1Despite the many vicissitudes it faces, economic sociology has never experienced such radical eclipses that one could diagnose, at any moment in its history, a state close to shambles. During periods of poor institutional visibility, a number of questions raised early on by pioneers of the discipline were in fact revisited, digested, and worked on by researchers who did not necessarily claim the “sociological” label. Throughout the twentieth century, for example, the attachment of labor economics to a certain tradition of economic sociology attests to this. J. J. Gislain and P. Steiner (1995) evoke the importance in this matter of labor economics in the United States dating from the 1920s, and later, spurred on by J. R. Commons in particular, the determining role of the institutionalists. Work dedicated to Industrial Relations  and then to the segmentation of the labor market falls into this same vein of inspiration over the course of the years that followed.
2This article is devoted to the contribution of this American specialty to economic sociology. The analysis can clearly only be partial, so abundant is the literature that, from the 1920s up to recent years, has emerged from this institutionalist tradition. For this reason, my area of study will be doubly restricted. I will reach first into the arena of labor relations only,  a field of study sufficiently broad to encourage reflection. I will then accord special, but not exclusive, attention to three authors – J. R. Commons, S. Perlman, and J. T. Dunlop – whose works seem emblematic of this ongoing articulation between history, economics, and sociology. Their research is of all the more interest in that they, unusually, enact a theoretical gesture in a universe in which the accumulation of empirical data predominates. I intend to follow several pieces of the thread that leads from the German historical school to the industrial relations movement in order to better identify the original place of this American tradition within the galaxy of what P. A. Hall and C. R. Taylor (1997) call “historical institutionalism.” 
3I also wish to underscore how this diachronic concern helped labor relations specialists uniquely distinguish themselves from the classical and the neoclassical economic models. However – and this will be at the core of my proposition – recourse to history does not run seamlessly from one author to the next. There are at least three different usages that I would like to highlight: the first gives priority to a hierarchy of factors (J. R. Commons), the second to a combination of factors (S. Perlman), and the last to a series of factors (J. T. Dunlop).  As original as they are, these approaches to labor relations are far from completely satisfying, not by reason of their analytical diversity, but quite simply because they are weighed down, in particular for S. Perlman and J. T. Dunlop, by ethnocentric presuppositions. Someone who wishes to evaluate the originality of an assertion that tightly interlaces sociology and economics may nonetheless profit from these unorthodox attempts. The last section of this contribution makes this argument and suggests that, beyond the difficulties of historicism, work on industrial relations has opened multiple tracks for research and paths for interpretation that are still relevant today.
1 – The Atlantic Passage
4In his history of American industrial relations, B. Kauffman (1993) shows that the institutionalist economy that emerged in the 1920s in the United States is distinct from classical and neoclassical theories by virtue of three elementary postulates. First, institutionalists refuse to analyze the labor market, arming themselves with a competitive model in order to better understand the numerous specificities aided by an empirical and historical stance. Then the importance of the power of the monopoly of firms in the “exploitation” of consumers and employees should be emphasized. Institutionalists lastly hold the conviction that laissez-faire policies are antinomic with individual liberty, provoke conflicts of interest, and are inefficient economically (monopolies that certain businesses enjoy have as a consequence low salaries and high prices, as well as factors that then lead to stagnation and unemployment). As I will now suggest, these three basic propositions are the product of the application to a North American reality of analytical and methodological precepts issuing from the German historical school.
1.1 – Deutsche Ursprungen
5The link between the German historical school and the American institutionalist tradition is today clearly established. It should be recalled that for the United States, the 1870s and 1880s constitute an important turning point to the extent that, based on economic depression and social movements (strikes, demonstrations, the success of proselytizing by the Knights of Labor), they registered the growing dissatisfaction of young students – John Bates Clark, Henry Carter Adams, Richard T. Ely, Edmund J. James, Simon Patten, Edwin R. A. Seligman, et al. – with the dominant economic theories. The intellectual alternatives that these sons of good families steeped in Protestant culture (with the exception of E. Seligman) hoped to promote are directly indebted to the teachings from which they benefited on trips to Germany, privileged moments in which they soaked in the theses of the historical school. Upon returning to their native country, these young American researchers were quite decided in developing a point of view that borrows largely from the Germanic approach. The organizational declension of the project took form in the American Economic Association, which was brought to the baptismal fount in 1885. Founded thanks to a decisive impulse by R. Ely (1886a), this association wanted to be a place open to the winds of change and, in opposition to the “laissez-faire”-type approaches, it promoted inductive, historical, and statistical investigative methods (Hamilton, 1919). In 1918, the charter of the association explicitly stipulated the goal: “Institutions, and no longer value, will henceforth be the object of the economic science. The word Institutionalism is thus proposed for the first time to designate the school of economic thought which will shape all those who accept this charter” (Guéry, 2001, 19).
6Richard T. Ely (1854-1943) is one of the fathers of this institutionalist economics. To the extent that his work directly influenced the development of “Industrial Relations,” it may be helpful to provide a quick survey of his trajectory and major theses.  After his studies at Columbia, Ely spent three years in Heidelberg, where he earned a doctorate (1879). Returning to the United States, he contributed to the diffusion of the theses of the German historical school, which further led him to become involved in an American Methodenstreit with his colleague S. Newcomb, a mathematician, physicist, and astronomer early converted to the marginalist point of view. A teacher in the history and politics department of John Hopkins University (beginning in 1881), then at the University of Wisconsin (dating from 1892), Ely differentiated, without great originality two eras of political economy: that of the English school (Malthus, Ricardo, Mill) and that of the German school. The first, an orthodox one, is according to him characterized by a deductivist approach, a propensity for reasoning based on the simple principle of Selfishness, an absence of statistical and historical references, a validation of laissez-faire, a tendency to cut economic phenomena from other social facts, the weakness of services it could render for practical action, the lack of realism of its theses (like that, for example, of the equalization of profit rates), and so on. Ely placed himself in the shadow of works of the second school, that which embodies the names of B. Hildebrand, C. Knies, W. Roscher, and others. “They took the name Historical School, in order to ally themselves with the great reformers in Politics, in Jurisprudence, and in Theology. They studied the present in the light of the past. They adopted experience as a guide, and judged of what was to come by what had been. Their method could also be called experimental. It is in many respects the same which has borne such excellent fruit in physical science” (Ely, 1883, 233). The experience and experimentation in question here take shape thanks to history, statistical treatment, and the comparative method. A double demand is thus asserted: that of empiricism on the one hand – “the first thing is to gather facts” (Ely, 234) – and that of the rupture with a too narrowly individualistic conception of man on the other hand – “as Blackstone has it, ‘Man was formed for society’” (Ely, 234).
7Where does he stand, more precisely, on questions related to labor relations? For Ely (1886b), three factors explain why employees are in an inferior position on the labor market. The first lies in the imperfection of this particular market and the inequalities that result. The economist states the extent to which, empirically, the existence of employer coalitions can harm employees. Monopsonies and oligopsonies of employers or, quite simply, collusion between the latter (through professional associations in particular) are sources of inequalities. This concretely means discriminatory practices at the expense of minorities, immigrants, and women. These management styles are particularly constrained by the injustice that employee mobility is limited by the loss of seniority rights (and other associated rights) that a change in business could not fail to provoke. The second factor points to the fact that, in spite of political rights acquired in the nineteenth century by laborers, the employer-employee relationship remains largely marked by the seal of authoritarianism. This absence of industrial democracy is at the root of multiple problems: weak involvement of employees in their work, a tendency to abandon businesses where labor relations have become unbearable, labor conflicts, and so on. Following the example of J. R. Commons and S. Perlman, who also attach a great deal of importance to it, R. Ely maintains economic insecurity as the third determining factor of inequalities that construct the relationship between employers and employees. He particularly decries the ill effects of the commodification of labor which renders employees vulnerable to the widest range of hazards (goodwill of the employer, work accidents, illness, etc.).
1.2 – (Industrial) Democracy in America
8Faced with the blatant failure of the labor market to autonomously reabsorb the difficulties and iniquities demonstrated by R. Ely, and more generally by the institutionalist economists who were supporters of German-style historical science, what could be done? For R. Ely as for J. B. Clark and H. Adams, socialist hope rests on the advent of an industrial democracy for which the great cooperative enterprise could be the preferred site of growth. “Our time is more democratic than others because it is more Christian,” wrote R. Ely in an introduction to a series of three articles that appeared in 1887 in Harper’s New Monthly Magazine. Confident in the brotherhood of man, R. Ely believes that the development of large enterprises through shares would bring a solution to the traditional authoritarianism of heads of enterprise, since it brings along with it a principle of industrial democracy (Ely, 1887a). Better than others, indeed, businesses financed by stock market capital could develop their production capacities and encourage savings. Their structure encourages above all the delegation of different tasks of production, management, and so forth, to those skilled in them. One could thus put a limit on managerial authoritarianism and, in the cooperative model, envision the undertaking by the employees themselves of the management of their enterprises (Ely, 1887b).  R. Ely is, for all that, not completely naive. He knows the dangers and the limits of the large industrial firm. Conferring the functions of piloting and supervision on management is not sufficient in itself. Management could waste resources attributed to it and prove to be inefficient in the realization of tasks. In addition, nothing can ensure that decent working conditions will be reserved for employees. The propensity for monopolistic stockpiling is also a danger that still looms, with its set of dubious methods (producing at minimum costs to incite bankruptcy in the competition, then buying up the competitor, etc.) and wrongdoings against the employees concerned. This is why economic and social regulation of business should be subject to responsibility differentiated by which businesses participate in the industry sectors at constant, decreasing, or increasing yields. In the first two cases, it is possible to satisfy the competitive principle to the extent that the monopolistic temptation is weak. In the last case, on the other hand, following the example of the gas industry, intervention is necessary. “The economic space concerned with growing returns should be abandoned by business in favor of the federal government, the State alone, or a combination associating the State with its different subdivisions, at the top rank of which are municipalities” (Ely, 1887b, 262).
9Ely’s logic is a mixture of multiple, typically German ingredients: a theistic background, concern for accumulating empirical data, denunciation of the inequalities suffered by employees, engagement favoring a certain state-level intervention in the operation of economic affairs, and, finally, a typically historicist schema for analysis. Thanks to an alliance between Christianity and economic theory, the whole serves an ethical project of which the fight against natural and social inequalities is the one most tangible manifestation. The link to the historical school becomes even clearer when one remembers that social economy (die Volkswirtschaft) as conceived by position-holding socialists “does not consist simply in a production society. What is more important above all is knowing not how to produce as much as possible, but in knowing how man lives, to what extent economic activities realize moral ends in life, these postulates of justice, humanity, morality which are present in every human society” (F. von Schoenberg, Handbuch der politischen Œkonomie, 1st booklet, 15; quoted in E. Durkheim, 1975, 270). R. Ely argues similarly in favor of an economic approach in which social justice constitutes the preferred horizon.  But, one could easily agree, the analyses of R. Ely remain unsatisfactory for opposing the English school, which, whatever one thinks of it, has a much stronger murmur of theoretical inspiration than that of the American heterodoxes. Indignation and empirical bends do not suffice to build a theory of labor relations. Yet it is this weakness that “Industrial Relations” will now and always regret. In 1958, John T. Dunlop recognized it at once: “The field of Industrial Relations today may be described in the words of Julian Huxley: mountains of facts have been piled on the plains of human ignorance. .. the result is a glut of new material. Great piles of facts are lying around unutilised, or utilised only in an occasional and partial manner. Facts have outrun ideas. Integrating theory has lagged far behind expanding experience” (1958, rev. ed. 1970, VI). If it remains partially valid today, this statement nonetheless merits qualification. Within the corpus of institutionalist works, one finds the traces of original attempts at theoretical formalizations which are so desperately lacking in numerous empirical investigations.
10This applies thus first and foremost to the research of J. R. Commons.  The latter manifests first in the will to systematize, after the German economists and R. Ely, the historical steps that punctuated the future of articulations between market and labor relations. The analyses that lead J. R. Commons to the labor movement also take root in reflections devoted to the logic of the development of an economic system. Capitalism moves successively through the steps of commercial capitalism, employer capitalism, and then banker capitalism (Commons et al., 1918). In the second place, Commons places rules and their negotiations at the heart of economic relations, and in doing so he sketches a veritable theory of transactions (Bazzoli, 1999). The influence of S. and B. Webbs is explicit here. In the early 1900s, J. R. Commons read Industrial Democracy. In the manner of the English pair, he places rules at the center of his analysis in order to demonstrate that labor practice and collective action can lead to the production of standards and be able to weigh on work activities and stabilize conflicts of interest. What is more, dissatisfied by the individualist logic that bases the analysis of labor relations on the contract, J. R. Commons developed an institutionalist reflection that assures a place for the collective and for hierarchical organization. These two axes of theorization were borrowed in a different way by the researchers whom I will now turn to: S. Perlman and J. T. Dunlop.
2 – The Historical Sociology of “Industrial Relations”
11Before examining the manner in which S. Perlman and J. T. Dunlop attempt to systematize labor relations in a historical sociological framework, I will first consider the article J. R. Commons dedicates to American shoemakers. This contribution constitutes an important piece in the institutionalist tradition to the extent that, thanks to the consequent use of empirical materials, J. R. Commons formalizes links between the market and social interactions. This article thus wholly presages the reflections S. Perlman and J. T. Dunlop would then devote to the dynamic of labor relations.
2.1 – Social Implications of the Extension of the Market Sector
12The article that J. R. Commons dedicated to shoemakers in 1909 falls unequivocally into the German tradition. The author recognizes this willingly, as he compares his work to that of K. Marx, G. Schmoller, and K. Bücher. In the first place, J. R. Commons concedes the interest of understanding the world in terms of economic steps. However, rather than reason based on the mode of production and a theory of capital gains, he proposes giving priority to an explanation by the markets. Industrial evolution may indeed be read through the prism of incessant expansion of the space for the distribution of goods produced. Attention should be paid, as such, to the many disturbances that have come to distort the effectiveness and good intentions at the heart of the community of producers (producing poor quality goods, using prisoners or children for labor, turning to nonstriking employees, etc.), as many troubles have sparked the implementation of protective measures. As for G. Schmoller and K. Bücher, J. R. Commons is grateful for their having avoided the paths of abstraction, which, he adds, still does not excuse the weakness of their typologies.
13In order to empirically support his own analytical framework, J. R. Commons offers the case of the shoemakers. On the cutting edge of combats and conflicts that mark the industrialization of the United States, these shoe manufacturers comprised a class of men who were skilled, scattered geographically, perpetually threatened by industrial and commercial changes, but who, thanks to their organizational abilities, always knew how to parry the risks related to the growing extension of the market frontiers. Thus, as he indicates unequivocally in the subtitle of his article (“A sketch of Industrial Evolution”), it means using the study of a particular professional group, in the manner of researchers in the German historical school, as a pretext for identifying the stages of industrial evolution. The comparison is not forced. Even with an interposed translation, K. Bücher’s book (1893) dedicated to the evolution of industrial societies serves explicitly as a reference for J. R. Commons. The method used is simple: it consists of speculating that different types of shoemaker organizations – J. R. Commons distinguishes six, from the Company of Shoomakers (Boston, 1648) to the Boot and Shoe Workers’ Union (1895) – correspond to defined stages of industrial evolution. One also notices another typically German hypothesis, according to which the extension of the market sector is the true motor of evolution of industrial forms and, consequently, the types of regulation of labor relations. The method thus consists of giving priority to an explicative variable to reason out what one wishes to explain (an approach by a hierarchy of factors).
14I will not detail the multiple stages that J. R. Commons comments on at length in his seminal article. At the risk of such a digression, let us linger on the first step, which, in the mid-seventeenth century, led the shoemakers to regulate their profession (installation of a commission, a professional charter, a table of penalties, etc.) on the grounds that some of them produced poor quality merchandise. Regulations forbade the manufacture of shoes at the customer’s home in order to better promote supervision of the less competent cobblers by skilled artisans and, in short, to oust the former from the shoe labor market. Thus reconfiguring the market spaces for goods and labor is laden with social consequences: at the risk of being expelled from the profession if they departed from their quality standards, the least skilled shoemakers would end up working in a shop, no longer under the heel of a customer (who supplied location, materials, and food); functions otherwise separated (those of the merchant, master, and worker) would then become united into a single status. J. R. Commons also shows how the passage of a custom-order market from neighborhood to a much larger market participates directly in this evolution. The institution of new professional standards goes hand in hand with technical transformations (warehousing merchandise, sale in stores, development of shop work in addition to outside production). In sum, economic development has a strong impact on the social classes. Under cover of market, regulation, and technical mutations, the master artisan changes status to acquire that of a merchant and employer. Throughout his long exposition of the different stages that follow, J. R. Commons gives an account of a similar manner of evolving articulations between merchant configurations, labor relations, and links between producers and consumers. In the straight line of the German historical school, he thus offers an original analytical outline for the conditions of the emergence of American labor relations.
15What are the limits of this economic historicism? By integrating the figure of the consumer into the analytical landscape, J. R. Commons first discredited himself in the eyes of certain labor relations specialists, who reproached him for not relativizing the importance of social ties between employers and employees. But the major weakness of the J. R. Commons analysis lies mostly in ethnocentrism from his point of view. The axiological framework (validation of private property, professional mobility, etc.), that structures his proposal, largely borrowed from American idealism, results in a concept of collective action that may or may not be able to cross borders with complete relevance. In other words, “the unions will thus be one pressure group among others, which will stand, at first, against the commercial capitalism that reduces the small landowner to the status of facility renter, but which, in its fight against the excess of the market, acquires a vital social function, that of holding responsibility for representative democracy in the industry” (Gombin, 1972, 550).
2.2 – S. Perlman’s Model
16S. Perlman (1928, 1935) would certainly not dispute this manner of conceiving union activity. This is not at all surprising from a Polish immigrant who was a student of J. R. Commons and successor to his chair in political economics at the University of Wisconsin. Like his predecessor, S. Perlman bears the historicist imprint of the German school. For him as well, the history of unionism should be understood in terms of stages. In the case of the United States, he distinguishes the following periods: antimonopolism, cooperatism, socialism, and finally Gomperism, the last stage being the happy outcome of a haughty step toward economic cooperation. But the analyses of S. Perlman are not reducible to a simple, new application of the German institutionalist paradigm. As L. M. Tremblay (1965) suggests, the theses that S. Perlman develops in A Theory of the Labor Movement (1928) must also be read as a theoretical counterpoint to the positions taken up by Lenin in What is to be Done? (1901-1902). I will briefly revisit the terms of the debate. In his book, Lenin has in his line of sight a group of “economists” sheltered by the Russian social democratic party, a group that swore that economic action (turning first of all to strikes, in the example of those that formed in Moscow and St. Petersburg in 1895) was more beneficial for Russian socialism than the battle against tsarism. Lenin, it is known, stood poles apart from such a position. “Trade-unionism is but the ideological enslavement of workers by the bourgeoisie. This is why our objective, the objective of social democracy, is to combat spontaneity, to redirect the labor movement away from the habit of spontaneity of trade-unionism, to seek refuge under the wing of the bourgeoisie, to draw them under the wing of revolutionary social democracy” (Lenin, 1901-1902, rev. ed. 1978, 90).
17A former member of a revisionist Marxist movement, S. Perlman stands close to the “economists’” positions. His objective is to show that unions are not and should not be revolutionary forces that aim to reverse the dominant market system. A defender of a reasonable capitalist order, Perlman lays the foundation for a model of intelligibility of economic development that rests on the following axiom. Each social group is equipped with a different economic philosophy. To the extent that they are holders of or control the means of production, capitalists live on the premise of abundance. They accept economic risks, promote individualist values, and are supporters of “laissez-faire.” The workers (manualists), for their part, are marked by an aversion to economic risk by reason of an acute consciousness of the rarity that characterizes their social world. “The manualist consciousness of rarity is the product of two major causes, one which is internal to him, the other which is external. The typical manualist is conscious of his innate inability to seize the economic opportunities that present themselves in the complex modern world of changing affairs. He himself knows that he was not born to take risks and he does not have a sufficiently agile spirit to feel comfortable in the realm of uncertain games which characterize competitive commercialism. We must add to that a conviction that, for him, the world is configured around the principle of rarity, which engenders an institutional order of things in which the best opportunities are selectively reserved for property owners, capitalists, and other privileged groups” (Perlman, 1928, rev. ed. 1970, 239-240). Trespassing freely on J. R. Commons’s territory,  S. Perlman hopes that this consciousness of rarity pushes workers to react together, particularly through unions. Belief in the rarity of jobs is thus a factor in mobilization that leads unions to practice a policy of exclusivism toward persons considered strangers to their groups and to establish an integral collective control over jobs. There is finally a third social group – that of intellectuals – whose major characteristic is to perceive in an equally false and abstract manner that which could be the realities of labor. Such is the case for the communists in Russia, or for the movement led by F. Lasalle in Germany. It is clear that, for S. Perlman, the influence of intellectuals on the labor movement is necessarily harmful. 
18Thus, S. Perlman produces a model (approach by the combination of factors) that articulates three principal variables: the capitalist power of resistance or, if one prefers, the power of dominant social classes when they are no longer protected by the State; the degree of influence of intellectuals in society in general and on the labor movement in particular; finally, the degree of maturity of the union movement in regard to the jobs question. Table 1 proposes a basic formalization of the manner in which S. Perlman combines these variables in his 1928 masterwork.
An elementary modeling of The theory of the labor movement by Selig Perlman
An elementary modeling of The theory of the labor movement by Selig Perlman+ : strong, or positive toward the influence of intellectuals
– : weak or nonexistent, or harmful to the influence of intellectuals
19Despite its apparent simplicity, S. Perlman’s model poses several problems. The first part of The Theory of the Labor Movement first of all contains inexactitudes and imprecisions in the descriptions of different revolutions and configurations (Russian, German, English, and American) that the author delivers. For example, contrary to what S. Perlman maintains, intellectuals were often solicited at the initiative of workers, and it is consequently erroneous to systematically attribute to bourgeois intellectuals, and not to worker unions, faults that they did not commit. Similarly, whatever the country, unionism was never completely disconnected from the political field, and, consequently, job-conscious unionism does not correspond to any historical reality (Sturmthal, 1951; Gulick, Bers, 1953). We must add to this that, in Perlman’s writing, the term intellectual is vague to say the least, since it designates alternately “the nonmanual, the revolutionary, the policy to refer to, and finally, a program that is not defined by opposition to job-conscious unionism” (Gombin, 1972, 554).
20There is still another problem: under the cover of a more coherent and tight-knit theorization than his predecessors’ at the University of Wisconsin, does Perlman not, in the end, rationalize the action of the American Federation of Labor (AFL), which, with Samuel Gompers, from 1886 to 1932 raised the control of jobs to the rank of philosophy, all without seeking to reform society or take control of businesses? It is difficult to deny this assertion. This restriction of union action to a purely economic function that locks workers into a role of producers cannot consequently maintain a position as a general and universal intelligibility scheme. This is all the more true as American history testifies, before the AFL, to union attempts to conquer for the benefit of the workers rights other than economic (Tremblay, 1965). S. Perlman later returned to his own theory in the 1940s to connect the notion of the job conscious as no longer a specific and limited action but a whole form of worker activity (as the formation of an independent party).
2.3 – J. T. Dunlop and the dynamic of labor relations systems
21After S. Perlman, the second major theoretician who branded labor relations with the hope of bringing the discipline out of the empirical stalemate was incontestably J. T. Dunlop (1944, 1958). His analytical ambition was first and foremost comparative, since, as he notes in the preface of the 1970 revised edition of his Industrial Relations Systems, two major questions are worth asking: to what extent, first of all, can we hold that there is or is not a tendency in labor relations systems to converge on a single model? In what circumstances and to what extent, then, can we transfer practices and institutions from one country to another? Thus bearing witness to his concrete implications in the actions of industrial mediations, J. T. Dunlop does not renounce his action. His ambition is nonetheless well and truly theoretical, since J. T. Dunlop wishes to produce a set of concepts that might have a universal impact on all levels of American economy and also across borders. We know the terms of Dunlopian grammar. A system of labor relations is the interaction of three types of actors (workers and their representatives, employers and their organizations, political institutions) joined by a common ideology, which produces rules relative to workplaces and work communities and those under the most varied forms (agreements, statues, orders, decrees, regulations, awards, policies, practices, customs). Thus, the task of the theory is to explain why particular rules are established in a system given the labor relations and how and why they change in response to variations in the environment that affect this system (Dunlop, 1958, rev. ed. 1970, IX).
22The perspective thus laid out nonetheless obliges a historical detour, and it leads back to very German tones. The reason for the institutionalist bias is given thus by J. T. Dunlop: “The major characteristics of a national system of industrial relations are determined at an early stage of industrial development in a country, and in the absence of violent revolution on a large scale, a national system of industrial relations retains these traits despite later evolutions” (Dunlop, 1958, rev. ed. 1970, 307). For J. T. Dunlop, systems of labor relations in for example, Great Britain, the United States, Scandinavia, Australia, or even New Zealand are directly accountable to social configurations that prevailed before World War I. With a similar concern as S. Perlman for not falling into the most unbridled of historiographies, J. T. Dunlop surveys four modes of constructing labor relations around the world: the “communist” path (a determinant influence, as in France, of communist parties on the formation of worker unions), the “International Labour Organization” path (an importance accorded, as in the United States, to rules laid out on an international level), the “State as employer” path (the decisive impact of the State/employer on the rules of labor), and, finally, the “professionalizing” path (the role of professionals in the negotiation and mediation that go hand in hand with a growing appeasement of social relations).
23J. T. Dunlop does not only hold himself to this framework, but maintains that three parameters merit consideration to understand the rapidity and the manner in which each national system dove in in its own way. The first is political. According to J. T. Dunlop, the calendar of national, political, and industrial revolutions is determinant. If, historically, the national question took precedence over all those related to recognizing the labor movement, or if the labor movement were intimately associated with fights for independence, then the labor relations systems would be characterized by close relations between the State and the labor movement. This signifies also that if, as in France, Italy, or Japan, the labor movement had to fight ardently for political rights, the labor relations systems would bear stigmas. Its composition is largely in debt to the structuring of the political realm and furthermore, it is quite susceptible to organizations and public political effects. The gap between industrial development and institutional affirmation of union organizations is equally important: when such a recognition occurs – as in the Netherlands, Germany, Yugoslavia, Japan, or Italy – through abrupt and belated adjustments rather than gradual recognition, the labor relations system tends to be more centralized than others and is conferred less autonomy on lower levels (e.g., regionally), and it tends to organize itself around an industrial base (not by trade) and to accord a larger place to the State.
24The second parameter maintained by J. T. Dunlop is economic. In his 1958 work, he lists, albeit in a more allusive manner, a set of variables that explains the correlations between economic development and affirmation of labor relations: integration of the workforce into industrial wage systems, bureaucratization of employee organizations, emergence of management professionals, augmentation of employee qualifications, regulation of labor relations by political authorities, collectivization and growing institutional regulation, and, finally, development of rules relating to labor. The place of the elite is the last parameter maintained by J. T. Dunlop. It thus distinguishes three models. The first, the dynastic-feudal model, is borne by landed or commercial aristocracy. These elites are occupied with the past; they validate the nation-state, the family, the church, etc. In the second model – that of the bourgeoisie (middle class), in this case – the elite issue from social groups occupied essentially with commercial activities. In conflict with the old order, they uphold the principle of free enterprise. The intellectual-revolutionary elite are at the core of the last model. They are more supportive of the rapid industrialization movement of modern societies.
25Unlike S. Perlman, J. T. Dunlop does not combine the preceding three parameters to logically engender the configurations of labor relations. The design, which draws from an approach by a series of factors, is simultaneously more simple and more complex. It is first simple: a variation of the technical, economic, or other kind of environment engenders in an almost mechanical way a transformation of rules (system output). For example, the more important a group is, the more rules will be formalized. But the design is also more complex: logically, to the extent that environmental variables are numerous, the products of a system are necessarily uncertain and, in all cases, difficult, if not impossible, to anticipate. To get around the difficulty, J. T. Dunlop reasons parameter by parameter, and he pays particular attention to links between types of elite, methods of structuring labor relations, and rules related to labor.
26Though it is more sophisticated than that of S. Perlman, the attempt at formalization proposed by J. T. Dunlop also suffers from a number of imperfections, beginning with a strong ethnocentric bent that gnaws at the modeling. Here too, indeed, the temptation is great to validate what the author knows best, the North American system. “According to Dunlop, the only truly stable model of labor relations is the Anglo-American type, directed by an elite middle class. This means, in particular: ‘liberal’ relations (in the American sense of the term) between leaders and employees; independence of workers with regard to management or another authority outside of the enterprise; acceptance of unions by the elite; a plurality of unions in competition for representativeness; a variety of forms of union organization; financial independence of unions; union leadership truly rooted in workers, with a strong business base; tolerance by the elite with regard to conflict; and so on. In the list presented here, certain traits pose problems. Why a competition between unions for representativeness? This is too clearly a reference to the judicial procedure established by the Taft-Hartley law in the United States in order for the union to gain recognition, and this cannot be a general characteristic of the system” (Crouch, 1990, 305-306). A number of other criticisms surfaced (Reynaud, Eyraud, Paradeise, Saglio, 1990), particularly to mention the equally over-insistent influence of the Cold War atmosphere on an analytical design that rationalizes to excess the present state of political forces.
3 – Beyond the bounds of historicism
27Already weakened by their excessive propensity to hyper-empiricism, are American “Industrial Relations” condemned, most notably by reason of a recurrent ethnocentric theme, to analytical incompetence? It is true that J. R. Commons, the first, did not immediately gain acceptance from his colleagues. But “that Commons had little in immediate theoretical posterity does not signify that he had no influence. This is manifested in not only the reformist ideal of the New Deal of which he would form the intellectual basis, but also on the institutionalist movement: in forging, through his study on labor questions, the Wisconsin school, Commons founded an original branch of this movement” (Bazzoli, 1999, 39-40). If, for his part, S. Perlman was never convincing in his attempt to model the action of labor movements, he did open up numerous decisive arguments that later would feed reflections on collective action and economic regulations. S. Perlman explains, for instance, that in the United States, a country where class sentiments were always weak, forming a collective conscience, setting up a union, and refusing to fold to the injunctions of a merchant logic were not possible because of a strong identification by employees with their professional groups. Contrary to what some presume, the condition for collective action rests thus less on a view toward a sociopolitical project and more on a strong anchoring in community. Works by labor movement specialists, beginning with those by Webb, fully confirm such a postulate and at the same time shatter the false opposition between corporate tradition and revolutionary unionism (Thompson, 1963; Sewell, 1983; Segrestin, 1975). There is another example: S. Perlman supported the point of view that unionism would necessarily tend in industrialized nations to break from the strategies of radical contestation in the existing social order to assimilate progressively into it and contribute directly to its proper regulation. This intuition, which R. Dahrendorf (1957), to cite just one example, would revisit in order to find reason in the development of conflicts at the heart of industrial societies, today passes for a truism. However, such an assertion did not hold against any evidence from when its author took a risk in formulating it.
28Because they join a landscape in which labor relations have acquired a social and academic recognition and because, even indirectly, they also benefit from the structural-functionalist aura, the theses of J. T. Dunlop had a more immediate intellectual impact on his contemporaries than did those of J. Commons and S. Perlman a few years later. His theory, however, “met a curious intellectual fate. Its direct legacy is limited and it was never subject to any deep discussion. It is at the same time an obligatory reference and few expressions have been used so broadly and universally. This curious situation that a classic did not create a school is explained in part by the fact that the analysis and research retained the elements of the theory the most easy to assimilate and the most easy to adapt to different theoretical contexts.. .. On the other hand, some of the principal articulations of the theory were put under much less examination or were treated as if they were self-evident” (Reynaud, 1990, 7). It would be useless to state my agreement with such a diagnosis, since it is precisely one of those unknown parts of the Dunlopian œuvre upon which I wished to elaborate previously.
29There thus remains, despite this curious destiny, a double Dunlopian heritage that should not be underestimated. The first is the emphasis on articulations between technological, economic, political, and rule-producing environments. Some conjectures are certainly fragile,  and neo-Parsonian determinism of the whole design is more than contestable. But the arguments posed by J. T. Dunlop are often more subtle than is typically acknowledged. Dunlop thus explains that certain rules really depend on the economic context and others on the political context; that the smaller the unit of analysis, the greater the influence of the context on the analytical system; that a diversity of procedures can stem from a similar complex of procedural rules; that political homogeneity is stronger in a national space than in the economic and technological fields and that these are, consequently, the political variables that differentiate the strongest among the countries (which fully justifies the sectorial comparisons that Dunlop undertakes in chapters 5 and 6 of his work), and so on. The second legacy of J. T. Dunlop is the formation of a semantics that has served and that still serves labor relations specialists. This last point could seem minor, but it is not, as it is so true that without a common theoretical language, it is impossible to connect observation with interpretation (Putnam, 1962).
4 – Conclusion
30There is one final lesson to take away from this American detour. Contrary to certain dogged beliefs, the distance that institutionalism in general and “Industrial Relations” in particular keeps from the classic designs of the labor market is not initially explained by the refusal to consider the lack of realism itself of hypotheses advanced by neoclassical economic theory. This is because they operate on the idea that competition works at the expense of a single social group (employees) and that institutionalists sharpen this critique, and some among them proclaim themselves in favor of models of negotiation that could restore a social equilibrium that in theory, but in theory alone, the competition model as applied to the labor market could just as well produce.
31That is why, under German influence, the “Industrial Relations” program converges so well with that of the neorealist economy, for which the major objective is to renew the analysis of labor markets by building bridges between facts and theory as well as between economics, history, sociology, law, and so on (Kerr, 1983, 313). The involvement of American institutionalists in the areas of decision and political and economic regulation  is the logical expression of a concept of social sciences that is not limited by a purely abstract formalization of the world of men. The tribute to pay, however, is quite high since, less present than others in the academic field (notably in terms of publications), institutionalist economists struggle with imposing their concepts and methodologies. In the 1960s, what is more, “a measure which blurs the innovative nature of collective negotiations, the interest in neorealist works declined; similarly, the prestige of the Wisconsin school eroded with the recognition of unions in the national economic life and the fact that they cease to bring hopes and fears along with them” (Kerr, 307).
32This decade doubtless signals the end of a first period, in which “Industrial Relations” flourished with an abundance of historicist references imported from Europe. Reciprocation will not be late in coming because, thanks to theories of segmentation and to interactionist offerings on negotiations and professions, European countries will next have their turn to benefit from the product of economists’ and sociologists’ work which, in certainly greatly different ways, have been able to renew the neorealist program of American pioneers of “Industrial Relations.” Thus, cleansing institutionalism with the water of a historical sociology that is too often prisoner even today of ethnocentric notions, or too weakly secured in methodological bases, will not be necessary (Badie, 1992). This would be much clumsier, as “Industrial Relations” still brings with it conceptual baggage, such as rules and negotiation, the current good fortune of which we know.
As a field of studies situated at the intersection of diverse disciplines (economics and sociology of course, but also law, history, and psychology), Industrial Relations is the product, in the United States, of the multiple social problems engendered in workplaces since World War I. Real recognition and institutionalization of this academic subdiscipline did not take shape until quite late when, during the 1940s, work and studies on labor relations acquired a relative visibility. This translated into three major events: the implementation of schools and institutes for industrial relations in the major American universities (Berkeley, Chicago, Cornell, New York, Wisconsin, Yale, etc.), the creation in 1947 of a new professional association (the Industrial Relations Research Association), and, finally, the foundation of the first academic journal dedicated to this field of study: the Industrial and Labor Relations Review (the first issue appeared in October 1947) (Kaufman, 1993).
I use here in an equivalent manner the terms “industrial relations” and “labor relations” to designate either the interactions between actors (unions, employers, and public authorities first and foremost) or the discipline that specializes in the study of such interactions. In the latter case, I refer to “Industrial Relations.”
According to P. A. Hall and C. R. Taylor (1997, 472), historical institutionalism presents four major characteristics: a tendency to conceptualize the relationship between institutions and individuals, attention given to the asymmetries of power associated with the functioning and development of institutions, a conception of institutional development that favors historical trajectories, critical situations, and unforeseen consequences, and finally, concern for combining different types of factors to explain the links between institutions and political situations. On a number of points, as we will see, Industrial Relations explicitly reveals this type of approach.
Here I freely adapt the distinctions suggested by J. Gadrey, F. Jany-Catrice, and T. Ribault (1999) for their international comparison work. Analysis by hierarchy of factors attempts to account for a variable to be explained by ordering the respective importance of explicative variables. The second means of doing this (through the combination of factors) focuses attention not on variables but on their relations with each other. The ambition is to bring forth a system of interactions whose coherence could be imitated. The analysis by series of factors distinguishes variables to be explained and variables that explain, but, unlike hierarchical analysis, it gives the same heuristic consideration to the whole of the explanatory variables.
To follow in closer detail the trajectory of R. Ely as well as the sometimes complex trajectories of other economists who were driven, in their early careers, by a socialist spirit (J. B. Clark, for example, who would end up adopting a marginalist point of view), cf. D. Ross (1991, 106).
In R. Ely’s mind, industrial democracy does not signify the erasure of status and social distinctions. “When businesses become veritable cooperatives respectful of the labor factor, the captains of industry will not disappear. The hierarchy is compatible with the most perfect of democracies” (Ely, 1887b, 260).
P. H. Nau and P. Steiner (2002) demonstrated that the moral concern is a common trait of Durkheimian sociology and the German historical school. Despite some nonnegligible differences, E. Durkheim and G. Schmoller bring a common interest to the question of reforms and social justice. American institutionalism, and primarily R. Ely, fully places its work within in this same set of preoccupations.
Although he did not succumb to the temptation of radical empiricism, this student of R. Ely did not necessarily reject connecting theory with practice. He thus participated in broad inquiries into the labor movement and immigration, and he also had an influence in this position over American worker legislation. He was also a member of industrial commissions charged with fixing salary rates in case of litigation in a profession. In short, J. R. Commons was an avowed partisan of a reasonable and popular capitalism that knew how to arrange its place in unionism.
For J. R. Commons, too, capitalism must be blamed because “it has not offered, as yet, to labor that security of the job which it has offered to investors in the security of their investments” (Commons, 1921, rev. ed 1967, 8).
S. Perlman distinguishes three types of intellectual elite according to the nature of forces that each attribute to labor. For Marxists, labor signifies a force of production and revolution against capitalism. In the case here, the worker is perceived as a person equipped with a strong revolutionary clairvoyance, capable of reacting to improve his material living conditions and to secure a dictatorship of his class over the rest of society. For ethical intellectuals (Christian socialists, anarchists, populists, etc.), labor is a source of liberty. Here, the vision of the worker is that of an individual who possesses potential to participate in the process of industrial development but who is simultaneously conscious that as a person, all his rights are not yet fully respected or recognized. The efficiency intellectuals (Fabian society) believe that society passes progressively from a stage of disorder to a stage of order. The figure associated with this concept of the world is that of a worker who has abandoned all claims on his employment. According to this, employers would be sufficiently powerful to implement recruitment procedures that secure an optimal adjustment between a worker and a job post. In the three cases, concludes S. Perlman, one notes a disconnect between the vision of intellectuals and the realities of life at work.
I think in particular of the link, which J. T. Dunlop makes too tightly, between types of machines and rules of remuneration.
The first institutionalist played a determining role in the implementation of the New Deal as well as the construction of a US Welfare State. Their successors were equally active since they participated in the War Labor Board, for example, during the Second World War and, following the example of J. T. Dunlop, they occupied roles as mediators in the industrial world (Kerr, ibid., 308).