The Divided Welfare State

The Battle over Public and Private Social Benefits in the United States
Published:   September 2002
ISBN: 0521013283 | 464 pages

"Hacker's study is an important resource for understanding why enactment of universal health insurance has been so difficult. His analysis is original and insightful. Politicians and political scientists alike will find this volume useful and thought provoking."
Edward M. Kennedy, U.S. Senator From Massachusetts

"A brilliant book that will force American left, right, and center to reconsider their ideas-and prejudices-about how our country provides social insurance and social services. An engaging and rigorous scholar, Hacker sheds light on a fact so obvious that scholars have often ignored it: In the United States, much of our 'welfare state' is actually a system of private sector support. This is an indispensable work of history, political analysis, and sophisticated common sense."
E.J. Dionne Jr., Brookings Institution

"A work of extraordinary historical sweep and analytic power, The Divided Welfare State offers a provocative new angle of vision on the politics of social policy in the United States."
Theda Skocpol, Director Of The Center For American Political Studies, Harvard University

Purchase this Book

Selected reviews of The Divided Welfare State are featured below:

The New Republic

Monday, October 14, 2002
The passing of the welfare state is a subject in dire need of serious thinking, and two impressive books devote themselves to the task. Jacob S. Hacker, a young political scientist, has turned his doctoral dissertation into an ambitious theoretical enterprise designed to explain why the American welfare state developed the way it did. Neil Gilbert, a professor of social welfare at Berkeley, has written a shorter and less theoretical book than Hacker's. Each tries to explain the significance of a development that seems to fly in the face of the welfare state's rationale: the growing reliance on private provisions for what were once understood as public functions, such as the use of tax credits or vouchers that permit individuals to choose from various service providers or the efforts to mandate business firms to provide pensions and health insurance.

Private provision characterized the American welfare state from the beginning, as Hacker's account demonstrates. Some private companies, such as the freight forwarder called American Express, first offered pensions to their employees as early as 1875, but the relative paucity of such efforts -- less than ten percent of the American workforce was covered by a private pension throughout the 1930s -- gave the Roosevelt administration its chance to pass comprehensive legislation. Typical of the politics of the New Deal was the debate over the Clark Amendment, named after the conservative senator Champ Clark of Missouri. Under its terms, any firm that provided pensions to its employees could "opt out" of Social Security, thereby undermining the program's universality. Roosevelt was opposed, and the amendment was defeated. Business calls many of the shots in American politics, but it did not call this one.

Still, Social Security never became a replacement for private pensions. The program was limited to those who paid taxes into the system. It was defended as an insurance program rather than as a benefit. One law passed by Congress in 1939 removed private pensions from wages, thereby excluding them from the Social Security tax, while another in 1942 enabled companies to reward their betterpaid employees by supplementing their governmental pensions with private ones. "By 1960," Hacker writes, "a two-track system of organized retirement protection had been firmly institutionalized in the United States." For roughly half the workforce, private pension programs developed by employers gave employees the prospect of a modicum of economic security in their old age. For the other half, Social Security offered some protection, but generally not enough to live on. Later policies, such as IRAs and 401(k)s, which were not seen as major innovations as the time of their passage, contributed to a situation in which better-off workers found themselves with more commodious retirement prospects than the less well-off.

Health insurance developed along similar lines in the United States, except that there never was a commitment to public provision as secure as the one represented by Social Security. As if they had learned a lesson from their defeats by Roosevelt during the New Deal, opponents of a national system of health insurance recognized early that private provision would undermine the appeal of publicly funded efforts. Demands for national health insurance came during the Truman administration, roughly ten years after Social Security had already passed, and by that time two-thirds of America's largest firms already had health plans (compared to the four percent who had company-sponsored pensions in 1935). The political effects of this relatively widespread coverage could be seen in the Revenue Act of 1954, which exempted private health care benefits from federal taxes. The Congress of Industrial Organizations, which ordinarily would be expected to line up behind a system of national health insurance, realized that the interests of its members would be best served by protecting private programs and supported the tax exemption. Without a significant push from the left, the way was clear for the American Medical Association to attack national health insurance programs as socialized medicine, which it was only too happy to do. As Hacker points out, indirect public support for private programs, such as exempting them from taxation, turned out to be an expensive proposition for the government to undertake. But no one projected the costs of, or indeed developed a rationale for, this rather odd way of financing the welfare state.

If national health insurance failed to become a reality in the 1940s, when elements of the New Deal coalition were still alive, it was even more unlikely to win support as the political impact of the Great Depression and World War II receded. Reformers hoped that a smaller program such as Medicare might lead the way toward a more comprehensive system of national health insurance, but that never occurred. Instead Medicare created a policy hodgepodge: physicians were given a blank check to charge the government, inflating costs; drugs (an important component of care for the elderly) were left out of the program; and interest groups such as the AARP developed stakes in the policy that complicated further reform efforts. If nothing else, the failed Clinton efforts revealed how truly complicated the system of medical insurance in the United States had become. Rather than representing a philosophical commitment to the principle that no one should be victimized by misfortune, health insurance policy in the United States, like pension policy, provided fairly good coverage for those already blessed in life and bad to non-existent coverage for those hardest hit.

As European and other non-American welfare states lose their innocence and begin to rely more on private means, America no longer looks like a laggard. Hacker calculates that when private benefits are added to public ones, there are no substantial differences between the United States and its close cousins in the amount spent on welfare; his figures show that the United States ranks higher than Ireland, Canada, and Australia when both kinds of benefits are considered and just below Holland and Denmark. This is not to say that differences between welfare states have been abolished, for public provision still retains elements of solidarity and equality that private provision does not. Still, as Gilbert points out, the differences between welfare states increasingly resemble what Freud laconically called the narcissism of small differences. Communism was overthrown quite dramatically in 1989. Less dramatically, but more important, social democracy crumbled around the same time. We are unlikely ever to go back to the untrammeled laissez-faire of Dickens's England; but we are also not likely to be hearing arguments to the effect that the correct public policies can usher in a new era of human dignity.

Hacker in particular, but also Gilbert in his way, wants to do more than describe the welfare state's troubles. Each seeks an explanation. Hacker finds his explanation in history. A public program such as Social Security creates a constituency -- I mean the elderly -- which then develops an interest in protecting the policy. The same is true of private actions, especially when they receive indirect support through tax advantages or subsidies: advocates for private pensions secured enough of a foothold in the system created during the New Deal to prevent any significant expansion of public pensions in the years that followed. In terms that have become increasingly popular in political science, policy creates politics and not just the other way around. Whenever government acts in the present, it influences how government will act in the future. As Hacker puts it, "By pushing policy development down a particular historical path, a policy passed at time T1 may significantly change the range of possible options at time T2."

All this sounds reasonable enough. Indeed, it has for years been the conventional wisdom of historians, who frequently chide social scientists for their lack of interest in the past. But Hacker is not content simply to leave his readers with the lesson that history matters. Instead he proposes that the story of the welfare state illustrates the power of what he, following other social scientists, calls "path dependence." According to this way of thinking, those who develop policy rarely if ever have blank slates. Just as Marxists argue that individual actions are constrained by structures, theorists of path dependence emphasize the degree to which our choices are limited by the choices made by those who came before us.

The great advantage of path dependence, in Hacker's view, is that it forecloses functionalist explanations that view "policy configurations" as "optimally tailored to current conditions by political agents." For too long, he believes, debates over the welfare state have been dominated by discussions of whether such programs serve the long-term interests of the capitalists or represent real progress on behalf of workers. From the standpoint of path dependence, they do neither. Path dependence gets one away from thinking teleologically in favor of thinking geologically. One comes to appreciate how contingent our institutions really are. Phenomena such as the welfare state are characterized, in an almost evolutionary manner, by complex traits fashioned at different times and in different ways in the course of their development.

Its advocates promise that the theory of path dependence offers an approach to the study of policy that explains more than we would know if we examined only contemporary events or treated history as a just-so story. Alas, this promise remains unfulfilled, at least in Hacker's book.

Although the private side of the welfare state has been studied by many (including Gilbert, whom Hacker cites), it is important to recognize that decisions not taken can be as important to the development of policy as decisions taken. Hacker is particularly good at explaining how non-decisions and decisions interact. His discussion of one particular example, the Employee Retirement Income Security Act (ERISA) of 1974, is exemplary in this regard. Designed to promote equality by enabling people to take pensions with them from one job to another, ERISA was ultimately unable to overcome the dual system of pensions created by America's mixed system and wound up as a brief and not very successful interlude in an otherwise conservative story. The problem is that path dependence is not required to make this case; all that is required is intelligence. The strengths of Hacker's approach stem from the exhaustive historical research that he undertook, his mastery of detail, and his clarity of presentation, not from his theoretical asides. If all the talk of path dependence in The Divided Welfare State had been deleted by a ruthless editor, the book would have been just as valuable and considerably shorter. And it would have been better written, since the more theoretical Hacker tries to be, the more turgid his writing tends to become.

Theory ought to help us to understand the present and to predict the future, but when it comes to contemporary debates about Social Security and health insurance and where they might lead, path dependence is worse than useless. Will President Bush succeed in his efforts to privatize Social Security? Hacker, rightly, cannot leave the issue alone, for anyone concerned with the future of the welfare state ought to address the plans for Social Security floating around Washington these days. Yet finally the theory of path dependence gives him little or nothing to say. On the one hand, he reminds us, our previous reliance on private benefits makes it understandable that Bush would propose such a course of privatization; if anything, the dependence of the present on the past suggests that Bush's plans ought to succeed. But, Hacker goes on to note, Social Security as we know it has its defenders and has proven itself extremely resistant to change, so maybe they will fail. The conclusion offered by Hacker's efforts to apply systematic theorizing to such a burning issue, then, is that the United States may go in one direction or it may go in another. To offer his readers more than platitudes on this question, Hacker would have been better off talking about such real-world events as the stock market decline, the failure of Republicans to rally behind Bush's plan, and the decision of the Bush administration to shy away from a term like "privatization" that it once endorsed. Only then does it become possible to conclude that Social Security, for the foreseeable future, is safe.

Hacker recognizes that the theory of path dependence is not a deterministic theory claiming that certain courses of action are inevitable. That is why he adds to the concept of path dependence the notion of "critical junctures." These are the moments when new policy avenues open up, thereby starting new paths. The present period certainly seems to qualify as such a critical juncture, for if European and American welfare states are increasingly resembling each other, then both have departed from the paths upon which they embarked in the 1930s and 1940s.

Unfortunately for Hacker, path dependence cannot explain why critical junctures -- the rest of us call them changes -- happen. The theory, in fact, seems doubly wrongheaded. Not only does it fail to grasp the realities of the American situation, it also concentrates our attention in the wrong direction as we try to understand Europe. Welfare states created by broadly based political coalitions built on middle-class voters around which powerful interests have been formed ought to be resistant to any efforts to cut back benefits; but the Scandinavian welfare states, which should be exemplary models of path dependence, are changing radically. Gilbert quotes one observer who describes the Danish version of those changes as a "paradigm shift" in social policy. Of course one can account for changes of this magnitude by adding to path dependency the idea of a critical juncture. But why add to one theoretical abstraction another that amounts to its exact opposite? And why begin with the theory that has the least relevance to the situation at hand and move backward from there? As one might expect from an approach that refers to the past as "T1" and to the present as "T2," path dependence offers the illusion of theoretical elegance rather than the reality of it.

If theory -- if that is the right term to describe such fashionable approaches in the social sciences as path dependence and rational choice -- does little to serve reality, it evidently does a great deal to serve other needs, particularly academic self-importance. Jacob Hacker is that rare academic whose doctoral dissertation has been transformed into his second book. His first book, The Road to Nowhere: The Genesis of President Clinton's Plan for Health Security, appeared in 1997, and it was widely and rightly praised for the quality of its analysis of the Clinton debacle. Hacker interviewed many of the key players involved and consulted huge numbers of documents that had been released under a court agreement. Lacking ideological baggage and methodological apologetics, the book established Hacker as a rising star in political science. I recall thinking how rare it was for a book by someone so young to demonstrate so much poise and self-assurance.

The Divided Welfare State is even more thoroughly researched than The Road to Nowhere, and it expands the author's scope from health care to pensions, but in many ways it is not as good a book as its predecessor. It introduces pomposity into Hacker's work. Not content to allow readers to draw their own conclusions, he proclaims what he believes to be the "key insight" as well as the "central goal" of his new book. (The notion is that path dependence matters as much in the story of private benefits as it does in the history of public ones, and the goal is "to specify exactly what it means to say that political processes are path dependent and to clarify the essential relationship between path dependence and arguments about timing and sequence.") He insists that his method is "comparatively informed" even if he rarely discusses non-American examples. He even praises himself for "the extensive sweep of my analysis," just in case the reader missed it. All this seems to be part of a trend these days among doctoral dissertations that are turned into books published by university presses. One must at all costs say something original, and evidently the best way to convince readers that one has done so is simply to announce the fact.

Like other doctoral dissertations turned into books, moreover, Hacker's is stronger in the middle than it is in its introduction and its conclusion. Compared to the days of my own doctoral work in the 1960s -- we were generally a lazy bunch -- younger scholars now typically engage in exhaustive research and frequently uncover fascinating empirical material. But they are also encouraged -- compelled, really -- to squeeze their material into whatever theoretical approach happens to be in fashion at the time. The theoretical chapters of dissertations turned into books inevitably tend to be worthless, however much they impress thesis committees and search committees. Since his research is strong, Hacker's book perversely illustrates the flaws of graduate training in the social sciences, for it provides such a strong contrast between substance (what Hacker has learned on his own) and form (what Hacker has to do to persuade readers that he made some theoretical breakthrough). Reading Hacker defend path dependence is like listening to a trial lawyer making the best case he can for a defendant whose innocence he doubts.

Academic convention insists not only that one must have a theory, but that it must also be a relatively fresh one. For a while, rational choice theory was all the rage; but as others have noted, rational choice theory never seemed to explain much of the world outside the academy, though its focus on the pursuit of self-interest did seem rather appropriate to understanding academic careers. The application to the fad of path dependence is obvious. Once large numbers of scholars invested in rational choice theory, the point of diminishing returns set in, and it was only a matter of time before a new theory began to beckon. "Historical institutionalism" became the preferred venue for many scholars who believed that there was more to human beings than self-interest, and the consequence of that turn was the publication of a number of important books that showed how contemporary political institutions were shaped by historical forces. Then along came path dependence as a way in which historically trained scholars could prove that they were as theoretical as the next guy, which in political science departments typically means the rational choice types. And so, although historical approaches began as an alternative to rational choice, they quickly started to mimic its economic language and its grandiose claims. -- By Alan Wolfe

Related Programs