Liberal political analysis is ordinarily based on a sharp distinction between domestic and international politics, and an assumption that domestic politics is the proper arena for democratic self-determination. But self-governing citizens have never exhausted the cast of characters who populate liberal states. Living alongside them there are often domestic aliens – permanent residents who are subject to the law, and may be protected by it, but who do not participate in making it. Refugees and remnants also inhabit liberal states. Whether citizens or not, they tend to bear the historical consciousness of victims or potential victims wherever they may live. A correlative fact is that in many now-liberal societies the meaning of citizenship itself is indelibly marked by the “missing” – the emigrant and the exile, the expelled and the extinct. Such identities – and the historical presence or absence of individuals who claim them – are generally regarded as messy details in the state-centered conceptual framework that dominates liberal political thought.

Not so long ago, nearly all African-Americans living in the United States were subject to a multitude of racial restrictions officially prescribed and enforced by state governments and their local subsidiaries. Most of the Jim Crow system dated from 1890–1910. By the middle of the twentieth century, this system was well established, so much so that many people assumed that it had always existed and that it expressed the timeless folkways of the South. However, in what strikes the historian as an astonishingly brief period during the 1950s and 1960s, the edifice was largely torn down. The puzzle is this: How could any institutional apparatus so deeply embedded, long-standing, and apparently strong be toppled so quickly? Although many scholars have discussed aspects of the puzzle, no one has offered a simple, clear, and compelling explanation. We aim to do so in this essay.

Only the United States supreme court justice has ever been impeached. In January 1804, the House of Representatives began a formal inquiry into the official conduct of Associate Justice Samuel Chase and approved eight articles of impeachment in November of that same year. The Senate held a trial of the justice in February 1805, which concluded with his acquittal on March 1. On the final article of impeachment, Chase escaped removal by four votes.

How do “ideas” help to account for political outcomes? The age-old question of the role of ideas has been given new life recently, partly in response to the imposing structure of theory that has been built around economic or rational choice explanations, with their emphasis on interests and material circumstances rather than ideas. Few outside of the most abstract economic model-builders hold that ideas have no impact on policies, but there is little agreement on how that impact occurs or on the observable criteria that would establish (or contradict) the impact of ideas.

A central paradox of the modern American presidency is that citizens regularly call for strong presidential leadership while at the same time their political culture predisposes them to be reluctant followers.1 One of the ways contemporary presidents resolve this paradox is by invoking an electoral mandate. By persuading others that he possesses a mandate from the voters to pursue a particular policy agenda, a president can disguise his leadership under the pretense of simply carrying out “the will of the people.” The presidential mandate thus enables presidents to lead while seeming to follow, to exercise power over people under the guise of empowering the people.

On 30 May 1914, Theodore Roosevelt fired the opening shots of the midterm elections against the party of Woodrow Wilson. Roosevelt framed the off-year elections as a referendum on the failures of the New Freedom, the Democrats' three-pronged program to curb the power of the trusts. Rather than bringing monopolies to heal, the former president asserted, Democratic policy had simply driven the economy into recession. “[T]he Democratic party,” Roosevelt explained on another occasion “has been engaged in what is fundamentally an effort to restore the unlimited competition of two generations back and to subject this to only an ineffective and weak government control”. To all, Roosevelt's counsel was constant: the prudent course of citizens that fall was to register a vote for social and industrial progress, to support the Progressive party candidate for Congress.

In what sense were the 1940s a turning point for the postwar development of the American state? This is an odd question since the decade is not usually considered a moment of lasting political transformation, as were the New Deal and the Great Society. Rather, it is more characteristically rendered as a period of political stalemate, memorably captured by Samuel Lubell's description of Harry Truman as “The Man Who Bought Time.” Yet the choices of the 1940s have been recently characterized by Ira Katznelson and Bruce Pietrykowski as far more significant than hitherto understood. Out of the conflict and debate over the New Deal emerge, they argue, a rather “crisp choice … about the character of the national state,” one that decided between alternative models of state-economy relationships.

In the late nineteenth century, public outdoor relief came under severe and sustained attack from reformers. Municipal reformers attacked it as a source of machine patronage and corruption, and charity reformers saw it as the cause of pauperism and moral turpitude among the poor. But in New York City, the critical decision to cut the municipal program came not from the reformers, but from the city's Democratic machine, Tammany Hall itself. In December 1876, the machine administration of Tammany Mayor William Wickham and Boss John Kelly terminated municipal outdoor relief funding for 1877, except for the distribution of coal. The previous “reform” administration had, by contrast, kept the program intact.

“Rebuilding the American State” was written in the manner of a bozzetto: it is a sketch drawn to reshape interlocking analytical and historiographical conversations and to suggest pathways joining the era of Roosevelt to the qualities and conundrums of postwar Democratic party liberalism. We underscored the key role of what might be called the long 1940s, stretching from the economic and political crisis faced by the New Deal in 1937–38 to the election in 1952 of the first Republican president since Hoover. We claimed that institutional and policy decisions taken across a number of domains in this period coherently recast the state and, in so doing, the contours and possibilities of American politics. We argued as well that old and new institutionalist approaches to state capacity have shared an unfortunate propensity to inventory organizational resources without regard to the normative and practical policy visions that define the content of what it is the state actually is meant to accomplish. In this light, simple dichotomous distinctions between weak and strong states appear as too blunt to sharply etch our understanding of the past half-century of American political development.

The development of the Earned Income Tax Credit (EITC) has been extraordinary. What started in 1975 as a minor amendment to a forget-table tax bill became in the 1980s one of the most popular programs in Washington. The national media and policymakers from both parties repeatedly praised the program. The New York Times, for example, hailed the EITC as “a wonderful but little-known anti-poverty weapon…the best means available for lifting the working poor out of poverty.” Republicans embraced the program as a superior alternative to welfare (meaning Aid to Families with Dependent Children), to a higher minimum wage, and to government-run day-care centers. President Bill Clinton made expanding the EITC central to his plans for reforming welfare, increasing the progressivity of the tax system, and improving the economic security of low-wage workers. One would be hard-pressed to find any government program that received such universal acclaim among political elites. By the early 1990s, it had become the policy equivalent of penicillin. Only in the spring of 1995, as congressional Republicans sought to balance the budget within seven years, did some policymakers seriously consider cutting this program.

In my essay I questioned two premises of Ira Katznelson's and Bruce Pietrykowski's interpretation of the failure of the Third New Deal: whether conflict over the New Deal during the 1940s can be understood as a debate over the developmental and fiscalist modes of state-economy relations; and whether an approach rooted in the analysis of bureaucratic missions and functions is a useful way to get at this question or, more generally, the broader issue of how to link state capacity with political choice. My own assessment of the evidence led me to reject their interpretation of the failure of the Third New Deal and to advance an alternative way of thinking about the question of state capacity and political choice. Now, in their very interesting rejoinder, Katznelson and Pietrykowski have raised some important issues that help clarify our differences and show what is at stake in this discussion. I should like to respond briefly to their comments and clarify my own views.

Long ago, Louis Hacker noted the anticapitalist bias in American historiography. The tendency of capitalism, particularly marked after the advent of industrialization, to concentrate wealth and to turn that wealth into political power suggests the problem. The Declaration of Independence with its twinned affirmation of equality and liberty provided the ideological underpinnings for national unity. Both seemed threatened during the long transition from an agricultural to an industrial economy as the rich got richer and more and more of America's free citizens were exposed to the unfreedom of the workplace. Progressive historians who systematically analyzed the impact of industrial capitalism in the United States wrote out of sympathy with labor in its struggle for recognition, and no doubt Hacker was thinking of this tendency when he made his remark. This anticapitalist bias in the writing of American history has continued, only now historians decry the steady incursion of commerce and industry out of deference to a traditional way of life.