Ideas do matter. These days the claim is nothing new, even less so in the field of welfare state research. Many studies have examined how cognitive and normative ideas shape welfare state dynamics (Cox Reference Cox2001; Steensland Reference Steensland2006; Schmidt Reference Schmidt2002). Recently many authors have concentrated on formulating a common conceptual framework for idea-oriented policy research, including constructivist institutionalism (Hay Reference Hay, Rhodes, Binder and Rockman2006) or discursive institutionalism (Schmidt Reference Schmidt2008). These comprehensive efforts have drawn idea-oriented policy analysis away from its earlier focus on defining ideas towards pointing out more accurately how ideas interact with other explanatory factors such as interests and institutions. It is the link between ideas and interests that stands in the focus of this paper.
The scope of idea-oriented policy research ranges from analyses that ask which economic and political factors explain the rise of specific ideas (ideas as explanandum) to approaches that give explanatory primacy to ideas for many forms of political and economic action (ideas as explanans). In parallel to this division idea-oriented research tackling the question of interests is divided between interests as ultimate or as immediate cause of action (Braun Reference Braun, Braun and Busch1999). Some studies show how interest coalitions are forged that support certain ideas and/or try to raise discursive legitimacy for their positions. For example, King (Reference King1995) has shown how ideas have helped to formulate work fare policies in the UK and in the US during the 1980s. Links between the Reagan and Thatcher governments and conservative think tanks have fostered long-term interest coalitions for reform. Moreover, many studies show how interests drive institutional change along the lines set by policy paradigms consisting of specific problem definitions and goals (Hall Reference Hall1993; Steinmetz Reference Steinmetz1993). In this perspective ideas are “focal points” or common points of reference for the forging of coalitions in situations of crisis (Ringe Reference Ringe2005; Sabatier Reference Sabatier, Sabatier and Jenkins-Smith1993). Still, interests are the ultimate forces that drive groups towards conflict while ideas help to discover shared objectives and converge discursively.
There is a second strand of literature in which ideas play a more substantive role. Here ideas are not seen as additional factors, competing with interests and institutions, but rather as a substantive element of interest formation. They help political or economic actors to make sense of the institutional and material conditions in regard to their basic needs. For example, Cox (Reference Cox2001) describes the importance of the discursive construction of reform imperatives for welfare state retrenchment policies that influenced preference formation among crucial political actors. Steensland (Reference Steensland2006) shows how the cognitive distinction between “deserving and undeserving poor” has influenced guaranteed income policies in the US. It is the close “interlock of ideas and patterns of practice” (1285) that gives ideas a prominent role for the formulation of interests and strategies. Interests in this sense reflect actors’ sense-making and goal-defining efforts in concrete economic and political situations that are potentially uncertain and open to different interpretations.
This article tries to forward this debate both conceptually and empirically: first, an analytical framework will be provided that tries to bridge the two described models of linking ideas and interests. While basically following the argument that ideas are an integral part of the situational formation of interests, it is argued here that the other perspective in which interests are ultimate driving forces points to a very important aspect that tends to be lost in radically constructivist perspective: Interests are never completely subject to situational interpretation but are indexed by the objective socio-economic position actors have within the distribution of power and resources. Therefore interests should be defined as the situational concretization of subjective goals actors build in regard to the basic impulse to defend (or improve) their objective social position. This is a three-dimensional view of interest that focuses on the need to triangulate between social position, knowledge and situational context.
Second, I will show empirically that this three-dimensional idea-oriented concept of interest can help to promote one of the most intense academic debates in welfare state research: The emergence of the American welfare state in the Social Security Act of 1935. The role of business interests in the emergence of the American welfare state has been subject to extensive debates (Hacker and Pierson Reference Hacker and Pierson2004; Hacker and Pierson Reference Hacker and Pierson2002; Swenson Reference Swenson2004). On one side of the debate, proponents of the welfare capitalism approach have stressed the positive interests that employers had in industry- and nation-wide social policy. On the other side of the debate, proponents of a more power-oriented approach have claimed that business supporters were not positively interested in the benefits of this new labor market policy, but they were in a position of historical political weakness to block it. For strategic purposes some American employers chose voice over exit and took part in the process to prevent worse.
I will show that by changing the concept of interest this debate could be re-directed towards a common explanatory problem shared by both sides: A situational concretization of interests, be it substantial or strategic, has to be oriented towards ideas that are plausible and legitimate within a public or managerial discourse. If economic positions substantially erode in a situation of crisis, business leaders have to direct their interests towards their expectations about the consequences of possible policy choices. The same is true if business leaders want nothing more than to strategically prevent the worse. They need a justification (for themselves as well as for potential followers) of why the second-best solution is better than any other options. In both cases business leaders depend to a certain degree on plausible and legitimate ideas on which they can base alternative interest definitions. If we investigate which ideas American employers found plausible and legitimate during the 1920s, we can understand how it was possible during the Great Depression for unemployment insurance to change from a threat to business profitability to a safeguard of economic stability. We will see why both ideational change and the crisis have to be seen in interaction to understand why some employers in the early 1930s turned to unemployment insurance instead of other institutional means.
The first section develops a theoretical framework for linking ideas and interests. The second section presents the debate about the role of business interests for the emergence of the Social Security Act (SSA) in more detail. The third section traces the evolution of business interests in US labor market policy from 1911 to 1935. This study is based on original archival work that demonstrates how ideas traveled between social reformers, politicians and employers. It builds on conference proceedings, congressional hearings and political pamphlets, which are presented here in tandem with socioeconomic data and institutional context variables.
The Role of Ideas in the Formation of Interests
In this section, I provide a conceptual framework which links ideas to the process of interest formation. It is based on crossing two important conceptual dimensions of “interest”: First, interests can be conceptualized as objective or subjective action orientations. Interests either contain abstract, ultimate goals that reflect objective structural positions, or they are concrete subjective goals in specific situations (Swedberg Reference Swedberg2005, p. 380; Dahrendorf [Reference Dahrendorf1957] 1959, pp. 180ff.). For example, while profit maximization provides a stable basic orientation for capitalist employers, the question of which labor market policy will safeguard or endanger this general interest will have to be answered by interpretation of the context. Second, interests can be material or ideal. Objective interests reflect social positions, but these positions do not necessarily have to be defined on material bases. For example, in science or religion better positions can be taken by non-monetary means such as intellectual capacity or spiritual charisma. The spread of certain ideas or worldviews can be an objective interest in itself. These two dimensions are different in principle but will be connected here in a specific way that helps to solve open questions about the origins of Social Security in the US.
Material and Ideal Interests
Classic liberals put trust in the rationality of individual interest pursuit as a basis for a peaceful social order. Interests were defined as material, almost physical self-love of the individual. Functionalist social theories shared this concept of interest but doubted Smith’s paralleling of individual and common interest. Instead, the Hobbesian problem of order was imminent: “A social order resting on interlocking of interests, and thus ultimately on sanctions, is hence hardly empirically possible” (Parsons [1937] Reference Parsons1979, pp. 404f.; Durkheim [1893] Reference Durkheim1997, pp. 179ff.). Subjective interests were thought to be much too volatile to make social stability possible. For a successful social integration norms based on cultural values have to limit or replace the pursuit of individual interest (Swedberg Reference Swedberg2005, p. 384).
In the Marxian tradition interests do not reflect the opposition between individual and society but can be derived from basic social structures of the economy. Interests are closely tied to class, rooting them within the objective structure of capitalist production. It is unclear how much structural determinism there was in Marx’s work, how confident he was that the class of itself would eventually turn into a class for itself. However, criticism within political sociology and political economy primarily referred to Marx’s reduction of interest to the distribution of material value. Many sociologists have stressed the influence of other social conflicts on the formation of political interests. Weber ([Reference Weber, Roth and Wittich1922] 1978, p. 302) pointed to religious and professional distinctions within the labor and the capital class. Lipset and Rokkan (Reference Lipset, Rokkan and Lipset1967) argued that emerging political parties reflect different social cleavages such as state vs. church or city vs. countryside. Dahrendorf ([1957] Reference Dahrendorf1959, pp. 182ff.) stressed the importance of organizational means for the formation of interest groups within structures of authority. Research on neo-corporatism has shown how capital and labor interests are shaped by the institutional structure of markets, firms and the political system (Hollingsworth and Streeck Reference Hollingsworth, Streeck, Hollingsworth, Schmitter and Streeck1994; Iversen and Stephens Reference Iversen and Stephens2008; Mares Reference Mares, Hall and Soskice2001; Hall and Soskice Reference Hall, Soskice, Hall and Soskice2001; Jessop [1993] Reference Jessop and Vij2007). Historical institutionalist approaches have described how existing policies shape the interests actors pursue, in the defense of institutionally privileged positions (Pierson Reference Pierson2000; Thelen Reference Thelen1999). The common root of all objectivist interest-oriented approaches is the assumption that the interests a political actor pursues can to a large extent be derived from his or her social position in a political or economic field. Bourdieu (Wacquant Reference Wacquant, Bourdieu and Wacquant1992, p. 25) has formulated this objectivist concept of interest in the most general sense, applying it to ideal (symbolic, cultural) as well as material interests.
Objective and Structural Interests
There is a second strand of criticism, in which the political economy perspective on interests is not criticized for focusing on material interests, but for its assumption that individual subjective interests can be derived from objective social structures. Instead, interests should be defined as subjectively meaningful goals in specific situations (Swedberg Reference Swedberg2005, pp. 362ff.). The most radical position on this side is taken by standard economic theory. Interests are only described as subjective preferences that are exogenous individual “tastes” and do not need to be traced back to social macro-structures. They should only be examined in their aggregation and interaction. A similar argument can be found also in early sociological theory of Georg Simmel. For him (Simmel Reference Simmel and Frisby2008, p. 469), sociological analysis could only examine which forms of community (Gemeinschaft) emerge from the interaction of subjective individual interests (that may well be material or ideal). Interactionist approaches, while they do not share the assumption that the formation of interests cannot be analyzed sociologically, still argue that interests are subjective action orientations. They stress the formation and negotiation of interests in concrete interactions (Whitford 2002). As Hochschild (Reference Hochschild, Goodin and Tilly2006, p. 290) argues “a person’s actions are directed by an understanding of his or her interests, which are derived from ideas or conceptions of the self in a particular context”.
A three-dimensional concept of interest
Although there are two independent dichotomies within the concept of interest there are also some hints in sociology at the question how these dimensions could be brought together. There are some approaches which argue that the formation of subjective interests or preferences can be understood as an idea-based sense-making of objective positions. One of the most striking is by Weber:
Not ideas, but material and ideal interests, directly govern men’s conduct. Yet very frequently the “world images” that have been created by “ideas” have, like switchmen, determined the tracks along which action has been pushed by the dynamic of interest (Weber [1915] Reference Weber, Gerth and Mills1961, p. 280).
This citation is remarkable in two respects. First, Weber refers to “material and ideal interests”. This points to the above argument that adherence to or spread of certain values or beliefs can be an interest in itself. Second and more importantly, Weber points to the role of ideas in any interest-oriented action. Even if material interests build the basis for action, ideas still influence “the tracks” of action like switchmen. Here Weber states that an examination of the objective origins of interests is not enough to explain observed behavior. Subjective action orientations are derived from objective (material and ideal) interests, a process for which ideas are essential.Footnote 1
In recent years, there have been many policy studies that follow Weber’s perspective and open space for an idea-based process of sense-making between objective positions and subjective action orientations (Campbell Reference Campbell1998). Schmidt (Reference Schmidt2008, p. 314) has argued that “it is the process in which agents create and maintain institutions by using what I call their background ideational abilities”. Structures of meaning that shape discourses create change and dynamics by influencing interest perceptions. Social policy researchers have focused on the discursive construction of social problems and cognitive categories that make administrative intervention technically possible. Especially in the field of unemployment insurance, many authors have stressed the evolution of “social knowledge” (Topalov Reference Topalov and Palier1994, p. 494) around the turn of the 19th/20th century as a pre-condition for state action, which shaped very much possible interests and strategies (Zimmermann Reference Zimmermann2006; Mansfield et al. Reference Mansfield, Salais and Whiteside1994; Steensland Reference Steensland2006). In her study of American employment policy, Weir points to the special problems entailed by the use of the concept of interests: “Potential group members [in an alliance] do not always know their interests in a specific policy area […] questions must be asked about why one policy is favored over the other” (Weir Reference Weir2005, p. 253). In his studies of paradigm changes in American and Swedish economic policy history Blyth has stressed that interests are “clusters” which contain material needs as well as cognates and beliefs (Blyth Reference Blyth2002, p. 30). The reason for this is basic economic uncertainty. In contrast to the assumptions of standard economic theory, actors most of the time do not have full information and it is very often not possible to calculate risks. Under profound uncertainty, “agents can have no conception as to what possible outcomes are likely, and hence what their interests in such a situation in fact are” (Blyth Reference Blyth2002, p. 32). As Steinmetz puts it in his case of early German welfare institutions “social policies […] are structurally underdetermined” (1993, p. 53). Cultural tools help the actors to build stable expectations about the surrounding economic conditions (Swidler 1986). Concrete interests cannot be derived from market position alone, because an element of uncertainty is necessarily involved (Beckert Reference Beckert2002). While they are aiming at their general, positional interests, agents need to realize their “given” interests in “risky” situations (Blyth Reference Blyth2002, p. 9; Braun Reference Braun, Braun and Busch1999, p. 12).
How can ideas give orientation for interests? This can be answered by looking into phenomenological approaches to rational action. Schütz defines ideas as typifications of the social world, which means they are deeply rooted in experiences of the actors (Schutz [Reference Schutz1932] 1967, p. 13). In Schütz’s view, the actor over-stresses only some aspects of the world that have high relevance from his perspective. This implies a “one-sided escalation” (“einseitige Steigerung”) of some aspects and a distortion of reality in the name of coherence of the picture drawn (Weber [1905] Reference Weber and Winckelmann1982, p. 191). The actor learns from past experience by treating the world at every step as if it were coherent. It is the escalation towards coherence that makes ideas an independent influence. Ideas define normative and cognitive – often deeply intermingled – images that guide this creation of coherence. It is a process of “disentanglement” which is essential for turning the world into a less complex, calculable environment (Callon Reference Callon and Callon1998).
As the above figure shows, the formation of interests cannot be reduced to the interplay between structural positions and a specific strategic context alone. In economic terms, it is not enough to look only at the interplay between structural market position and situational contexts. Instead, ideas help actors to make sense of over-complex, uncertain situations. It is also not enough to focus only on ideas and the situational formation of interests, as some constructivist approaches tend to do. The socio-economic position filters the possible interest definitions.
The aspects of plausibility and legitimation of ideas describe the willingness of political actors to employ a specific worldview for the formation of their interests. It is an empirical question which ideas are plausible and legitimate enough to guide interest formulation. Turning this concept into a dynamic model, there are two possible sources for a change of interests (beside a radical shift of socio-economic positions). First, the context can change, causing actors to reevaluate which concrete goals will best serve their positional needs. Second, new ideas can gain plausibility or legitimacy, inducing a change of how employers make sense of the material and institutional context and translate their positional needs into support or opposition to specific policies. It will be shown that these two processes have coincided to make a change of business interests in the emergence of the American welfare state possible.
Business Interests in the Making of the American Welfare State
The “giant leap” towards a modern welfare state in the early New Deal makes the introduction of the Social Security Act (ssa) in 1935 a crucial case for welfare state analysis. Although we know from many studies that there were important institutional precedents (Rodgers Reference Rodgers1998, pp. 245ff.; Skocpol Reference Skocpol1992, pp. 102ff.), social insurance for the unemployed and the elderly did not emerge at a national level before 1935. Many welfare-state researchers have turned to the institutional structure of the political system to explain the limited degree of social policy in the US. The fragmented political system of the US gives structural advantage to opponents of social legislation (Orloff 1993, p. 289). But what made it possible for the “welfare laggard” America to jump into a central social insurance program in the 1930s? Political scientists have argued that the implementation of the ssa is largely due to the initiative of independent social reform experts and their “capacity to carve out a viable political place for limited reforms” (Skocpol and Ikenberry 1983, p. 135). Social security in the early New Deal could only be implemented by a potentially autonomous administration that was expanding federal state capacity (Skocpol Reference Skocpol1992). In a time of crisis, when popular demand for social reform was high, independent middle-class reformers were able to implement an American alternative to the European models (Weir Reference Weir2005, p. 253). As Weir (Reference Weir, Skocpol, Evans, Rueschemeyer and Skocpol1985, p. 115) has argued: “Existing patterns of state intervention and the initiatives of political leaders often activate particular interests and coalitions within a range of alternative possibilities”.
These historical-institutionalist approaches have been criticized for falling into a trap of pluralism by other political scientists (Domhoff Reference Domhoff1996, pp. 253ff.). It is necessary to ask whose interests stand behind the emergence of the American welfare state. Accordingly, some authors stress the role of employer interests in the development of the Social Security Act. Domhoff (Reference Domhoff1996, p. 122) has shown the close personal and institutional links between the New Deal administration and the “Rockefeller network” for the ssa. Employer interest was to weaken labor unrest and strike activities by institutionalizing more peaceful industrial relations.
In recent years, a new argument about business interests in social policy has gained support. It criticizes the class struggle argument and claims that capitalists can have a positive, substantial interest in regulating labor markets. In market economies with a high degree of coordination in long-term relations, social protection against unemployment increases workers’ productivity and their willingness to invest in industry- or firm-specific skills (Iversen/Stephens Reference Iversen and Stephens2008; Hall/Soskice Reference Hall, Soskice, Hall and Soskice2001). Company size, skill specificity and exposure to international competition are crucial factors for the formation of business interests on the labor market. This is not only observable in the comparison of different countries but also for different employer groups within one country. As Mares (Reference Mares, Hall and Soskice2001) has shown, big and internationally oriented companies with specifically skilled workers have a high interest in controlling social benefits in order to use them as an incentive for human capital formation. Smaller, nationally bound companies with the need for general skills favor universally pooled social protection that is tax financed because, for them, the cost of social protection is a bigger problem than the control over benefits.
For the case of the ssa a similar argument can be found. After the breakdown of industrial corporatism that followed the First World War, many employers turned to efficiency wage strategies that included company-level social benefits (Nelson Reference Nelson1969, pp. 40ff.). During the Great Depression, those employers that followed this strategy of segmentalist wage-setting faced “cut-throat competition” from low-wage and low-price competitors (Swenson Reference Swenson1997, p. 69). Compulsory social security was attractive to help them reduce these competitive disadvantages by penalizing low productivity, low job security strategies. This created the potential for a cross-class alliance for social policy that the New Dealers anticipated (Swenson Reference Swenson2002, pp. 192ff.). They believed that employers would switch to supporting market regulations after these had been introduced, as they had done for other pieces of legislation before (Swenson Reference Swenson1997, p. 76). This gave Roosevelt confidence that Social Security would later be accepted, although the National Association of Manufacturers (nam) and the US Chamber of Commerce (uscc) opposed the bill beforehand.
Critics of Swenson argued that even if employers influenced legislation towards a business-friendly model this does not mean that they were actually favoring Social Security over the status quo. The centralist turn of the New Deal administration left employers no choice but to strategically adapt and take part in the process to prevent worse. Employers found themselves “operating in a climate of diminished influence, selecting from a list of options not of their own choosing” (Hacker and Pierson Reference Hacker and Pierson2004). Some employers turned to their second-best choice, a business-friendly model with marginal benefits demonstrating that revealed preferences in a legislative process do not necessarily reflect a substantive change in interests (Hacker and Pierson Reference Hacker and Pierson2002, pp. 283ff.). A substantial weakening of employers’ structural power (exit) can also cause a strategic reorientation towards instrumental power (voice) which means direct lobbying and taking active part in the political process.
Empirically this debate was heated by the shared observation that very few employers explicitly supported ssa (Hacker and Pierson Reference Hacker and Pierson2002, pp. 299ff). The nam and the uscc were particularly harsh in their opposition throughout the legislative process. Still, a handful of supporters were highly present and active promoters of the ssa in the legislative process during 1934 and 1935. For Swenson, these few employers were important for creating trust among New Dealers that employer opposition would vanish afterwards (Swenson Reference Swenson2004, p. 4). However, his critics doubt the structural change of interests and claim that employers with company plans by the 1930s represented a “miniscule portion of the overall business community” (Hacker and Pierson Reference Hacker and Pierson2004, p. 191). Swenson then countered that the historical sequence contradicted the thesis that employers were merely strategically adapting: the small group of supportive employers was already promoting unemployment insurance in 1931, while Hoover was still president (2004, p. 20). Moreover, as Swenson shows by referring to debates within the employer camp, those segmentalists supporting Social Security risked acquiring a reputation for radicalism. This debate is an example of an interpretative fight: As Hacker and Pierson (Reference Hacker and Pierson2004, p. 193) rightly stress, it cannot be solved by adding more historical material. It boils down to the interpretation of interests articulated explicitly by American employers.
While struggling with the “honesty” of business leaders’ description of their self-interest, an important point is left out of the analysis on both sides. This concerns the direction of the strategic reorientation. For both sides the Great Depression is crucial as ultimate cause, as impulse, for a change of business strategies. For Swenson the crisis meant cut-throat competition for segmentalists; for Hacker and Pierson it meant a loss of structural political power and a centralization and interventionism by the state. However, it is much less clear why business favored a certain type of welfare state institution; why the new interest definition developed in a certain direction. To Hacker and Pierson the open question is: Why did business leaders direct their strategical lobbying towards unemployment insurance as a “means for stabilizing the economy”? Other possible goals could have been a “self-financed risk pool for the employees” or a “tax-based system of means-tested unemployment benefits”. There must have been a re-definition of interests on their side (even if only second best interests) that was more than just reducing cost or blocking social policy in general. To Swenson the question would be why segmentalists (or the segmentalist avant-garde), facing low-cost competition in their markets, turned to national social policy programs to help them instead of turning to strategies of cartelization or market regulation, as they had done in earlier crises? Both questions refer to a learning process on the side of business leaders that cannot be explained by the crisis situation alone – the idea-aspect of interest points towards the fact that there are always different ways out of a situation of crisis and ideas influence where to seek a solution.
Unemployment Insurance in the United States from 1911 to 1935
In this section I will demonstrate that the change in business interests toward the support of the ssa in the early 1930s cannot be sufficiently explained by the economic context of the Great Depression and the political context of the New Deal administration. Although both aspects created a problem for employers that made it necessary to redefine their interests, we can only explain the direction of this reorientation when we take into account the change in legitimate ideas among business leaders, which we can trace back to historical experiences before the 1930s. To illustrate this, I will focus on the question of unemployment insurance (UI) within the ssa.
Four steps are necessary to illustrate the argument. I will first describe the two sets of ideas that shaped the American discourse on UI among social reformers between 1911 and the SSA. I will then show that initially all employers were opposed to state-led unemployment insurance before the Great Depression, even during earlier economic crises. Third, I will examine how new experiences in management and economic policy as well as the institutionalization of workmen’s insurance in the 1920s raised the legitimacy of liberal-corporatist ideas among some American employers. Finally, I will illustrate how employers changed their positions towards unemployment insurance in the early 1930s in accordance with the cognitive and normative ideas that gained legitimacy in the 1920s. By comparing UI with other policy options, I will show that UI, while not the only solution to business problems of competition and growth, was the most plausible to a group of employers who had been engaged with a set of social reform ideas before.
Between 1911 and 1929, there were two perspectives on unemployment insurance that shaped the American debate among social reformers: the Ohio and the Wisconsin school of thought. The Ohio School’s main representatives were Isaac M. Rubinow, Charles Henderson, and Abraham Epstein. In 1913 Rubinow, a physician of Russian origin, proposed nationwide compulsory social insurance systems for the unemployed, the sick, and the elderly, very much like those introduced in Germany and Great Britain (Rubinow [1913] Reference Rubinow1916). He justified unemployment insurance with the results of studies on the endangered social and medical well-being of the working class in America. Alleviating the destitution of unemployed workers, which carried with it economic, psychological and moral hazard, was the primary aim of social policy (Rubinow [1913] Reference Rubinow1916, p. 44; Ohio Commission on Unemployment Insurance 1932, p. 11). As the reasons for unemployment lay in the unavoidably cyclical character of the capitalist economy, the federal government had the responsibility to create an UI system capable of protecting workers’ standards of living (Epstein Reference Epstein, Sellin and Young1933, p. 21; Rubinow [1934] Reference Rubinow1976, p. 314; Rubinow [1913] Reference Rubinow1916, p. 44). Labor market policy was tantamount to a weather forecast: the only solution was to select appropriate clothes (Henderson Reference Henderson1913, p. 172). Wherever private charity and local relief were not able to compensate for destitution, the federal state had the right and duty to introduce a compulsory system of UI. The low wages paid in most industries provided no chance to build up sufficient protection individually (Ohio Commission on Unemployment Insurance 1932, p. 13). Rubinow described UI as “true class legislation” (Rubinow [1913] Reference Rubinow1916, p. 491). Its purpose was to force those groups of employees and employers in the more stable sectors to pay for the risk faced by workers in other industries as an act of solidarity. The Ohio school of thought can be characterized as regulatory-reformist: reformist because it favored the adoption of social policy as a counterforce against capitalist dynamics inevitably leading to cyclical unemployment; and regulatory because its proponents turned to state coercion to introduce a system of compulsory unemployment insurance.
The second and eventually more successful school of thought did not fully develop before 1920. It was led by John R. Commons, an institutional economist at the University of Wisconsin. He was induced to follow this topic after the First World War by John B. Andrews, his doctoral student and secretary of the American Association for Labor Legislation (aall). Commons criticized the European models for their “un-American” coercive elements. American employers would never accept the idea of making some companies pay for the poor employment performance of others (Commons [1934] Reference Commons1961, p. 863). Commons had been advisor to the Wisconsin Industrial Commission (wic) since 1911. Commons made the problem of unemployment a question of appropriate market regulation: how to even out the economic cycle by setting the right incentives for employers. He advocated penalizing companies that laid off workers too easily, thereby inducing them to apply long-term management strategies. In this regard, his proposal followed the logic of workers’ compensation for work-related accidents, as it favored prevention over benefits (Commons [1921] Reference Commons, Rutherford and Samuels1966, pp. 294f.). Instead of the needs of the unemployed, the regularization of the economic process became the primary goal of UI (Andrews Reference Andrews1915, p. 189). In contrast to Rubinow, Commons believed that stretching out the business cycle could enable the market to prevent cyclical crises. Unemployment was not seen as inevitable in a capitalist production system; it was rather a sign of “maladjustment” (Leiserson Reference Leiserson1914, p; 326). Social policy was not about forecasting and reacting but about economic engineering (Commons [1934] Reference Commons1961, p. 8). The primary means for reaching that goal was not coercion but institutional incentives for employers. The contributions to a system of unemployment insurance should induce employers to voluntarily regularize employment over time where coercive regulation would have been circumvented by employers anyway (Andrews Reference Andrews1920, p. 239; Andrews Reference Andrews1914, p. 213). UI was considered to be an instrument of economic policy and not of social policy (Commons [1934] Reference Commons1961, p. 875). The Wisconsin school traced labor market problems back to the inevitable contradiction between long-term and short-term strategies of employers. Even if managers wanted to calculate over a longer period, short-term cost threats made this impossible. UI aimed to counter those cost threats by making contributions dependent on every single firm’s employment record (Commons [1934] Reference Commons1961, pp. 865ff.). The Wisconsin School can be characterized as liberal-corporatist: liberal because there was the assumption that market self-regulation would lead to maximum societal welfare if appropriate institutions set the right incentives; and corporatist because they wanted to integrate employers (and trade unions) into the stabilization of the economy on a voluntary basis (Commons Reference Commons1934: 134).Footnote 2
From the First World War through the 1920s, the two schools conducted their debates by publishing monographies and writing articles in the American Labor Law Review and other social reform magazines. Social reformers also took part in most of the congressional and state hearings on the question of unemployment insurance that took place between 1916 and 1935. Still, their impact on politics remained marginal before the Depression.
Rejection of Unemployment Insurance before the Great Depression
Although there were severe crises in 1914-1915 and 1920-1921, rejection of any form of collective UI was consensual among employers – and also among trade unions and the federal government in those years. In 1914-1915 and 1921-1922, employment, growth, productivity and wages strongly decreased with a growth of – 6.5 percent in 1914 and – 2.44 percent in 1921. Unemployment rose from 5.4 percent in 1913 to 12.9 percent in 1914 and from 4.3 percent in 1920 to 21.20 percent in 1921 (Carter 2006b, pp. 23ff.; Carter et al. 2006a, pp. 265ff.; Commons et al. 1935: 128). Although the downturn of the Great Depression was much more disastrous, it is surprising that in these earlier downturns, which at the time were unprecedented, employers showed no hesitation in their rejection of UI. The aall hosted two conferences on unemployment in 1914, and trade secretary Herbert Hoover established a similar conference in 1923. Most participants were skeptical about the possibility of calculating a system of UI unless more accurate labor market data could be collected (National Conference on Unemployment 1914, p. 251f.; President’s Conference on Unemployment 1923). Most of the employers participating in the 1914 conferences doubted the extent of unemployment (National Conference on Unemployment 1914, pp. 227ff.). Some employers accepted that there were structural causes of unemployment, such as Samuel A. Lewisohn, director of the Bank of America (US Senate 1929), but still heavily doubted the feasibility of UI due to a lack of statistical labor market data and experience in the field (US Senate 1929). Additionally, employers saw unemployment benefits as a “dole” that endangered the independence of the American worker and the sources of business productivity (Bremner Reference Bremner1991, pp. 17ff.; Katz Reference Katz1996, p. 25). This argument they shared with the American Federation of Labor (afl). Even those employers who built social benefit systems at the company level saw no use in bringing the state into the organization of unemployment insurance. Some studies have explained the failure of social insurance in individual states before the Great Depression by the fragmented authority structures and party patronage systems in some states (Amenta, et al. Reference Amenta1987; Orloff Reference Orloff, Weir, Orloff and Skocpol1988). Employer opposition was often the force behind those institutional vetoes. At neither the central nor the state level did employers have an interest in unemployment insurance before the Depression. Indeed, American employers considered UI a threat to their interest in profitability and stable growth during this period. Even in times of crisis, when these common interests could not be safeguarded by a free labor market, UI was not seen as a plausible or legitimate answer to economic downturns.
The Ideational Shift towards Liberal-Corporatist Ideas among Employers in the 1920s
In this section I will show how liberal-corporatist ideas gained legitimacy and plausibility among American business people in the 1920s. There are two sources of ideational change we must consider. First, actors and networks can succeed in forming and institutionalizing “advocacy coalitions” to disseminate certain worldviews (Sabatier Reference Sabatier, Sabatier and Jenkins-Smith1993; Haas Reference Haas, Haas and Peter1992). The members of the Wisconsin network were extraordinarily successful at building up a community of young politicians and business people around the liberal-corporatist approach. Second, another mechanism is the existence of successful examples or tests of a certain idea. Ideas can gain influence if they have an intrinsic affinity to institutions and policies in other fields that are considered as successful. Long-term oriented management strategies, a growing optimism in economic planning, and the spread of workers’ compensation all had an ideational affinity to the Wisconsin school without being directly linked to the question of UI.
The Wisconsin Network and the Settlements
The Wisconsin group saw the spread of liberal-corporatist ideas as a common policy enterprise (Haas Reference Haas, Haas and Peter1992, p. 3). The community was made up of mostly students and assistants of Commons and Ely at the University of Wisconsin. These two professors put their doctoral students to work on a variety of practical political tasks in Wisconsin and other states (Schlabach Reference Schlabach1969, p. 25). Many of the administrative executives of the ssa had worked with Commons in Wisconsin during the 1920s, among them Edwin Witte, executive director of the Committee on Economic Security, Robert M. La Follette, Jr., who would replace his progressive father in the Senate in 1925, and Elizabeth Brandeis, later counselor to President Roosevelt. Many legislators of the ssa first worked together on academic and practical work in Wisconsin; as Edwin Witte wrote in his biography, “I owe to Commons my entire outlook on life and a great many of my ideas”. (Schlabach Reference Schlabach1969, p. 19). Witte developed the Wisconsin Workmen’s Compensation Act together with Commons, became consultant to the wic in August 1912, and worked for Robert M. La Follette, Sr., in the election campaign of 1924.
Moreover, the liberal-corporatist worldview was organizationally diffused through the urban settlement movement in Chicago and New York (Davis Reference Davis1994; Schlesinger [1957] Reference Schlesinger2002, pp. 26ff.). Many progressives at the turn of the century hoped to improve the urban social and cultural infrastructure through the newly built “settlements”, which were supposed to provide education to the poor in order to reduce dependence on crime and the dole (Bremner Reference Bremner1991, p. 57). The settlements were meant to bring together all social classes into one urban neighborhood, thereby improving social integration and modernizing American cities (Katz Reference Katz1996, pp. 164f.; Davis Reference Davis1994, p. 3). They provided space for living, but also for political and cultural activities. Many reform-oriented groups – such as the women’s movement or trade unions – held their meetings in the settlements. At the same time, the settlements offered rooming for academics who wanted to study social problems (Davis Reference Davis1994, p. 171). The two most prominent settlements were Hull House in Chicago, founded in 1889, and Henry Street, founded in 1883.
Ideologically, the Wisconsin School was much closer to the settlements than was the Ohio School. Both Commons and Ely often lectured in the settlements (Davis Reference Davis1994, p. 171). Rubinow, in contrast, had been a founding member of the American Socialist Party in 1901 and frequently attacked the Progressivist voluntarism in the settlement movement (Lubove Reference Lubove1986, p. 40). Among those New Dealers that lived for some time in settlements as students were Frances Perkins, secretary of labor from 1933 to 1945; Robert F. Wagner, chief legislator of the New Deal in the Senate; and Adolf A. Berle, economic advisor to President Roosevelt. Moreover, Joseph B. Eastman and Gerard Swope, who were later to become two very prominent business supporters of the ssa, had lived in settlements (Schlesinger [1957] Reference Schlesinger2002, p. 25). During the 1920s a network between universities and public policies in cities like Wisconsin, Chicago, and New York, was formed in which many important actors who would later form a coalition supporting the ssa were engaged. The peak representatives who would participate in the Committee on Economic Security (ces) of 1934 were quite familiar with one another.Footnote 3 While this does not mean that their political agendas concerning UI were fully set in the 1920s, the settlement activities do show how the New Dealers were able to find some employers and labor activists who shared basic beliefs stemming from a shared political background. The supporting group for the ssa were not only familiar with these ideas from book reading but were at least for some time involved in a specific communication process among social reformers that focused on certain ideas and rejected others. They put resources into keeping up an alternative path (Crouch and Farrell Reference Crouch and Farrell2004) in a deeply conservative era on the federal level, moving their perspective in a privileged ‘starting position’ for the next period of political change.
New Emphasis on Management and Macroeconomic Planning
To become influential, ideas need legitimation beyond the groups injecting them into the public discourse (Parsons Reference Parsons2002; Schmidt Reference Schmidt2002). Otherwise, it is doubtful that these new administrators could have been able to find support for their policies. There were new trends in management and a rise of new economic policy paradigms in the 1920s that had a special affinity to the liberal-corporatist worldview.
The depression of 1921-1922 was the eighth economic downturn that the US had faced since the end of the Civil War. This seemed to prove the existence of a six-year cycle of recession that had not been recognized before (US Senate 1928, p. 24). The most influential book on macroeconomics in the early 1920s was The Business Cycle by Wesley C. Mitchell, in which he argued that “the irregularity of other sequences arises from varying combinations among sequences themselves regular”. (Mitchell Reference Mitchell1913, p. 450). There was widespread optimism among labor market actors that it would be possible to counter the business cycle and guarantee stable long-term growth (Leuchtenburg [1953] Reference Leuchtenburg1993, p. 179f.). Management and economic policy in the 1920s became centered on the goal of regularization.
New Emphasis on Management
A new management legacy in American industry began with Henry Ford in the early 20th century. Ford paid higher wages to his labor force in order to raise productivity and consumer demand, while at the same time aiming to weaken trade union activity in his company. The “geometric” growth of the automobile industry in the 1920s made Ford a role model (Leuchtenburg [1953] Reference Leuchtenburg1993, p. 185). This segmentalist labor market regime built on the high fragmentation of wages within his industry was proposed by more and more managers in the 1920s (Nelson Reference Nelson1969, p. 35; Swenson Reference Swenson2004, p. 5). However, most 1920s employers who created company plans against unemployment wanted to achieve temporal regularization of the business cycle rather than a long-term formation of human capital (Commons et al. Reference Commons, Lescohier and Brandeis1935, p. 151).
The first company-level UI was established in 1916 by the Dennison Manufacturing Company, a paper producer from Framingham, Massachusetts . By 1919, the company had built up a reserve of $150,000 and was able to provide its unemployed with half of their weekly wages for a few weeks. Similar systems were adopted by S.C. Johnson in 1922 in Racine, Wisconsin and by Leeds & Northrup in 1923 in Philadelphia (Nelson Reference Nelson1969, p. 53f.). The arguments for adopting such systems were based on the idea of inducing management to prevent layoffs: unemployment benefits were a method for employers to discipline themselves into planning for the long term (Nelson Reference Nelson1969, p. 51). Procter & Gamble introduced a “job guarantee” of 48 weeks of paid work a year for its 5,000 employees. Similar plans were adopted by other paper and food companies (US Senate 1932, p. 9). For Edward A. Filene, from Boston, cyclical unemployment was a problem of “waste”: a waste of production factors and of possible demand (US Senate 1931, p. 237).
The Rochester plan was mutually established by 14 companies of different size (48 to 13,000 employees) in Rochester, New York (US Senate 1931: 102). It was a regularization plan that contained many elements of managerial planning, of which social benefits were a part. Marion Folsom of Eastman Kodak, who supervised the plan, explained it this way:
Some of the methods which they used are accurate forecasting of sales, careful planning, scheduling of production, and building up inventories during slack seasons, diversification of products, education of public versus seasonal buying, changing hours to meet changes in volume, price concessions during offseasons (US Senate 1931, p. 102).
The industries involved had vast structural differences: companies in fields such as chemicals, optics, telecommunications, machine tools, textiles, and lithography took part. Other important regularization plans can be found in retail, paper and energy companies, among them General Electric and Standard Oil. Clearly, neither the degree of firm-specified skills nor exposure to international competition could predict who followed the “New Emphasis” in management (Nelson Reference Nelson1969, pp. 35ff.). From a historical perspective, the growing plausibility of anti-cyclical stabilization did not come from considerations concerning employee skills, but from a new encompassing strategy of anti-cyclical management techniques.
The ideational process towards regularization in the 1920s was not limited to managers; industrial relations consultants and policymakers were also inspired by similar concepts. The Senate hearing on unemployment in December 1921 opened with a statement by Wesley Mitchell on business cycles (US Senate 1921, pp. 4ff.). Ideas of collective institutional regulation were connected to high hopes for stabilizing growth (Schlesinger [1957] Reference Schlesinger2002: 130ff.; Mitchell Reference Mitchell1913). They aimed at “taking off the top of the wave” (US Senate 1921, p. 11). Anti-cyclical adjustments of public works and infrastructural investment were seen as new modes of governance to breed economic progress (US Senate 1921, p. 22). As Sam A. Lewinsohn, president of Miami Copper and Tennessee Copper & Chemical, put it:
My point is that economic history indicates that it is often a small stimulus that starts the resumption of business activity, it is a stimulus that we need and it is just that that a postponement of these public projects to a time of depression might accomplish (US Senate 1921, p. 22).
Even the conservative federal administrations of the ‘20s were sure that active government was needed to ensure maximum production. Anti-cyclical financing of public works and construction thus became the core growth strategy of conservative administrations in the 1920s (Rothbard [1963] Reference Rothbard2005, p. 188). Herbert Hoover – then Secretary of Commerce – argued in 1923:
slumps are in the main due to the wastes, extravagance speculation, inflation, over-expansion, and inefficiency in production developed during the booms, the strategic point of attack (President’s Conference on Unemployment 1923, p. vi).
Elimination of waste by anti-cyclical planning became a legitimate orientation for economic policymaking.
Workers’ Compensation
Ideas can gain legitimation if they can meaningfully connect to successful institutional examples. Thus, another reason for the growing legitimacy of liberal-corporatist ideas was the success of the corporatist model of workers’ compensation before the First World War (Orloff 1988, p. 53). After the turn of the century, US civil courts had gradually changed their jurisdiction concerning employer responsibility for industrial accidents (Leiby Reference Leiby1978, p. 203f.). The older clauses in the law had meant that employers had no obligation as long as any failure of the workers was involved, and this no longer fit the reality of industrial workplaces (Skocpol Reference Skocpol1992, p. 293). Juries were reluctant to abide by these older regulations and tended to decide in favor of the injured. Lawsuits exploded, and (Theodore) Roosevelt declared in 1908 that the US should follow the European countries in establishing workers’ compensation (Rodgers Reference Rodgers1998, p. 247).
In 1911, the nam decided to endorse workers’ compensation laws (Leiby Reference Leiby1978, p. 205). Between 1910 and 1913, 22 states adopted workers’ compensation. The wic established a system in which employers paid contributions into a mutual fund that provided benefits for workers injured in work-related accidents. The wic was given the right to legislate the administrative details (Katz Reference Katz1996, pp. 198ff.). Commons’ idea was centered on prevention, while the benefit side remained marginal. Employer contributions could be reduced if companies were successful in preventing new accidents, thus turning the field of accident prevention into a profitable area for good management (Commons Reference Commons1934, pp. 141ff.). The wic hired factory inspectors who gathered data within the companies and discussed their findings with representatives of employers and trade unions. The aall was not successful in its mission to make workers’ compensation compulsory and universal in all states (Skocpol Reference Skocpol1992, p. 299). However, it was successful in putting administration into the hands of industrial commissions, along the lines developed in Wisconsin. By 1920, 32 of 45 states with workers’ compensation laws administered those laws through industrial commissions. By 1933, 44 of 52 states were following the Wisconsin model.
It is important to stress that workers’ compensation was not a direct institutional precedent for UI. Establishing a labor market administration would have meant expanding state capacities, while workers’ compensation was still possible within the fragmented patronage system of state politics. Moreover, such a system was attractive to employers only under the cost threat of court decisions (Skocpol Reference Skocpol1992, p. 298; Orloff 1988, p. 58). However, the spread of workers’ compensation and the institutionalization of corporatist forms of social welfare provided positive examples that liberal-corporatist ideas of regularization and prevention were plausible and feasible.
To sum up, between 1911 and 1929, many American employers experimented with management strategies aimed at governing the business cycle. They also noted the success of the institutions that brought together labor and capital interests in order to carry out such strategies. These institutions were not introduced in labor market policy, but in the field of work-related accidents. During the 1920s, some of those businessmen who were later going be involved in the process towards the ssa promoted their companies’ benefit systems within the business community based on their positive experience.
The important aspect of the mechanisms described above is that they provided evidence for the plausibility and feasibility of the liberal-corporatist worldview. This endowed the ideas of regularization, prevention and stabilization with more legitimacy than their competitors from Ohio who were very much entangled with American socialists. The argument is not that UI gained legitimacy as an institution, but that those cognitive and normative ideas that made up the common Wisconsin perspective passed important “reality tests” in other fields. Even though there was little reason for employers in the economic prosperity of the 1920s to reframe their interests in unemployment policy at all (Nelson Reference Nelson1969, p. 76), the success of the liberal-corporatist community in shaping the American economy provided the main “contender” for conservative policies when the country fell into crisis again in 1929.
The Change of Employer Interests during the Great Depression
This section demonstrates how a group of employers during the Great Depression reframed their interests towards UI along the cognitive and normative ideas of the Wisconsin School. American employers were split on UI in 1935. On one side, the nam maintained its sharp opposition to any form of UI, arguing that even after the Great Depression there would be no alternative to a free, self-regulating market. As employer representatives John C. Gall, James L. Donelly, and James A. Emery argued in all congressional hearings on the topic, UI was still seen as a threat to federalism, profitability and future growth of the American economy (National Association of Manufacturers 1935, pp. 17f.; US House of Representatives Reference Orloff, Weir, Orloff and Skocpol1935, pp. 1021ff.; US House of Representatives 1934, p. 403f.). On the other side, some employers supported federal UI and testified in Congress in support of UI during the Depression – some of them were invited to the Advisory Council of the ces in 1934. I will show that these employers were mainly recruited from the ranks of firms that had oriented themselves towards the Wisconsin worldview in the 1920s. In contrast to the crises of 1914-1915 and 1921-1922, the economic crisis fell on radically changed ideational grounds among those business leaders.
American Employers in the Depression
The pressure that the Great Depression put on American employers is obvious. Industrial output of consumer goods dropped by a third between 1929 and 1933. Output in producer goods dropped by more than half (Carter 2006b, pp. 115ff.). Moreover, a dramatic deflation that could not be permanently followed by wage cuts intensified labor cost pressure on most companies. That led to a sharp decline of profitability. The percentage of American firms that operated at profit fell from 95 percent in July 1929 to 40 percent in July 1932 (Carter Reference Carter2006b, p. 129). What Swenson calls the “cut-throat competition” of the Depression years, which was much more pronounced for those companies that offered social benefits to their employees, made employers seek a new institutional frame for stabilizing competition and industrial growth in the future (Swenson Reference Swenson1997, p. 69; Swenson Reference Swenson2002, p. 194f.). Even the US Chamber of Commerce had to admit that all measures to attract employers to regularization on a purely voluntarist basis had failed (Department of Manufacture of the US Chamber of Commerce 1930, p. 3). Employers uniformly acknowledged the severeness of the profitability crisis, but were divided on how to return to economic stability (US House of Representatives 1934, pp. 313ff.). While the nam saw UI as a cost threat that was now less bearable than ever before, other employers reframed their interests positively towards UI. Filene noted that “we are now up against a condition, and not a theory” (US Senate 1931, p. 240). In truth, there were more ideational elements in his new position than Filene was willing to acknowledge: without liberal-corporatist ideas, UI would not have been a plausible response to the “condition”.
Liberal-Corporatist Ideas and Unemployment Insurance
As early as 1931, Gerard Swope, president of General Electric, published a regularization plan for the American economy. His perspective recommended federal social policy to induce regularization.
The foregoing plan […] places on organized industry the obligation of coordinating production and consumption, and of a higher degree of stabilization. (Swope and Frederick Reference Swope and Frederick1931, p. 44).
UI was turned from a threat to business interests into a measure that would safeguard stable profitability by smoothing out the business cycle (Swope and Frederick 1931, pp. 37ff.). Many employer statements in congressional hearings after 1930 clearly show that the meaning of UI had changed according to liberal-corporatist ideas. Marian Folsom argued:
There is no question about it that with the adoption of a plan of this sort, if you have unemployment reserve legislation adopted, in the future employers will do a better job in stabilization (US House of Representatives 1934, p. 71).
Folsom’s words indicate that unemployment was seen as a cyclical but manageable problem. Its consequences were not primarily the destitution of the unemployed, but the destabilization of economic progress, as Edward A. Filene described:
if you start with a large number of unemployed in any business for any reason, you come inevitably, unless you stop it, to an epidemic. It is like an epidemic disease. It spreads and spreads until you get conditions such as you have now (US Senate 1931, p. 240).
Control of the business cycle seemed possible if collective institutions set the right incentives by establishing fiscal sanctions against laying off employees (Kohler in US Senate 1931).
diminished unemployment would mean stabilized industry, more even production, and freer opportunities to increase business profits (Draper Reference Draper1932: 29).
The government was expected to establish an institutional frame to discipline employers for the employers’ own benefit, while still refraining from any direct intervention into labor market structures (US Chamber of Commerce 1933, p. 3f.). The following dialogue between Senator Wagner and Marian Folsom of Eastman Kodak in a congressional hearing shows how the interest shift was rooted in regularization thinking:
Mr. Folsom: — I am very much opposed to governmental unemployment insurance, but I think that industry, employers generally, must do something instead […].
Senator Wagner: — So as to lessen the hardship and suffering resulting from unemployment? Is not that the idea?
Mr. Folsom: — Yes; primarily, though, to furnish an incentive to them to reduce unemployment (US Senate 1931, pp. 111ff.).
The state was to come in only as an indirect enforcer, setting incentives for employer-led social policy for “naïve” employers who lacked foresight.
The attempt to cut down wages and put people out of work, making them unemployed, is the natural, as easy-as-falling-off-a-log idea of meeting over-expense and failure to make adequate profits. One of the greatest things that could be done with all business and for the general prosperity of our country is to really make all of us understand how much greater the field of waste is for saving and increasing profits than is the cutting down of wages (filene in US Senate 1931, p. 242).
Workers’ compensation was frequently cited as a positive example, as Draper’s statement shows:
Moreover, like accident compensation which has stimulated the movement to prevent accidents, unemployment insurance will help to focus the attention of management, in good times as well as in bad, upon the problems of providing steadier employment (US House of Representatives 1934, p. 283).
It was not the pressing need of the actual labor market crisis that provoked this position. Rather, UI was seen as an institution that would induce stable self-regulation in the future. Responsibility for stable employment was still ascribed to the individual employer:
The hazard of what may be called ordinary unemployment, as contrasted with extraordinary unemployment due to a business depression, is more or less within the control of the employer (National Industrial Conference Board 1933, p. 7).
In fact, all five liberal-corporatist ideas can be found where employers justified a change in labor market interests in favor of UI. It is remarkable that the same general impulse (profitability, efficiency and growth) were now invoked in favor of UI that had been cited against it in earlier economic crises. Even though in earlier crises UI had been seen as a danger to economic stability, it was now seen as a stabilizing instrument. The difference between the two crises was a change in legitimate ideas in favor of a liberal-corporatist perspective. Applying these ideas could turn workers’ compensation into evidence that proved the feasibility of UI, even though employers had stressed the very different character of the two institutions ten years earlier. The change in business interests was thus the employers’ own idea-based reconstruction of labor market interests, triggered by a severe crisis.
The Debate about American Business Interests and the ssa
How does the argument developed here connect to the debate between Swenson, and Hacker and Pierson? Although sometimes statements such as “parasitic producers” or “welfare laggards” (US Senate 1931pp. 147ff., p. 244) point in this direction, much less statements can be found in the hearings that connect to human capital formation than to the temporal regularization of the business cycle. However could not the fact that this argument can only scarcely be found simply mean that they were strategically covering their real, structural interests behind the rhetorical reference to regularization? The structure the group of supporters makes this doubtful.
As figure 5 shows, the explicit supporters of UI cannot be described by structural lines of industry, skill level or size. Filene as well as Draper represented industries that were not high-skill oriented. Size also does not seem to be an argument in regard to the problem of risk-pooling. Moreover, the Rochester plan advised by Folsom contained many very different industries such as chemicals, opticals, telephone, textile, lithography and tools with a varying size between 50 and 15,000 employees (US Senate 1931, p. 102). For business supporters of UI the idea of regularization encompassed a broader strategy than organized competition on a market covering employment policies as well as customer relations and investment strategies. The ideational framing of the Wisconsin perspective infected many employers with diverse structural positions on the labor market. This is also true the other way around. First, many industries with strong segmentalist structures such as automobile or machine industry were not among the supporters of the ssa. Second, even for those in a favorable structural position on the labor market, UI was anything but a natural reaction to economic distress; public works, price regulation and cartelization had been much more typical reactions to earlier crises. Ideas can connect to more than one market position and it is possible for two actors to have a similar market position but not a similar idea-based perception of interest. This holds true even for the same actor in two historical moments of crisis. Those who did support UI were linked by their ideational closeness to the Wisconsin school and their experience with scientific measures for engineering the business cycle in the 1920s.
If we explain business support as a strategic adaption to a loss of structural business power, as argued by Hacker and Pierson, a similar argument can be raised. Strategic voice in a legislative process requires the definition of a second-best option that can find at least some acceptance among employers. Before the 1920s other options would have been ranked much higher than UI, such as public works programs or factory-level labor-management cooperation. Only because of the growing plausibility of the Wisconsin worldview could UI be ranked higher – now as a regularization measure and not as an element of socialism as in earlier crises. Although the debate about the role of business interests in the ssa cannot be solved towards one side it has now become clear that both explanations need to pay more attention to the role of ideational changes and learning processes to sufficiently understand why certain business leaders supported ssa while others did not.
This article has provided a conceptual framework for analyzing how ideas and interests interact. There are positional needs that show a high degree of temporal stability, such as profitability, income maximization or economic stability. However, these general orientations need to be translated into concrete, subjective interests, a process of interpretation which entails the evaluation of different policy options. We cannot explain this process of translation solely by the institutional and economic context because in many economic situations, especially in times of crises, outcomes and expectations are uncertain and open to more than one interpretation. This is even more important for the complex question of how possible social policies will influence macroeconomic trends and market structure. Cognitive and normative ideas guide how political actors make sense of this context, which then explains the direction of a re-formation of interests. Ideas should not be opposed to interests and institutions. Instead, they systematically shape the interaction of economic conditions, institutions, and interests, because even highly rational actors like employers need to make sense of the world.
As has been shown in the theoretical section of this paper, the ideational turn of policy analysis is to a large extent a rediscovery of an old problem of interest theory: how do objective structures turn into subjectively meaningful action orientations? This is the point at which social communication about the plausibility and validity of cognitive worldviews and normative justifications enters directly into the perception of social reality, even for highly rational and strategic economic and political actors. Interest analysis in this sense becomes reflexive.
Why do we need a reflexive turn in policy analysis? In this paper I have tried to show how a broadened concept of interest helps to answer lingering questions about the origins of the American welfare state: employers did not discover their interest in UI in the economic downturns of 1914-1915 and 1920-1921, but in the Great Depression. Ideas influenced the constructive efforts of employers to frame UI as something in accordance with their general economic needs. During the course of the 1920s, a liberal-corporatist worldview gained in plausibility and legitimacy among employers. It contained a specific normative and cognitive problem definition centered on the regularization of the economy through state incentives and voluntary employer cooperation. Seen through the liberal-corporatist lens developed in Wisconsin, UI seemed to be a feasible and legitimate means of stabilizing economic development. Yet before the 1920s, UI had been viewed as a threat to economic interests and was associated with European socialism. A change in the ideas that business leaders considered to be legitimate prepared the grounds for the reorientation of employer interests that would be triggered by the Great Depression.
As figure 4 shows, a change of context is not sufficient to explain a change of concrete interests. In the economic downturns of 1914-1915 and 1920-1921, business interests eroded and provoked a search among business people for new orientations. Interests did not change, however, because at the time there was no legitimate worldview that framed UI as a potential answer to safeguard the general interest in economic stability. Liberal-corporatist ideas gained legitimacy in the 1920s, but this growth was not sufficient to reframe interests in a time of economic prosperity. New ideas may be legitimate, but as long as positional needs are safeguarded, there is no need to revisit concrete goals (Hall Reference Hall1993). In the Great Depression, erosion and delegitimation interacted to effect a change in business interests.
Fig. 1 The Sociological Concept of Interest
Fig. 2 Labor Market Policy Worldviews in the US between 1911 and 1935
Fig. 3 Company Backgrounds of Business Supporters of Unemployment Insurance in the 1930s
Fig. 4 The Dynamic of Interest
However, the change of business interests was not a landslide; in fact, it was only partial. This observation has provoked intensive debate in American welfare state analysis. Why did some employers change their attitude towards UI while most others did not, even those in favorable structural positions on the labor market? The answer I propose here could help overcome the interpretative dead-end of the debate: business interests changed only partially, because only those actors inspired by and experienced in liberal-corporatist ideas were able to reframe their interests towards UI. Swenson may be right in claiming that a segmentalist wage setting explains the structural needs some employers faced in the Depression. By reducing employer orientation to market position and driving the aspect of sense-making out of the analysis, however, he fails to explain why so many employers did not discover these material advantages and why the group of supporters was structurally diffuse. Hacker and Pierson, on the other hand, may be right in describing the importance of a change in structural power for the emergence of the SSA, calling the business reorientation a strategic second-best choice. By reducing employer interests to a structurally fixed rejection of social policy, however, they fail to explain why UI could be seen as a second-best choice at all, outplaying all other possible options employers would have favored in earlier times. From a longitudinal perspective it becomes possible with the help of ideas to understand the processes of interest changes while a cross-sectional perspective can explain the moment in which a change of interest becomes probable.
Acknowledgments
I would like to thank the editors of aes/ejs for their support and valuable comments. Sincere thanks are given to Peter A. Hall, Peter Swenson, Jens Beckert, Wolfgang Streek, Daniel Béland and Elisabeth Anderson for comments on earlier drafts. As a matter of course, sole responsibility for all mistakes and lacks of clarity lies with the author.
