Management Prerogatives, c. 1886-1904 Jeffrey Haydu Department of Sociology
U.C. San Diego In The Critical Study of Work: Labor, Technology, and Global Production, Rick Baldoz et al., eds. (Temple University Press, 2001) Abstract
Studies of the labor process treat capitalists as individual vehicles for larger economic forces. In these studies, capitalists transform work by pursuing straightforward interests in response to competitive pressures and workplace conflict. I argue that employers' collective identities are also important ingredients in the development of the labor process. In the case of turn-of-the-century battles over craft control, the formation of a "business community" was important for both strategic and ideological reasons. Strategically, the extension of social and organizational ties among proprietary capitalists enhanced their capacity to impose open shops. Ideologically, the lines along which proprietors constructed social solidarities and boundaries were peculiarly favorable to strident assertions of managerial prerogatives.1
Do Capitalists Matter in the Capitalist Labor Process?
Collective Capacities, Group Interests, and
Management Prerogatives, c. 1886-1904
A common and early criticism of Harry Braverman's Labor and Monopoly Capital (Braverman 1974) was its neglect of workers. "Labor" appeared in Braverman's story, of course, but largely as a generic victim of capital's relentless quest for control. What critics found missing in Braverman's account was a dialectic between the labor process and its victim. They called for more attention to the impact of production relations on workers' ideology and action, and they insisted that workers' resistance to capital also shaped the development of the labor process (Thompson  reviews the debates). Sympathetic scholars soon filled the gap. Ethnographic and historical research has given us a richer sense of how the labor process influences labor consciousness and politics (Burawoy 1979; Vallas 1993; Hanagan 1980; Haydu 1988a). That research also restored workers' agency, celebrating the ways in which working-class action altered the technologies and control strategies deployed by management (Edwards 1979; Hyman 1980; Noble 1986)
As the 1998 conference on Braverman's legacy has demonstrated, this scholarly tradition is alive and well. Contemporary research investigates the subtleties of management control in new arenas, from software designers to part-timers and workfare recipients (O’Riain 1998; Whalley and Meiksins 1998; Krinsky 1998). And in exploring the interaction of the labor process with worker identities and resistance, scholars have broadened the scope of the field to include race, ethnicity, and gender, along with class (Webster 1998; Baldoz 1998; Heide 1998).
In this article I argue that students of the labor process owe employers the same kind of attention they have lavished on workers. Much as workers' solidarities and identities mediate their impact on the labor process, so do employers' ties with their peers. The character of those ties, I argue, shapes their policies vis-a-vis employees and modifies their ability to implement those policies. For both reasons, our understanding of the labor process requires that we treat employers as we treat workers: as agents and as members of larger communities.
This recommendation clearly applies to turn-of-the-century struggles over craft control, the same struggles that provided the basic story line for Braverman's theory of the labor process. My paper draws on recent research in business history and on various local histories of business elites to show how employer class formation mattered in these workplace transformations. It mattered for both collective capacities and for common definitions of employer goals. At a time when many employers were by no means settled on how they should deal with craft regulations, a realignment in their solidarities and identities favored a belligerent assertion of "management rights." And at a time when "capitalists" were mostly proprietary employers with limited individual resources, local class formation gave business interests the ability to defeat craft control. After developing these arguments about employer class formation and the labor process in the late 19th century, I will briefly illustrate their relevance to capitalists and models of management in the late 20th century. Capitalists and the Labor Process The strengths and weaknesses of labor process theory can be illustrated by two brief anecdotes. When machinists struck the Coe and Stanley Company in 1913, Bridgeport's Manufacturers' Association rallied behind one of its own. Secretary Bennett arranged to import and house scabs; the Executive Board agreed to cover Coe and Stanley's strike-breaking expenses; and Board members, recognizing the special difficulty of replacing highly skilled workers, increased their subsidy to the firm by $1.50 a day as a bonus for each toolmaker employed. By early 1914, Secretary Bennett was pleased to report Coe and Stanley "progressing with its work with practically no outside interference."2 Another dispute that year did not go as well for the Association. Members of the International Association of Machinists at American Graphophone raised the same demands as machinists at Coe and Stanley: a 50-hour week with no reduction in pay. In this case, however, management capitulated. Officials of the Manufacturers' Association protested that this concession "would have a very injurious effect on the interests of the other members," and they organized a special meeting with American Graphophone's manager to discourage this breach of open shop principles. Not even the added moral weight of guests from the New Jersey and New York branches of the National Metal Trades Association and from the American Anti-Boycott Association was enough to reverse the company's decision.3
The literature on the labor process illuminates some aspects of these two disputes. That literature has long been concerned with management control and occupational deskilling. These theoretical interests converge in historical studies of craft control in the late 19th and early 20th century. Employers' offensives at this time targeted both workers' skills and customary mechanisms of workplace control. Their victories paved the way for the dilution of labor and the elaboration of open shop management techniques (Montgomery 1979; Nelson 1975; Haydu 1988a). Coe and Stanley's troubles with toolmakers and employers' alarm over American Graphophone yielding to a union clearly fit this model. Graphophone's heresy can also be accommodated by more recent contributions to the labor process literature. Management strategies now tend to be thought of as more varied and less sure to succeed than Braverman imagined (Littler 1990; Friedman 1990; Guillén 1994).
Where the literature on the labor process is less helpful for understanding these two disputes is in its tendency to treat employers in the singular rather than the plural. Management strategies are many (Friedman 1977; Burawoy 1985), of course, as are managerial factions. These differences may, in turn, be linked to diverse external influences, such as professional ideologies (Burawoy 1979, 183) or "elite mentalities" (Guillén 1994, 22-27) The unit of analysis, however, typically remains the single firm. Contributors to the 1998 conference often refer to "employers," as in "employers perpetuate workers' conditions of dependence" (Chun 1998, 65). But the plural form does not signify employers as a group so much as employers in the aggregate. Neither in case studies nor in surveys of capitalist labor policy does this aggregate shape employers' capacities or dispositions. Instead, it reflects the assumption that individual employers respond to external market forces or internal labor relations needs in similar ways. Indeed, nothing is lost by substituting for "employers" the abstract "capitalist," whose behavior can be explained in terms of systemic pressures and imputed interests. One conference paper reminds us that "what is of interest to the capitalist . . . is the amount of labour time taken to produce each unit of value." The necessity of reducing labor time is what "drive[s] and structure[s] the character of the capitalist labour process at all stages of its development" (Cohen 1998, 93, 94, original emphasis).
This asymmetry in studies of the labor process -- agency and group dynamics for workers, stick figures and atomization for employers -- has many sources. The field attracted scholars seeking to understand divisions and quiescence among workers, not employers. Lembcke's recent and welcome proposal to reintegrate labor history with studies of the labor process, for example, presents joint organization and collective capacity as problems for workers (Lembcke 1995). For employers, by contrast, neither power nor goals have seemed problematic. American intellectuals in particular, surveying post-war industrial relations, find little reason to doubt capital's ability to organize the labor process and exclude unions from workplace control. And while different management strategies for controlling the labor process did become a major focus of the literature, employers' underlying interests are still either taken for granted or deemed irrelevant, given the constraints of the market. What varies are the problems facing employers and thus the strategies they adopt to exert control and extract surplus -- not employers' own understandings of their priorities. Two more examples from the 1998 Braverman conference illustrate this logic. James Rinehart invokes the external forces of "heightened international competition, declining profits, and fiscal crises" to explain management's shift from Quality of Work Life reforms in the 1970s to an emphasis on teamwork and participation in the 1980s (Rinehart 1998, 337). And Peter Whalley and Peter Meiksins analyze the dilemmas created by different types of employees (technical professionals working either as independent contractors or as on-site part-timers) to explain differences in employer control strategies (Whalley and Meiksins 1998). In both cases, the explanation can move directly from problems to employer policies because each employer's underlying interests and goals are known and constant, accepted as givens within the capitalist system.4
Whatever the reasons for this skewed analysis, however, labor process theory obscures the formidable influence of the business community on individual employers. Coe and Stanley could count on the support of competitors as well as managers in other industries. An even broader coalition of local, regional, and national employers sought to steer American Graphophone towards particular industrial relations objectives. If these employers were wage earners, their encompassing solidarities, joint organization, and shared understanding of industrial conflict would surely be labeled "class consciousness."5 It is the role of employers' class formation in battles over craft skills and craft unions that is largely missing from debates on the labor process.
In the remainder of this paper, I offer a preliminary exploration of two links between employer class formation and the labor process. In the first section, I examine collective capacities. My focus will be not on large corporations and their managers but on the smaller firms and proprietors that were most strident in their attacks on craft control and union interference around the turn of the century. These employers did not necessarily have the clout to triumph over craft customs and unions.6 The obvious alternative of cooperation with other employers to control work practices, labor markets, and employee representation was easily undermined by competition or mutual indifference. The cultural practices and collective organization associated with employer class formation helped overcome these obstacles. In the second section, I turn to the ideological side of employers' offensive against craft regulation. Class formation for employers, as for workers, involved the construction of both differences (from other classes) and similarities (among members of a putatively common community). The ideological lines along which employers drew distinctions and extended solidarities, I argue, were also peculiarly favorable to belligerent assertions of management prerogatives. In turn, this common rallying cry reinforced employers' collective capacities. Contemporary class alliances and ideologies of work, I will suggest, may be analyzed in similar terms. Employers' Class Formation and Collective Capacities Craft Control and the Problem of Collective Action among Employers One of the most common criticisms of Braverman is that he overestimated management's ability to have its way with wage labor. Through formal organization and informal work group practices, employees often thwart capital's drive to separate conception from execution. With or without the boss's knowledge, workers also may exert control over the methods, pace, and remuneration of production tasks, even when those issues have been declared management prerogatives or officially settled off the shop floor by corporate executives and union bureaucrats (Burawoy 1979; Zetka 1992; Zuboff 1984). The political effects of this stubborn attachment to customary job controls vary and may (as Burawoy argues) even benefit employers. It remains the case, however, that these controls are not easily or inevitably overcome by management.
This general criticism of Braverman applies to turn-of-the-century as well as more recent conflicts over craft control. Much like Braverman's work, early surveys in the new labor history highlighted pivotal battles in which large firms like Carnegie or Ford vanquished craft unions and revolutionized the labor process (Brody 1960; Nelson 1975; Meyer 1981). More recent historiography, however, has found these epics doubly misleading. First, they obscure the persistence of craft practices in hat-making shops, stove foundries, pottery plants, construction sites, and textile factories, among others (Bensman 1980; Cebula 1976; Stern 1994; Christie 1956; Scranton 1983). Even in the heart of the second industrial revolution and the open shop -- early 20th century Detroit -- craft practices persisted in smaller metal-working shops, and many employers quietly accommodated union representation for their skilled employees (Klug 1993, 762, 767-768). Ford and Carnegie are misleading for another reason. Most employers lacked the financial resources and strength in product and labor markets to single-handedly abolish craft work practices and purge unions. The typical firm in the late 19th century was still a small one, owned and managed by a single proprietor or a few partners (Bruchey 1980; Roy 1997). More than 80% of all manufacturing firms in 1880 had fewer than 50 employees, and in most cities these modest establishments accounted for a clear majority of the workforce (Griffen and Griffen 1980, 124). Such firms got the better of craftsmen if and when employers acted together -- honoring the same blacklists, establishing vocational training programs to replace union apprenticeships, fusing political resources to limit union influence in municipal government, and maintaining a united front against strikes (Ramirez 1978; Harris 1991; Cohen 1990; Klug 1993).7
Neither employers nor business historians needed game theory to recognize the obstacles to such cooperative action (Akard 1992; Bowman 1989). Common organization and collective action among firms could be every bit as problematic as solidarity among workers, but for rather different reasons. In contrast to wage earners, employers were not usually divided by ethnicity or religion in late 19th century America: most were native born and high Protestant (Zunz 1982, 204-206; Folsom 1981, 77; Ingham 1978). Nor did personal hardship drive them to put immediate needs ahead of collective interests. Still, cooperative action often proved difficult. In any given industry, employers might profit by not coming to the rescue of competitors under siege from union demands or strikes. If one firm made concessions to unions, others gained a competitive advantage in production costs; when one company suffered a strike, rivals could poach market share. At various times, employers did make agreements to curb "ruinous competition" by standardizing labor costs and conditions, but these pacts were vulnerable to free riders and defection (Bowman 1989; Haydu 1988c). Managers of railroads in the Chicago area, for example, organized in 1886 to combat railway union demands. Rival lines, however, spurned the General Managers' Association's efforts to cooperate on a wider scale (Stromquist 1987, 251-252; McMurry 1953).
Cooperation among employers in different economic sectors was not sabotaged by direct competition. Employer unity across trades, however, was no more "natural" than solidarity among workers. The uneven development of markets and the labor process put employers in very different strategic positions. Shop owners who still worked at the trade might have little in common with full-time managers; custom and large-batch producers did not have the same labor needs; and firms manufacturing for local markets experienced different opportunities and constraints than did those that sold their wares nationally. Even among employers in closely related industries -- stove and general foundries, for example, or general and "sanitary" pottery -- differences in product markets and ownership patterns could rule out common action (Barnett 1912, 441-442; Stern 1994).
Whether their relations were competitive or merely distant, two additional considerations made employers reluctant to take part in collective organization. One concerned proprietary rights. Employers who inveighed against union attempts to "run their businesses" were little more willing to cede authority to employers' associations. Detroit managers, for example, often declined to share personnel information with the Employers' Association's Labor Bureau or even to establish administrative mechanisms for gathering this information (Klug 1993, 792-793, 802-806). Walter Drew of the National Erectors' Association added that collective action against labor was bad politics. Arguing against proposals to establish a national federation of open shop employers after World War I, Drew warned that such "class organization" might stimulate workers to follow suit (Fine 1995, 202). Solidarity and Organization among Employers Despite these obstacles, formal organization to deal with labor spread rapidly from the mid-1880s (Bonnett 1922, 1956). It was often a common threat from the labor movement that first drove employers to organize or to adapt existing trade associations to industrial relations tasks. Reflecting both the speed of mobilization and its roots in confrontation with the Knights of Labor, employers' associations accounted for 75% of lockouts in 1884-86, compared to 30% in 1881-1883 (Bonnett 1922, 22). Many of these early combinations faltered once labor insurgency ebbed. By the turn of the century, however, employer organizations assumed a more permanent footing and more regularized functions, including local employment services, data gathering, and industrial boosterism, as well as ad hoc strike-fighting assistance.8 These organizations also commonly exacted substantial dues and even the posting of bonds to guarantee compliance with association policies.
The fact that employers' associations spread, institutionalized, and acquired significant authority over member firms even in the absence of immediate labor challenges suggests that employer solidarity had deeper roots than the contingencies of class struggle. When labor historians seek to explain worker solidarities and divisions, the usual suspects are social networks and cultural practices rooted in residence, leisure, and politics (Faler 1981; Rosenzweig 1983; Wilentz 1984). Students of the middle classes (Blumin 1989; Gilkeson 1986) and upper classes (Baltzell 1964; Jaher 1982; Domhoff 1974; Roy 1991) classes have followed their lead, substituting for pubs, mechanics' institutes, and ward politics the elevated realms of social clubs, prep schools, and civic reform associations. Through such institutions, local upper classes developed common codes of behavior, regulated entry into both business and social circles, and stigmatized the rabble. And most studies highlight the 1880s and 1890s as the key decades in which local employers, bankers, and professionals actively differentiated themselves as a distinct class, withdrawing into separate neighborhoods, cultivating the fine arts, organizing country clubs, and steering their children into the right schools and colleges (Gilkeson 1986, ch. 4; Couvares 1984; Ingham 1978; Eggert 1993). Naturally, the character and timing of class formation varied. The three major metropolises (Boston, New York, and Philadelphia), for example, displayed more social differentiation between older elites and an emergent middle class than was common in newer, middle-sized manufacturing cities such as Pittsburgh, Detroit, or Cleveland (Jaher 1982).
Personal ties and status conventions developed in elite schools, private clubs, and local charities may have smoothed the way for joint organization among men otherwise divided by business competition or by the industry-specific concerns of their workaday lives (Rosenzweig 1983, 14-15; Roy 1991).9 Case studies of urban elites during this period point to an overlap in the membership of cultural institutions and business associations. Members of San Francisco's Industrial Association, for example, came from varied business backgrounds but shared service with the local art association and Boys' Club; most activists in Detroit's Employers' Association, similarly, belonged to the same high Protestant churches and Republican party organization (Issel and Cherny 1986, 96; Klug 1993, 745-752). And the development of common cultural institutions and social identities among businessmen, professionals, and the upper crust of the white-collar workforce, in turn, had clear implications for employers' collective capacities. As Gutman emphasized (Gutman 1977), it became more likely that employers could count on "community" support and municipal government backing in their battles with labor.
Under these conditions, the trend was toward more inclusive organization for dealing with labor. Often, early employer associations were local branches of industry-specific bodies like the National Founders' Association, National Metal Trades Association, United Typothetae, and National Erectors' Association. By the early 1900s, however, a growing number of citywide employers' associations combined representatives from different trades (Jackson 1984; Cohen 1990; Harris 1991; Klug 1993). Even without formal peak organizations, sectional bodies were increasingly linked by a dense network of interlocking memberships, mutual endorsements, circulating newsletters, and rotating guest speakers (Bonnett 1922, 362-377, 551-552; Cohen 1990, 170). Employers' Class Formation and Collective Interests Employer class formation was important not only for the balance of power in labor relations, but also for employers' understandings of their interests and rights at work. A variety of competitive pressures and technological opportunities led employers to attack craft practices and unions in the late 19th and early 20th century. The ideological terrain of class formation, however, may have helped turn practical considerations into uncompromising moral principles.
In their dealings with craft control and unions, late 19th century employers pursued different policies and justified them in different ways. Alongside the familiar pattern of deskilling, textile proprietors in Philadelphia and specialty steel manufacturers in Pittsburgh treated handicraft skills as a resource for innovation and "flexibility" rather than as obstacles to management control (Ingham 1991). And amid movements for the open shop, stove founders, coal mine owners, and pottery manufacturers relied on unions to help stabilize their industries rather than driving them from the workplace (Klug 1993; Bowman 1989; Stern 1994; Haydu 1988b).
The range of strategies for organizing work and labor relations can be explained at least in part by the constraints that technology, markets, and ownership patterns imposed on employers. But this diversity also suggests that individual capitalists had more discretion in their approach to unions and workplace control than labor process theory would allow.10 The convergence of employer policies vis-a-vis craft regulation towards the turn of the century also suggests that there were forces at work other than competitive pressures and capital's unvarying quest for control. Businessmen who had seemingly taken a more pragmatic stance with regard to craft unions now insisted that they would in no way compromise their management prerogatives. And at least in their public statements, this antiunion consensus brought together employers in very different economic circumstances. To explain this shift, we should take our cue not from studies of the labor process but from reinterpretations of labor politics. Those accounts have emphasized the links between social networks, collective identities, and the framing of group interests (Melluci 1989; Berlanstein 1993; Hall 1997). We might thus ask if employers' definitions of proper labor policies (as well as their capacity to implement them) also reflect the "imagined communities" to which they belong.