existing distribution of power.1 Throughout these explanations, the state appears as a relatively passive actor, responding to the needs and demands of others.

Skocpol and others have challenged these theories by “bringing the state back in” (a phrase used as the title of a book of which Skocpol is a coeditor).2 They have attempted to show the ways in which the state itself—its own institutional and political structure and its conception of its own interests—shapes social policy in crucial, indeed decisive, ways. To understand why some approaches to policy prevail and others do not, they insist, it is necessary to look at the way power is distributed within government bureaucracies; at the kinds of tasks state institutions are capable of performing; and at the political interests of civil servants and other public officials responsible for shaping and implementing social policy. As Eric Nordlinger, one of the early theorists of this “state-centered” approach to policy, wrote in 1981:

The preferences of the state are at least as important as those of civil society in accounting for what the democratic state does and does not do; the democratic state is not only frequently autonomous insofar as it regularly acts upon its preferences, but also markedly autonomous in doing so even when its preferences diverge from the demands of the most powerful groups in civil society.3

Skocpol herself made an important contribution to this model in her first book, States and Social Revolutions, published in 1979. Unhappy with prevailing explanations for revolutions, which emphasized the power of social groups and insurgent movements, Skocpol argued that the state itself was the principal agent in causing the great upheavals in France, Russia, and China. Prerevolutionary state institutions in those societies proved unable to accommodate the social or military pressures bearing down upon them; their own incapacity, much more than the power of revolutionaries, was the cause of their demise. “Both the occurrence of the revolutionary situations in the first place,” she argued,


and the nature of the New Regimes that emerged from the revolutionary conflicts depended fundamentally upon the structures of state organizations and their partially autonomous and dynamic relationships to domestic class and political forces, as well as their positions in relation to other states abroad.4

Throughout the next decade, Skocpol explored the idea of the autonomy of the state in a number of important theoretical articles and empirical studies of episodes in American policy-making. Much of her recent work has concentrated on the New Deal, which she has also portrayed as a state-centered reform movement, affected by social forces and interest-group pressures, certainly, but not determined by them. In one of her most challenging articles, for example, she and her collaborator, Kenneth Finegold, argued that the National Recovery Administration of the New Deal failed in large part because the state lacked the experience and the administrative capacity to manage effectively such an ambitious program of industrial regulation. The Agricultural Assistance Administration, by contrast, was more successful because a highly developed agricultural bureaucracy already existed.5

Protecting Soldiers and Mothers marks a significant theoretical departure from much of Skocpol’s earlier work. Instead of a “state-centered” explanation of social policy, she talks now of a “polity-centered” model. The state itself remains important, but, she argues, nongovernmental organizations, institutions, and movements that are politically active—for example, unions, veterans’ groups, and women’s voluntary organizations—may also crucially affect policies. Still, the effectiveness of the nongovernmental groups she examines depends largely on the way they operate politically and bureaucratically to get what they want from the state. The public world she describes in this book is a broader and more fluid one than the relatively enclosed world of state bureaucracies she has generally been concerned with, but it remains a world of institutions.

In tracing how a system of social welfare gradually emerged, Skocpol concentrates on three related episodes. She describes, first, the emergence after the Civil War of an elaborate system of pensions for Union veterans—a system that at its peak accounted for nearly half the federal budget and gave benefits to something close to a majority of aging males (and a significant number of widows and children of veterans as well) in the northern states. She looks, second, at the failure of efforts to transform Civil War pensions into a permanent system of universal insurance for workingmen and the elderly. By World War I, the Civil War generation had dwindled, and the pension system constructed for its members dwindled (and ultimately died) with them. And she examines, finally, the emergence in those first decades of the twentieth century of a set of social policies designed to assist women and mothers—policies that won favor within the polity precisely because they were not universal. They represented, rather, assistance to a group of people widely considered vulnerable, dependent, and specially deserving of social support.

Skocpol describes the two models of social provision—the universal model, of which Civil War pensions were a crude early version, and the categorical model, which established women as a group requiring special protections—as “paternalist” and “maternalist” respectively. The paternalist model was rooted in a concept of the family wage, in which the earning power of the male breadwinner was the principal target of public concern. The maternalist model was based on a vision of women (and, above all, mothers) as a special class, with unique claims on society’s attention precisely because they could not—and should not—be expected to fit into the prevailing wage-earners model. The United States, she concludes, rejected the paternalist model and embraced the maternalist one. The question she tries to answer is why.

Civil war pensions, Skocpol maintains, flourished in the late nineteenth century because of the power of the Republican Party and its stake in securing the loyalties of Union veterans. The pension system was popular with voters (and above all with the Grand Army of the Republic, the Civil War veterans’ organization closely tied to the Republican Party). Its administration, moreover, became an important source of patronage for the party and, not coincidentally, an important source of graft.6 But the pension system also became the basis of what Skocpol calls a “precocious social spending state.” As such, it challenges the conventional view that the federal government took no important part in protecting social welfare before the New Deal, and that the United States lagged far behind other industrial nations in providing for its citizens in these years. The Civil War pension system was, in fact, more elaborate and more comprehensive than the social insurance schemes being created in much of Western Europe at about the same time. (It was also one of the first spending programs to offer benefits to African Americans, many of whom had fought for the Union.)


Why, then, did this vast system of pensions shrivel and ultimately disappear in the United States while considerably more modest systems took root and expanded in Europe? One reason, Skocpol suggests, is that the public justification for the pensions (as opposed to the more cynical political calculations that really underlay them) rested on the “moral worthiness” of the recipients. “Civil War pensions were idealized,” Skocpol writes, “as that which was justly due to the righteous core of a generation of men [and their survivors, both widows and children]…for their valiant service.” Although many people who had not served in the Union Army received pensions through fraud or political favoritism, there was never any general presumption that the veterans’ pensions should be universal. Confederate veterans—even those who had fought for the United States in earlier wars—and post–Civil War immigrants were excluded.

But the more important reason for the frailty of the pension system was its intimate connection with patronage and political parties. To the generation of “progressive” reformers emerging in the late nineteenth century, nothing seemed so important as limiting what they considered the corrupt and reactionary power of the entrenched political parties. Civil War pensions, they came to believe, were notable above all for the graft and fraud they produced and for the way they reinforced the dangerous dominance of machine politics and “patronage democracy” in public life. Stories abounded in the late nineteenth century of fraudulent pension claims and undeserving claimants (similar in kind, if not in detail, to stories Americans still hear of how “welfare queens” and others cheat the system). Rather than as a model for future forms of social provision, the Civil War pensions came to be seen by reformers as an obstacle to “good government.” Their principal legacy, in the end, was to serve as an example to reformers of what government should not do.

And yet there were, Skocpol makes clear, significant forces at work struggling to make Civil War pensions the basis for a system of “more permanent public benefits for all working and elderly Americans.” Particularly active was the American Association for Labor Legislation (AALL), an organization that lobbied strenuously for various systems of universal working-men’s insurance throughout the late nineteenth century. But the AALL was a “top down” organization, whose members consisted mainly of academics, lawyers, progressive businessmen, and other reform-minded elite white males. It had no significant following among workers and members of the lower-middle class. It tried to compensate for this weakness by making alliances with labor organizations; but the American Federation of Labor, the AALL’s most likely ally, consistently spurned them. Samuel Gompers and the rest of the AFL leadership considered the state essentially hostile to their aims. Hence they opposed most government efforts to assist workers, fearful that such efforts would serve ultimately to limit rather than expand labor’s achievements. Collective bargaining through a strong trade-union movement, they believed, was the best way to advance the interests of the working class.

In the absence of a strong political alliance that cut across class lines, the advocates of universal social insurance could not overcome the many obstacles in their path. The courts were hostile to the idea, and so were most members of Congress, which lacked strong central leadership comparable to that in European parliamentary democracies; and there was no permanent civil service with a stake in social programs with which the AALL could make an alliance.

In the end, the only element of the AALL’s universalist vision to become law was the system of workmen’s compensation laws that proliferated throughout the nation in the early twentieth century. Between 1910 and 1920, forty-two of the forty-eight states adopted some form of compensation for workers injured or killed on the job. But that was because there was already a considerable judicial, institutional, and political basis for such laws in the practices that most states (and many employers) had been following for years. Workmen’s compensation laws, Skocpol writes, represented “a reworking of the American polity’s way of handling problems already under public jurisdiction,” not an effort to impose fundamentally new missions on governments.

At the same time that these efforts to create a universal, “paternalist” welfare state were failing, attempts to create a “maternalist” system of social provision—aimed at the needs of women and children—were having considerable success. More thoroughly than any previous scholar has done, Skocpol has examined political battles in the various states that led to the passage of forty-one maximum-hour laws, sixteen minimum-wage laws, and other important protective measures for women workers in the first two decades of the twentieth century. She describes the “wildfire spread” of mothers’ (or “widows”‘) pensions in forty states between 1911 and 1920. She chronicles the more familiar stories of the establishment in 1912 of the Children’s Bureau, a federal agency designed to promote the health and welfare of children; and the passage in 1921 of the Sheppard-Towner Maternity Act, which provided federal funds to states to finance agencies and programs to help children, particularly the very young children of poor mothers.

Why did these “maternalist” measures enjoy such impressive success while comparable “paternalist” initiatives consistently failed? Again, Skocpol emphasizes institutional factors (as opposed to more traditional explanations based on popular sympathy for women or on the prevailing male notion that the proper role of women was to care for children). First, the campaigns for maternalist laws succeeded because they had the support of an impressive and powerful network of women’s organizations. Unlike the AALL and other male-dominated institutions, these institutions were broadly based and national in scope. With millions of members and thousands of activists, they were able to mobilize a large and influential element of middle-class opinion. They were skilled at public education, political agitation, and legislative lobbying. Particularly important were such groups as the Women’s Christian Temperance Union, the General Federation of Women’s Clubs, and the National Congress of Mothers—all of which were remarkably successful in mobilizing educated, middle-class women throughout the country to work simultaneously on behalf of “maternalist” goals.

Here Skocpol challenges (or at least expands upon) the work of a number of women’s historians who have written extensively about such eminent reformers as Jane Addams, Florence Kelley, Julia Lathrop, and Grace Abbott, and about the relatively localized or elite institutions (settlement houses, universities, the National Consumer’s League, and others) in which they worked, in winning passage of women’s and children’s protective legislation. Such women and institutions were important, she concedes, but the success of their efforts was crucially dependent on the mass mobilization of “ordinary” middle-class women by less well known female organizations.

Most American women, of course, could not vote during the first two decades of the twentieth century, the period in which women’s organizations enjoyed their greatest political successes. But Skocpol argues that women—because their efforts could appear genuinely nonpartisan, unconnected to the corruption of patronage democracy—actually became more influential politically because they were not voters. Voluntary organizations such as the National Conference of Mothers were virtually the only outlet for the political energies of many women. They were remarkably energetic in lobbying state legislatures, petitioning courts, and gathering the kind of “expert” data on social problems that reformers so venerated in an age of investigation.

Indeed, the political power of women’s organizations seemed to decline after women won the vote in 1920. That was partly because the victory of the suffrage movement convinced many women that efforts outside the political parties were no longer necessary, that their reform hopes could now be achieved through elections. The fate of the Sheppard-Towner Maternity Act revealed how wrong that assumption was. Congress passed Sheppard-Towner in 1921 in part because many members assumed that the newly enfranchised women would vote more or less as one on issues of concern to women and that suffrage would greatly increase the political costs of opposing programs to assist mothers and children. It quickly became clear, however, that women would not vote as a bloc, that the distribution of the female vote would, in fact, break down along almost precisely the same lines as did the male vote. By the late 1920s, when Sheppard-Towner came up for renewal, members of Congress no longer feared the “women’s vote” and voted it down.

Skocpol’s story, therefore, has both a clear beginning and, to some degree, a clear end. The creation of the Civil War pensions program, she argues, marks the real birth of modern social welfare in the United States. The death of Sheppard-Towner in 1929 marks the effective end of the first phase of the construction of the welfare state. The “paternalist” efforts that the Civil War pensions began left behind a very limited number of programs intended to apply to all wage-earners, most notably the workmen’s compensation laws in most of the states. The “maternalist” efforts of women’s groups left a much denser network of laws and state institutions to protect and assist mothers and children. But both efforts petered out rapidly in the decade after World War I. On the eve of the Great Depression, few contemporaries would have predicted the enormous strides the American welfare state would make in the 1930s.

Skocpol’s account of these developments in the history of social welfare is particularly impressive in its reconstruction of the detailed history of welfare programs that relatively few scholars have examined before. But the principal importance of her book lies in its important new interpretation of social policy.

It may seem obvious to say that one cannot understand how a government behaves without looking at the nature and structure of that government. But anyone familiar with scholarship in history and social science during the last thirty years knows that many books have been written explaining state policies with almost no reference at all to the state. Skocpol’s most important achievement in this book is to make it virtually impossible for anyone to explain the history of the welfare state in America without paying serious attention to the structure and behavior of political institutions.

At the same time, she offers a valuable corrective to the way many scholars interested in social policy (including at times Skocpol herself) have worked in the past. By expanding her definition of the public world to include voluntary associations, labor unions, and other nongovernmental organizations, she also suggests ways to connect the often insular world of political history to a larger social and cultural world.

And yet to a historian, at least, it sometimes seems as if Skocpol has not followed some of the paths she has herself suggested, and that she has been reluctant at times to pursue some of the logical implications of her own movement from a “state-centered” to a “polity-centered” approach to policy-making. Time and again in her book, social forces that badly need detailed discussion remain unexplored. While the growth of Civil War pensions, as Skocpol argues, was a result of a series of political calculations by Republican politicians responding to institutional pressures from veterans’ groups and others, those pensions grew most dramatically during the 1890s—a period of unprecedented social and economic turmoil characterized by labor unrest, urban disorder, agrarian protest, and a depression worse than any in American history to that point. Is it really imaginable that these social forces had no effect at all on the way policy-makers behaved? Skocpol would no doubt agree that they did have an effect, but she makes almost no effort to explore what that effect might be.

The failure of male reformers to establish a universal system of pensions in the early twentieth century was, Skocpol argues, largely a result of the AALL’s institutional weakness and the AFL’s hostility to government. But what role did inherited cultural assumptions and moral judgments about work, dependency, and public responsibility have on the way these organizations (and other parts of the polity) behaved? Skocpol largely leaves out such cultural influences, in part because she equates “values-centered” explanations with a discredited vision of American history based on the assumption that there is a single, universal liberal tradition.7 Skocpol is right to say, as she does in her introduction, that we must not only ask what American values are, but also “whose ideas and values? And ideas and values about what more precisely?” In her own work, however, “ideas and values” get relatively little attention.

While America’s patchwork version of the welfare state has no doubt been shaped by the political and institutional forces, it is also a result of powerful (if contested) moral convictions that have affected discussions of social policy throughout American history. As Joel Handler and Yeheskel Hasenfeld argue in a recent study of twentiethcentury welfare reform, “social welfare policy cannot be fully understood without recognizing that it is fundamentally a set of symbols that try to differentiate between the deserving and undeserving poor in order to uphold such dominant values as the work ethic and family, gender, race, and ethnic relations.”8

Finally, Skocpol argues powerfully and persuasively that the nature of the welfare state cannot be understood without taking into account how gender differences helped create it. But her use of gender at times seems curiously limited. It is never entirely clear, for example, why the universalist model of social provision she describes so effectively in the first section of her book is, in fact, “paternalist” in any sense other than that its principal supporters were men. Those who favored “paternalist” reforms, she argues, wanted to sustain the “family wage,” a wage based on the income potential of the male head of household. But supporters of “maternalist” reforms believed in the family wage, too. They usually departed from their male counterparts in the early twentieth century not by envisioning an economy in which men and women had equal access to wages, but by trying to create programs for widows, abandoned mothers, and other unfortunate women who were left without a male provider. The question remains, then, how did ideas about welfare (as opposed to the institutions supporting different versions of the welfare state) reflect differences in gender.

Here Skocpol’s book is related to (and challenges) new work by the historian Linda Gordon, who has published several important articles about twentieth-century social policy that suggest different ways of exploring these questions.9 Some of the features of welfare programs that have favored men over women were conceived, Gordon shows, by women reformers themselves. In a discussion of the Social Security Act of 1935, she argues, “Female welfare activists designed programs that would benefit women directly when male wage-earners failed them, but they did not usually support programs that encouraged women’s independence from men.” Instead of supporting programs to give women education, job training, child care, or access to employment, female reformers favored welfare payments to women with dependent children. They did so because both men and women embraced many of the same assumptions and values about “maleness” and “femaleness” that shaped (and limited) their sense of what was possible or desirable in public policy. Skocpol clearly disagrees with some of Gordon’s conclusions; but even her disagreements would be more persuasive if she had given more attention to attitudes toward gender. 10

Protecting Soldiers and Mothers seems to me the most important study yet written on the early origins of the welfare state—a brilliant corrective to previous scholarship and a work with which everyone approaching this subject must now contend. But while Skocpol has answered many central questions, she has also—perhaps without meaning to—raised others that her own theoretical assumptions leave her unable to answer. Historians hoping for a fully satisfying picture of the creation of the American welfare state will have to find a better way than either Skocpol or her critics have yet done to resolve the deep but ultimately artificial distinction between “statecentered” and “society-centered” explanations of political change.

Skocpol’s account of the origins of social policy ends (for now) with the start of the Great Depression, when most Americans believe the story of the welfare state actually begins. Thomas H. Eliot’s brief, charming memoir suggests something of how dimly aware the framers of New Deal social policies were of the institutional and ideological legacies that shaped their actions.

Eliot was a young lawyer serving in the Roosevelt administration in the mid-1930s. He worked with Frances Perkins, Harry Hopkins, Edwin Witte, Arthur Altmeyer, and others to draft the first Social Security Act; and he served on the first Social Security Board. And while he later went on to become a member of Congress from Massachusetts and the chancellor of Washington University in St. Louis, he apparently considered his years in the New Deal the most important and rewarding of his life. His memoir, published only a few months after his recent death, describes nothing else.

Several things strike one about this book. The first is its picture of the almost haphazard way in which policymakers constructed the social security system as they adjusted their plans at every point to the decisions of the Supreme Court (which they considered hostile to their aims) and to their sense of public opinion (which they considered fickle and unreliable). Eliot adds little to what we already know about the history of Social Security (or the New Deal generally), but he does convey a vivid sense of the uncertain, groping quality of the process by which both were created—a process that now seems quite similar to what we know of the one that created President Clinton’s health care plan.

Eliot’s account is also significant, however, for the things it does not say. He writes as if he and the others drafting the Social Security Act were operating in something of a historical vacuum; as if there were no precedents or structures already in place guiding or inhibiting them. That is, no doubt, the way many of the drafters felt—that they were doing something genuinely new. But of course there were precedents and structures shaping what they did—including many Skocpol has revealed—that were all the more powerful, perhaps, for being unacknowledged.

Eliot says nothing about differences in gender in this book, just as most New Dealers (including many New Deal women) were never aware of the effect of such differences on their thinking. But the Social Security Act was crucially shaped by assumptions about gender. It created one set of programs—old age pensions, unemployment, and disability insurance—primarily for men. (Large categories of women workers, most notably domestic workers, were excluded at first from those programs.) It created another set of programs—aid to dependent children, most notably—for women. The enormous differences between those programs—the former securely established as “social insurance,” the latter weakly structured as something like a “dole”—reflected an inherited set of ideas about the relationship between gender and economics.

In the process, Social Security also codified the distinction between “deserving” and “undeserving” welfare recipients. Retired and unemployed workers received their benefits automatically. Women with young children and others who claimed some form of “dependency” had to undergo elaborate screening and humiliating “means tests” before receiving their benefits. Whatever else has changed during the nearly sixty years since passage of the Social Security Act of 1935, the different ways in which social policies treat men and women, and the distinction they make between the “deserving” and the “undeserving” poor have proved remarkably durable.

As the Clinton administration attempts to reform our health care system and “end welfare as we know it,” it might find the long, often tortuous history of social provision in America both encouraging and sobering. Early twentieth-century women reformers, as Skocpol argues, “found ways to create encompassing and geographically far-flung women’s organizations” and often “stressed solidarity between privileged and less privileged women, and honor for values of caring and nurturance.” Reform efforts in our time, no less than in the past, will need to enlist broad public support; and today’s feminists, Skocpol suggests, might well profit from these earlier examples as they attempt to promote and shape the changes the Clinton administration has promised.

And yet as Thomas Eliot inadvertently reminds us, even at moments of great optimism, unacknowledged preconceptions—about politics, about gender, and about morality—can shape and distort the boldest programs. Understanding the ways in which damaging and invidious distinctions have crept into our welfare system is a first step toward thinking about ways to remove them.

This Issue

May 26, 1994