This file is also available in Adobe Acrobat PDF format
THE AMERICAN WELFARE STATE is known far and wide as a chronic underachiever. Historically, the United States created social programs later than many nations. Currently, essential pieces of the American welfare state are poorly developed or completely missing. Eligibility rules and benefits can vary widely from state to state. The United States spends less on social programs than most affluent democracies, and its rates of poverty and inequality are substantially higher. Analysts have cataloged these shortcomings for years and asked variants of the same basic question: How could a nation that has been so wealthy for so long fail to develop a real welfare state? Some have then attributed the failure to national values, saying that Americans have never really wanted to have much of a welfare state. Others have pointed to the fragmentation of public authority, within and among institutions, that makes any significant policy change hard to enact. Alternatively, the weakness of organized labor or the absence of strong left-wing political parties may have been the main culprits. The list goes on.
This book takes issue with the thinking behind the question and many of the stock answers to it. In the process, it offers a much different portrait of U.S. social policy. Although the American welfare state did not turn out just like its relatives in Europe and Canada, it did become large and far-reaching. The general public has called for more government involvement, and elected representatives from both parties have responded. U.S. officials have created a variety of social programs throughout the twentieth century. In recent decades, they have managed to expand some programs for the poor and near poor. There has actually been a remarkable amount of activity in the American welfare state—but you need to know where to look. And if you do know where, then all sorts of stories about failure and obstruction in social policy become suspect.
On the other hand, all this activity has not had much impact on the core problems of poverty and inequality. If traditional accounts of the American welfare state boil down to “little effort, little progress,” my account highlights the paradox of lots of effort and little progress. No matter how successful advocates have been in creating and expanding social programs, certain features of American politics have limited the benefits of many official changes in policy. Some of their accomplishments have even made it harder to reduce poverty and inequality. The American welfare state contains greater possibilities and tighter constraints than commonly believed. Recognizing these possibilities and constraints is essential for anyone who wants to understand the American welfare state or change it.
This book challenges a number of myths, misunderstandings, and half-truths about the American welfare state. Some pieces of conventional wisdom are based on questionable assumptions, or just plain wrong. In other places, analysts have the facts right but the explanations or implications wrong. These errors are not random; they spring from at least four sources. The first is that many analysts use European welfare states as their reference point when investigating the American welfare state. Historically, European welfare states (and Canada) have relied heavily on social insurance programs. Because the United States does not meet this standard, it seems to lag behind. While this approach may lead to a number of interesting and important comparisons, it misses alternative ways of addressing social needs. The American welfare state happens to rely less on social insurance and more on tax expenditures, loan guarantees, and social regulation than welfare states elsewhere. Once we include these other tools of government, our understanding of the size, shape, and historical development of the American welfare state changes substantially.1 Social insurance is one way to build a welfare state but not the only way. The significance of this error is introduced in chapter 1.
A second source of error has been overreliance on a few social programs, particularly Social Security and welfare, to support more general claims.* Social Security is supposed to represent everything that is right with U.S. social policy and welfare everything that is wrong. From these two programs we are supposed to make all sorts of inferences about social policy, many of which turn out to be faulty because many social programs do not work like Social Security or welfare. It’s like visiting California and Louisiana and assuming that you now understand the entire United States. This error is first discussed in chapter 2.
Third, while many economists and policy analysts examine the recent performance of social programs, scholars interested in the politics of social policy have concentrated primarily on developments between the late nineteenth century and the early 1970s.2 Scholars who have ventured into the latter decades of the twentieth century usually focus on a single program, legislative milestone, or group of beneficiaries.3 Few scholars have tried to generalize broadly about recent developments in the American welfare state.4 Chapter 3 begins to make the case for looking more closely at recent decades. What was true about social policy in the middle of the twentieth century was not necessarily true at the end.
Am I arguing that the conventional wisdom has always been wrong? No. In a number of instances, the old truths worked well for a time. Unified Democratic control of government was essential to passage of new social programs in the 1930s and 1960s. Many of the means-tested programs emerging from the New Deal and the Great Society were politically vulnerable. Public support for the welfare state was lower in the United States than in Canada and Europe in the 1970s. But none of these “old truths” does a very good job of capturing the world we live in now. The American welfare state has changed in recent decades, and our understanding must change with it.
All three of these errors contribute to faulty inferences based on limited evidence. The final source of error is methodological. Unlike other parts of the social sciences, the study of social policies has seldom devolved into battles over the One Right Way to acquire useful knowledge.5 Scholars have certainly disagreed about the relative influence of social movements, public opinion, political elites, and other factors. But seldom have they faulted each other for choosing a fundamentally flawed research design or relying on qualitative versus quantitative evidence. Some scholars are quite good at admitting the limitations of their own research methods and encouraging readers to consult work by scholars from alternative research traditions. The prevailing attitude is “live and let live.”
One problem with this attitude is that it can lead to peaceful segregation rather than genuine integration. True research communities, with scholars working on similar problems using different but complementary research strategies, are rare. We have many qualitative case studies analyzing developments in the first two-thirds of the twentieth century and many large-N quantitative studies aimed at the last third of the century.* We have fairly distinct groups of people working at the cross-national, national, and state levels. What emerges from the literature is a long list of important insights, affecting any number of social policies and historical eras, that never quite add up to a coherent whole. So much has been written by scholars marching happily, and often productively, in very different directions.
This book is unusual in pulling different research traditions together in a single study of U.S. social policy. It draws on three of the most important traditions—the policy analytic, the behavioral, and the historical—each of which has something valuable to offer. Scholars working in these traditions hail from political science, history, economics, sociology, and law. In some chapters, one tradition is featured; in other chapters, two or three traditions work together. If these traditions remain separate, then myths and misunderstandings about the American welfare state will likely persist. At a minimum, integrating these approaches will enable us to ask better questions about the politics of U.S. social policy.
THE PLAN OF THE BOOK
This book grew out of my experiences teaching a course called The American Welfare State to college students. I originally designed the course to convey general patterns of social policy making. Over the years the course evolved into a list of the many exceptions to these patterns, and recent editions of the course could fairly be titled (How Not to Think About) The American Welfare State. This book is my way of trying to state those objections more clearly and connect them in interesting ways. Each chapter typically opens with some piece of conventional wisdom, followed by evidence that it is wrong or seriously incomplete. Wherever possible, I have tried to offer some new generalizations about social policy to replace the old ones. Essentially, this book is a recurring process of demolition and reconstruction.
My attacks on the conventional wisdom cut across ideological lines. I challenge claims that liberals make about the size and shape of the American welfare state, and about the vulnerability of means-tested social programs. By the same token, the chapters concerning the impact of public opinion, of federalism, and of race on social policy will probably not sit well with conservatives. Such evenhandedness could signify a relentless pursuit of truth, or it might simply reflect a perverse desire to pick a fight with as many people as possible (personally, I prefer relentless).
Because I want this book to be used in classrooms, I presume little in the way of prior knowledge. Social science jargon is kept to a minimum, and most of the quantitative analysis is displayed in simple tables. These steps should have the added benefit of making the book accessible to a more general audience interested in social policy. In doing so, I run some risk of losing my audience of scholars and specialists, but I hope that a wide-ranging attack on the conventional wisdom might hold their interest. Some of the chapters end with ideas for future research, and I hope that some readers will be intrigued enough to investigate. The book is intended to spur debate over what is and is not possible in U.S. social policy, and I do not expect to have the last word.
Part 1 surveys the size, shape, and historical development of the American welfare state. The underlying premise of this “Basic Tour” is that we need to know what the American welfare state looks like and how it began before we can start analyzing its politics. Rather than cite a long list of numbers and dates, I organize the most important facts around a few key debates. Chapter 1, “She’s So Unusual,” challenges the notion that the American welfare state is considerably smaller than its European counterparts. For one thing, different measures of size yield very different results. Moreover, the closer you look at the full range of policy tools, the bigger the American welfare state appears. The key question to ask is not why the United States does less than other welfare states but why the United States does things differently.
Chapter 2, “Tracks of My Tiers,” analyzes the overall shape of the American welfare state. We have been told repeatedly that the American welfare state has two distinct tiers. The upper tier of social insurance programs is supposed to enjoy numerous advantages, politically and programmatically, over the lower tier of public assistance programs. One has only to contrast Social Security with welfare to appreciate the gulf separating these two tiers. And yet, the more we learn about other social programs, the more suspect the two-tiered model becomes. Important programs such as Medicaid, the Earned Income Tax Credit (EITC), unemployment insurance, and workers’ compensation do not fit very well in their designated tiers. Tax expenditures, social regulation, and other tools of social policy further compound the problems of the two-tiered model. It may make more sense to think about how different tools of social policy are used to help people at different stages of their lives.
Chapter 3, “Twice in a Lifetime,” lays out a basic chronology of the American welfare state with emphasis on the origins of social programs. Everybody knows, for example, that the American welfare state emerged dramatically during two “big bangs” of activity in the mid-1930s and mid-1960s. Social Security, welfare, unemployment insurance, Medicare, Medicaid, and a number of other programs originated during these two brief moments in time. Political scientists, historians, and sociologists have long debated what made these moments possible—extraordinary presidents, social movements, policy experts, struggles between business and labor, or something else. That debate is less compelling if a number of social programs were created at other points in U.S. history, which they were. They may not have been the kinds of social programs found in Europe, but they addressed many of the same problems and groups. This revised history leads to a greater sense of political possibilities, particularly with respect to political parties.
The emphasis in Part 1 is on breadth: comparing the size of welfare states in many countries, comparing important features of many U.S. social programs, and finding patterns in the enactment of many social programs across the entire twentieth century. Part 2 of the book offers more analytic depth. Chapters 4–7 focus on interesting anomalies and causal relationships that can lead us to a richer understanding of the American welfare state. This part is titled “New Horizons” because one common theme is how scholars have overestimated the constraints on social policy. Chapter 4, “Ogres, Onions, and Layers,” picks up where chapter 3 left off by comparing three recent cases of social programs enacted under divided government—the Employee Retirement Income Security Act, the Americans with Disabilities Act, and the Child Tax Credit. The most surprising finding is that Republican officials were pivotal actors in each case. The party usually known for bashing the welfare state has been instrumental in creating new social programs.
Problems with the two-tiered model of the American welfare state are analyzed more fully in chapter 5. After all that we have read and learned about the vulnerability of means-tested programs, how in the world did the EITC grow faster than any other major U.S. social program between 1980 and 2000? Why were so many politicians, liberals and conservatives, singing its praises? Did its success have anything in common with the equally remarkable expansion of Medicaid during the same period? These questions are not asked very often. Many scholars believe in the old saying “Programs for the poor are poor programs,” meaning that they are doomed to fail politically. Inclusive programs like Social Security are supposed to offer the only blueprint for sustainable social policy. By comparing the recent history of Medicaid and the EITC with that of other means-tested programs, I will demonstrate that some programs for the poor are much less vulnerable than others, and I will offer reasons why this is so.
In chapter 6, I return to the cross-national comparisons introduced earlier in the book. National values have often been cited as the main reason why the American welfare state is relatively small, but if analysts have been wrong about its size then they may be wrong about the impact of values as well. This chapter opens by comparing public support for the welfare state in a handful of affluent democracies. While support in the United States may lag behind the support in other countries, it is strong in a number of areas. The rest of the chapter shows how durable and broad this support has been. Contrary to reports of a “right turn” in American politics, public support for the American welfare state has not waned in recent decades. Even people who call themselves Republican want to spend more on some social policies. Public opinion data help explain the size and distinctive shape of the American welfare state, as well as some of the curious behaviors described in chapters 4 and 5.
Despite recent efforts to help the disabled, the poor, and families with children, the fact remains that the largest U.S. social programs benefit the elderly. These older individuals just happen to belong to one of the largest and most feared interest groups in the country, AARP, and just happen to be very involved in elections. Much less public money is spent on younger families with children, who are less organized than senior citizens and either cannot vote or do not vote as often. Therefore, older Americans, through sheer political muscle, must be controlling the direction of social policy—unless, of course, lots of people who are not elderly agree with these spending priorities. Then AARP would not look so imposing, and claims about generational war would be highly exaggerated. Chapter 7, “The World According to AARP,” raises doubts about the political clout of senior citizens by looking carefully at the history of AARP and comparing the policy preferences of older and younger Americans.
Part 3 of the book, “Checkpoints and Roadblocks,” helps explain how a welfare state that is larger, more diverse, and more popular than commonly believed can nevertheless fall short in addressing social problems. These chapters indicate how federalism, race, public opinion, and existing social programs may be larger barriers than commonly believed. Many policy makers think of the American states as “laboratories of democracy” where different experiments in social policy can be conducted and evaluated, which ultimately leads to better outcomes. The history of workers’ compensation offers clear evidence to the contrary. In chapter 8, I explain how the American states have retained control of workers’ compensation despite several efforts to remedy its flaws by increasing the role of the national government. The main message of this chapter is how political power trumps problem solving.
Many scholars believe that the negative influence of race on social policy has been eliminated or confined to welfare. I find in chapter 9 that race continues to affect a number of social programs. Its impact even extends beyond programs for the poor. The main evidence comes from comparing benefits across the American states. That race still matters seems clear; exactly how it matters is explored but not settled.
In the final chapter, I start to explain how a welfare state that is large and dynamic can nevertheless do relatively little to reduce poverty and inequality. A number of observers believe that these problems persist because the have-nots in the United States lack political power. Most of the poor and near poor, however, do not want the government to redistribute income. In fact, the public may be more sensitive to the gap between the rich and the middle class than between rich and poor. And social policy may reflect those concerns. Chapter 10 ends on a rather pessimistic note, finding that many of recent achievements in social policy benefit middle-and upper-middle-class individuals more than the poor and near poor. We seem to have trouble making progress against inequality because large parts of the American welfare state are not designed to do so. There is a large gap between creating social policies and programs and actually improving the lives of needy citizens.
Had I been trained as a historian, these chapters would have been arranged chronologically. As a political scientist, I am inclined to organize my thoughts based on specific questions or debates. Each chapter is therefore devoted to a single question such as “Is the American welfare state unusually small?” or “Why are some programs for the poor quite large and popular, while others are politically vulnerable?” The organization of my arguments is not so much linear as layered. We will move back and forth across the twentieth century, depending on the question at hand. Doing so will help us understand how the American welfare state is itself layered.
Had I approached this book like most political scientists, I would have narrowed the focus and explored one or two questions in depth. Readers who want to learn a lot about Social Security, poverty, the rise and fall of the Clinton health plan, or any number of topics might want to read the studies cited earlier in this introduction. It is in large part because we have so many good studies of specific programs and policies that the time is right to draw general lessons from them. What I offer instead is breadth of coverage and some synthesis—a trip past well-known and lesser-known trees, and a sense of the whole forest. My story is not simple: it includes many social programs and a variety of policy tools; it highlights decisive moments and long-term trends; it moves between different levels of government; and it tries to account for the influence of prominent officials, political parties, interest groups, public opinion, federalism, race, and policy design. But the American welfare state is not simple, either, and trying to grapple with its complexity is essential to understanding its politics.
METHODS AND DATA
One of the distinguishing features of this study is that it draws on three separate research traditions, using the strengths of one to compensate for the weaknesses of another. The first of these traditions is policy analysis. Policy analysts are adept at describing and evaluating programs, policies, and government agencies. They possess considerable factual knowledge about the American welfare state: they understand the ins and outs of eligibility, financing, and benefits. This wealth of specific knowledge enables policy analysts to challenge facile generalizations. Another distinguishing feature of this tradition is the attention paid to the real-world impact of social policies. Policy analysts want to know if governments are using scarce resources effectively and efficiently to address social problems. They want to know whether a particular government action reduces poverty, and at what cost. Their research is filled with surprising results and practical wisdom, and many of my students have enjoyed learning “what works and what doesn’t” from these authors.6 The influence of policy analysis on this book is most obvious in my analysis of the size and shape of the American welfare state, the remarkable growth of some means-tested programs, and poverty and inequality (chapters 1, 2, 5, and 10).
What you will not usually learn from these scholars is who or what influences the direction of social policy. Behavioralists offer a number of important insights into the influence of public opinion and interest groups on social policy.7 I will explore the often surprising contours of public opinion in chapters 2, 6, 7, and 10, using evidence from national and international surveys; interest group behavior will be analyzed in chapter 7 as well. Behavioralists gravitate toward statistical analysis of numerical evidence and therefore gather information about lots of cases in order to be confident of their findings. In the literature cited above, the individual citizen is the main unit of analysis. Other scholars working within this tradition take advantage of the federal structure of American politics and compare the behavior of individual states, such as their spending patterns.8 My analysis of race and social policy in chapter 9 adopts this latter approach.
The behavioral approach has its limits. These studies tend to focus on a fairly narrow slice of time and on recent decades, and thereby have trouble supporting broad generalizations.9 Further, the influence of elected officials, policy experts, political parties, and institutional design—no matter what the era—is harder to quantify than the results of opinion polls. Scholars investigating these factors need to rely more on textual evidence from government documents, periodicals, biographies, and secondary sources, and this is exactly the point where more historically oriented scholars can add value. A large number of historians, political scientists, and political sociologists have analyzed processes of development and change in the American welfare state.10 For these scholars, history offers a storehouse of evidence for testing theories about social policy making, an approach I use in chapters 3, 4, 5, 7, and 10. History also offers a way of understanding how choices made in the past, even small ones, can shape politics for decades (i.e., path dependence). Chapter 8, concerning the history of workers’ compensation, provides a vivid illustration of this process.
I use history in conventional and unconventional ways. The usual way is to analyze long-run processes or events that happened long ago, and I do that in chapters 3 and 8. This approach can also illuminate important developments in recent decades, and I use history to identify patterns in the enactment, the expansion, and the retrenchment of social policies since the 1970s (chapters 4, 5, 7, and 10). Like Paul Pierson, I reject the “Faustian bargain” in political science whereby historically oriented scholars are valued as long as they stick to subjects that seem to have little contemporary relevance.11 Some of the most interesting trends in U.S. social policy are relatively recent.
The historical approach also has its limits. Sometimes the urge to generalize is so great that scholars overlook or downplay exceptions. On these occasions, it helps to think like a policy analyst and investigate social programs in more detail. As we will see in chapters 2 and 3, sweeping claims about the existence of a two-tiered welfare state and about two “big bangs” of welfare state formation do not hold up well under close scrutiny.
Drawing on these three traditions has implications for my research design. Some chapters entail quantitative analysis of many cases (though my use of statistics is never complicated). This design is good for detecting general patterns and identifying associations among different variables. Other chapters are more intensive investigations of one, two, or three cases, using mainly qualitative data. Careful comparisons over time or across a small number of cases can help identify interactions among important factors and the causal mechanisms at work in a given relationship. When drawing on the work of policy analysts and behavioralists, I tend to use numerical evidence from government documents and public opinion polls. The historical approach leads me to consult books and periodicals. The unit of analysis in each chapter may be nations, U.S. states, demographic groups, individuals, policy tools, or social programs. It all depends on the question. Staying firmly rooted in just one of these research traditions limits the kinds of questions we can ask as well as the quality of the answers we can offer.12
Return to Book Description
File created: 8/7/2007
Questions and comments to: email@example.com
Princeton University Press