Louise Tilly often quoted Marx's famous dictum from The Eighteenth Brumaire of Louis Bonaparte that “men (and women) make their own history, but they do not make it as they please, they do not make it under self-selected circumstances, but under circumstances existing already, given and transmitted from the past” (Marx Reference Marx1935: 13). Louise believed agency should be understood in that way. What we learn as we study how people cope with different social circumstances is how strategies shift over time. No one moment will produce great transformations in behavior or mind-set, but an accumulation of shifting strategies for handling the daily realities of life can result in important social/historical changes.

To understand how individuals build new structures even as they are constrained by existing ones, Tilly's work with Joan Scott on women, work, and family in England and France across two centuries made use of detailed analysis of local areas along with a focus on large-scale trends in the two countries. Working out large-scale processes through close attention to microlevel analysis remained central to Tilly's approach to social history. Toward the end of her career, when she turned to world history, she offered her vision of how scholars could remain committed to human agency even as they attended to large-scale transnational processes.

In her 1994 address as president of the American Historical Association (AHA), Louise paid tribute to the recently deceased social historian Edward Thompson for his attention to the local-level context and for his understanding of the how local-level events were connected to “far reaching structural transitions [such as the movement] from market to industrial capitalism” (Tilly Reference Tilly1994: 1). Addressing the crisis surrounding social history by the mid-1980s, she acknowledged that local case studies had allowed scholars to detail the lives of ordinary men and women, but were difficult to use in answering larger historical questions, in particular, questions about processes and structures. For Tilly, the task was not to abandon the specific, for it is at the level of the specific that we can study agency, that is, “the connections between and among structures, processes of change and human action” (Reference Tillyibid.: 2). Moreover, because social historians encounter stubborn empirical details that cannot be compared simply to cases from other settings or time periods . . . narrative is therefore central in describing and interpreting temporal sequences of events” (Reference Tillyibid.) Nonetheless, Louise believed that one could and should do history that addresses the relationship between the specific case and large-scale trends by studying the “spatial links between groups, regions, political units connected by trade, production, migration, religion or political relations” (Reference Tillyibid.: 3). Beyond the connection between human action and structure, and the spatial links between processes that cross countries, the historian can make temporal connections by paying attention to the ways in which “the past constrains or facilitates the present action” (Reference Tillyibid.).

To illustrate these three kinds of connections, Tilly focused in 1994 on the effects of early industrialization on spinners and weavers in India, England, and France. She looked at the spatial connections between the processes in each country, as well as the ways that structure and action operated through the power balance and bargaining situation in both labor markets and families in the three countries (Reference Tillyibid.: 20). Finally, she showed how past structures influenced subsequent developments, for example, pointing to India's highly specialized labor system and the traditional patriarchal control of children's lives as important to understanding Indian responses to English technological innovations (Reference Tillyibid.: 6, 10).

Tilly's analysis of industrialization and the textile industry in England, France, and India barely touched on politics, but her other work on politics and collective action stressed similar kinds of connections between structures and action. Thus, in her overview of American women's politics, coauthored with Patricia Gurin in 1990, she emphasized that changes in woman-centered ideology in the last two centuries could be tied to “demographic, economic and political transformations [such as] the integration of women into political structures” (Tilly and Gurin Reference Tilly, Gurin, Tilly and Gurin1990: 23). Just as she argued in her AHA address, she cautioned that finding similar patterns in women's political history in the nineteenth and early twentieth century does not mean that history repeats itself, for each historical moment is unique. Nevertheless, she concluded, the past shapes and limits the present in politics as well, by “structuring the circumstances that offer possibilities to historical actors at any moment” (Reference Tilly, Gurin, Tilly and Gurin1990: 30).

As my work turned from a focus on women and the family to a comparative study of the welfare state, Tilly's perspective regarding structure and action, her emphasis on the ways that transnational processes both affect the local and the national and are remade by human actors, remained influential. For our project analyzing the development of the welfare state in England, France, and the United States from 1870 to 1950, my coauthor, Michael Hanagan and I pay attention to transnational processes such as industrialization and migration as they affect circumstances nationally and locally. Differing political programs are appreciated as strategies implemented by actors who are very much constrained by the social and institutional realities of their communities, but who also must be flexible and make pragmatic choices as they create and recreate political outcomes.Footnote ¹

This perspective is illustrated by focusing on one aspect of state initiatives in the three countries at one moment of crisis, the Great Depression, concentrating on the United States, where the Depression hit first and hardest. The development of Western welfare states is a vast subject and by now there are a number of important comparative analyses. In order to deal with a long time span we have selected some key issues—income supports over the family cycle, the eradication of sweated labor, and the spread of state-sponsored schooling. To get at the interaction between human agents and structures, we have selected a small handful of urban locations within each country—Paris and St. Etienne in France, London and Birmingham in England, and New York and Pittsburgh in the United States—and some critical time periods on which to focus. We do comparative history in order to get at national-level patterns but we focus on the narrative in each country; like Tilly, we emphasize that social policy is not simply the working out of long-term and enduring features of social political life, but also is shaped by the particular hour and the special circumstance.

The Second Industrial Revolution, which marked the development of mass industrial production and the modern business institutions that supported it, provides the long-term transnational context for our study. We see the formation of the national welfare state and the characteristics of political coalitions as influenced by two main processes, demography and democratization, broadly defined. Demography refers to trends in fertility, mortality, and migration in each country, all areas of focus for Louise (Gillen et al. Reference Gillen, Levine, Tilly, Gillen, Levine and Tilly1992: 1–12; Tilly Reference Tilly1994) and for her students such as Leslie Moch (Moch Reference Moch1992; Hoerder and Moch Reference Hoerder and Moch1996) and Donna Gabaccia (Reference Gabbacia2000). These characteristics influenced the nature of the industrial problem, the definition of the solution, as well as the composition of reform coalitions. As representative governments, the welfare states in the three countries were shaped by political coalitions that came together to push for government legislation. Not all efforts were successful, however, either in terms of their ability to enact legislation or to ensure that governments enforced the law. Democratization includes the differences in the pace of enfranchisement among large sections of the population, the balance of power between labor and capital, and the relative weight of elitist state structures. The democratization of state structures and the democratic mobilization of newly enfranchised masses in mass political parties, which differed in the three countries, shaped the character of political coalitions, what problems would be addressed, and what solutions tried. Finally, broad coalitions of workers and middle-class reformers succeeded when they could tap into widely held beliefs about the legitimate role of government.

How does our emphasis on the material—the demographics and the politics—help to explain the differing approaches to unemployment policy during the Great Depression? Historians have noted that among democracies, the most robust response to the Depression occurred in the United States (Rodgers Reference Rodgers1998: ch. 3). When Franklin Delano Roosevelt took office in 1933 unemployment stood at 25 percent. In addition to implementing a national economic recovery plan that allowed major American industries to collude on the matter of prices, production levels, and wages, Roosevelt moved to address the unemployment crisis through massive aid to localities for direct relief and a variety of work relief initiatives. Indeed, in historian Scott Smith's words, the public works programs were “an extraordinary successful example of state sponsored economic development” (Reference Scott Smith2006: 19). In 1935, the New Deal administration and their allies in Congress instituted a national plan of unemployment insurance in the Social Security Act of 1935 involving state participation in insurance plans for categories of workers.

Why did this happen? In part, the decentralized nature of relief that had long characterized the US approach to welfare created pressures from the grass roots up to the national state. Focusing on New York and Pittsburgh, we can see that local and state governments, in combination with private charities, could no longer handle the crisis and thus were clamoring for federal aid. But the expansion of American democracy was especially important. As a result of the economic crisis, newly mobilized social groups could exert political influence in the next five years, at the same time that the business community was on the defensive. The labor movement, in the doldrums since the 1920s, became active on behalf of the unemployed. At the local level, street demonstrations, organized both by local labor councils and leftist political parties, put pressure on local and state governments (Goldfield Reference Goldfield1989; Kelly Reference Kelly1990; Naison Reference Naison1983; Rosenzweig Reference Rosenzweig and Green1983:169, 171–72). From 1933 to 1937, through massive organizing and important strikes in various parts of the country mass production unionism created a place for itself in the American polity (Bernstein, Reference Bernstein1970 [1969]; Dubofsky Reference Dubofsky1994; Fraser Reference Fraser, Fraser and Gerstle1989; Jenkins and Brents Reference Jenkins and Brents1989; cf. Gordon Reference Gordon1994). From the early days of the New Deal, organizations active in the unemployment campaigns along with the American Federation of Labor, especially unions in the building trades, supported the implementation of massive works programs (Dorrance Reference Dorrance2010: 82; Scott Smith Reference Scott Smith2006: ch. 2).

This is a story of the increasing political power of America's white working class in the 1930s, and requires an appreciation we share with Tilly of how a large-scale process can transform politics. In this case, we focus on the changing demographic characteristics of the American labor force. So many of America's new working class in the early part of the century were immigrants without voting rights. Since the 1920s, national policy virtually cut off immigration from Europe; as the unemployment crisis deepened in the 1930s, a large portion of working class, particularly the unionized working class, were either naturalized citizens or children of immigrants. Citizenship rights became an increasingly important determinant of difference among America's working class. For example, the large influx of Mexicans in the 1920s declined not only because of a lack of work; policies of deportation and repatriation resulted in the return of almost 20 percent of the Mexican population in the early days of the Depression. As Mai Ngai shows, the policies of deportation and exclusion during the 1930s worked mostly to the disadvantage of nonwhite immigrants, most especially Mexicans (Ngai Reference Ngai2004: 75, 81–83, 135).

By contrast, the newly emerging group of American citizens among the European immigrants became participants in fashioning the American welfare state during the 1930s. Sidney Hillman, and others in Roosevelt's labor coalition who cooperated with New Deal administrators, could influence both the implementation of public works programs and the construction of unemployment policy in the Social Security Act of 1935 because they now represented a new political power, the votes of many immigrants, especially of European background, and their children, who were now voting American citizens (Lubell Reference Lubell1952).

New Deal policies thus exemplify increasing democracy but also the limits of that citizenship, as Quadagno (Reference Quadagno1998), Katznelson (Reference Katznelson2005), and others have shown. Most African Americans were Southerners and were not empowered to vote, but even enfranchised Americans (including many women, minorities, and poor whites), not fully organized to exert political power, were left out of the unemployment insurance guarantees and had less access to work relief projects.

Finally, American governments tightened the labor market by expanding schooling. As Tilly noted in 1994, groups produce structures and in turn are affected by such structures as labor markets within “cultural contexts of shared understandings” (Reference Dubofsky1994: 2). Because schooling was traditionally viewed as a legitimate province of government resources, many states localities and the national government put money into this expansion, even during this time of economic crisis. In the United States large coalitions in favor of expanded state schooling had long facilitated the American tendency to put faith in education as the answer to American social problems; we can see the pull of this tradition during the Depression (Cohen Reference Cohen2005). On the national level, government built schools through its various work relief programs—indeed, almost one-half of the counties in the United States could boast at least one school built by the New Deal—and also provided money through the National Youth Administration to keep adolescents in school (Scott Smith Reference Scott Smith2006: 87).

If natives and newcomers alike shared the American crisis, the French Depression at first appeared to be a problem borne mostly by migrants. Understanding why requires us to look again to how past strategies regarding labor markets shape the present. While the post–World War I years saw the United States restricting migration, as Clifford Rosenberg (Reference Rosenberg2006) has pointed out, Paris was the leading capital of immigration in the Western world. Long concerned about depopulation and now almost panicked in the wake of the World War I debacle, the French in the 1920s tried a variety of measures to increase their labor supply, such as recruiting workers from their empire, liberalizing citizenship laws, secretly obtaining foreign orphans, criminalizing abortions and birth control, and devising a guest worker program that would be widely imitated by Europe in the immediate post–World War II years. As the Depression deepened, Paris expelled large numbers of workers recruited in good times (Bavarez Reference Bavarez1991).

In addressing the unemployment crisis, shared understandings about the rights of French citizenship brought together the nationalist right and even elements of the republican socialist left. Except in the case of political refugees, the Popular Front that came to power in 1936 brought little relief for emigrant workers. Even the Communist Party's (CP) commitment to the migrants’ cause based on proletarian internationalism wavered. For example, in the Longwy steel basin, home of many Italian Communists who had fled Mussolini, the local CP in 1937 held a rally around the theme of France for the French (Noiriel Reference Noiriel1984: 306–9).

Ultimately, while much of France's unemployment policy was put on the backs of migrants who were encouraged to leave, those French citizens who were unemployed also suffered because of a lack of systematic unemployment insurance. The inattention to the unemployment crisis exemplifies how the institutionalization of one approach to the problem of labor markets can constrain efforts to adapt to changing conditions. In France, wide coalitions in favor of encouraging population growth had for decades resulted in important advances in the area of family allowances. By the 1920s, French employers were using payments to workers with children as a means of attracting and keeping their workforce. Beginning at the local level, payments to French families by individual employers and networks of industrialists had spread like wildfire in heavy manufacturing (Huss Reference Huss, Wall and Winter1988).Footnote ² By the time the Depression hit, the structure of family allowances conditioned the government to approach the unemployment crisis by expanding this benefit. At a time when labor surplus actually dominated the French economy, when emigrants were expelled in mass, and when French workers stood in soup lines, French politicians remained, as ever, preoccupied with falling fertility. The greatest contribution to the French welfare state in the 1930s was large-scale intervention by the French government in the development of family allowances, with repressive consequences for women, who still did not have the right to vote.

The Popular Front government had fashioned a united program that included generous references to social reform, but the true basis of the coalition for the CP and socialist leaders was opposition to fascism both in France and internationally rather than social legislation. Certainly the increase in labor's power affected the course of state building; the major social achievements during the Depression came as a result of a strike wave in 1936, which ended in an accord promising a forty-hour week, paid vacation, and compulsory arbitration, but these conquests were soon rescinded.

If the French, long preoccupied with scarce labor, turned to the transnational labor market, so too did the English, who for decades were contending with a surplus labor market. English politicians, social theorists, and many ordinary citizens considered emigration to the Dominions and the colonies a very viable solution to the problem of unemployment. Once again, we see the long-term influence of one approach to the problem of labor markets; the English push for emigration continued even when changing colonial labor requirements and the onset of agricultural depression in the 1920s greatly lowered the Dominion's need for labor. As the Great Depression set in, English politicians persisted in this fruitless effort to send laborers abroad.

When the Depression hit England in the early 1930s, the ensuing political crisis left Labor very weak. England already had unemployment insurance for some, and an outright payment program for many more, but with Conservatives basically in power throughout the decade, unemployment policies took a more punitive attitude toward the long-term unemployed. Massive local protests forced some revision in benefits, but the basic outlines of the system remained in the Depression and the ruthless standards used to enforce eligibility; stories of its cruelty and inanity were to become a staple of British working-class culture (Chamberlain Reference Chamberlain1989: 97). During the years of Conservative hegemony, innovative thinkers who championed increased government spending, including massive public works, and aggressive fiscal policies were ignored or shunted aside by the ruling Conservative party. But the leadership of Labour also resisted innovation, rejecting calls for stimulus spending in the 1930s.

While Tilly acknowledged the importance of cultural contexts and shared understandings, her own work, ultimately, did not focus on how values shaped behavior. In attending to the importance of historical traditions and the beliefs of political actors to understand both the nature of broad-based social movements and their success, our comparative study of the welfare state is influenced by scholars of political culture who emphasize how attitudes about the scope of government, and gender or race or class norms influenced the political responses in the three countries (Jacobs and Zelizer Reference Jacobs, Zelizer, Jacobs, Novak and Zelizer2003: 7–8). But we emphasize that while cultural values are deeply rooted and can affect political developments, their meanings are not fixed. Social movements, political scientist Sidney Tarrow has argued, are not merely “symbolic message centers” about the culture. Their meanings are constructed in the context of specific interactions as activists try to make use of, but also adapt, their political rhetoric in light of the power structure, the forces of opposition, as well as the values of their supporters (Tarrow Reference Tarrow1998: 107). In analyzing how social movement activists as well as politicians were strategic, framing their arguments to tap into the cultural understandings of audiences, which might vary at any one moment and often change over time, we continue to share Tilly's commitment to both strategy and agency. For us, Louise Tilly's perspective is still relevant.