Chapter Three
In Defense of the Trade:
From Local Struggles to National Settlements, 1890–1901.

Skilled workers' responses to industrial change during the 1890s were largely defensive and sectional. Engineers and machinists sought to protect their exclusive privileges by blocking dilution and banning piecework, and they relied on their craft power and sectional strategies to enforce work rules.

Features of the labor process favored the mobilization of protest along defensive and exclusive lines. Skilled men still occupied an important place in manufacturing processes. Their position at work permitted effective action on an independent basis, with no need to compromise sectional interests. The organizational vehicles for protest reinforced this outcome. Rank-and-file protest in both Britain and the United States developed in a context of unsupportive national unions and localistic and informal industrial relations. Machinists and engineers, accordingly, relied on district and shop-floor organization to pursue their interests. Before 1900, these local bodies were constituted along narrow craft lines, and this too favored the mobilization of exclusive goals and disputes.

Local autonomy in defending craft privileges created problems for union leaders and employers. The problems for union executives included internal discipline and obstacles to the organization's recognition as a responsible bargaining partner. For employers, lo-

cally based craft protest meant opposition to managerial innovation, the inconveniences of sporadic strikes, and competitive inequalities in labor conditions from one shop or city to another. These irritations led in both countries to national confrontations—the 1897–1898 engineering lockout and the 1900 and 1901 machinists' strikes. The initial outcome in each case was a national trade agreement designed to resolve the dilemmas of employers and union leaders alike. In neither industry did these agreements meet employers' ambitions. Yet Britain's Terms of Agreement endured, whereas American metal trades employers turned instead to the open shop after 1901. This divergence would provide the basis for subsequent contrasts in the development of factory politics, creating a new framework for the selective mobilization of craft grievances.

Local organization played a crucial role in the factory politics of the 1890s. District and shop committees, organized on a sectional basis, channeled protest along exclusive lines. They also resisted the authority of both employers and union officials, resistance that contributed to national disputes. The character and importance of local organization, in turn, may be traced to the union and industrial relations conditions under which craftsmen pursued their interests.

The Context of Protest in the 1890s

Unions

In their efforts to defend traditional rights and practices at work, engineers and machinists of the 1890s found their own union executives at times unhelpful and at times actively hostile. Neither the International Association of Machinists (IAM) nor the Amalgamated Society of Engineers (ASE) had firm, coherent, national policies to meet management offensives against craft control. The IAM constitution prohibited members from working two machines or accepting piecework. But the union executives did not specify whether new machines were to be run exclusively by unionists, by apprenticed men, by "skilled" operators (e.g., those who had worked four years in the trade, whether apprenticed or not), or by those in receipt of the standard craft rate (which varied from one

locality to another). Nor did ASE leaders provide such guidelines. The IAM had no rules regulating the operation of payment by results when such systems were in force; and after 1892, the ASE explicitly left control of piecework issues to district committees. In these matters, and even more so with regard to contentions over workshop discipline, the unions involved themselves in shop-floor struggles only after local negotiations broke down or strikes occurred.^[1]

Even at such times, however, union executives often proved reluctant to sanction strikes or back local struggles over control issues. For two years after the IAM's founding in 1889, its constitution formally eschewed strikes, advising members "to act in a manner becoming men of our calling, and treat with our employers as gentlemen in making known our wants, and we are satisfied we will receive the proper consideration."^[2] Even after constitutional changes permitted strikes, union officials remained cautious. IAM leaders, like their ASE counterparts, feared not only for the union treasury: excessive resort to strikes, in their view, could also endanger the union's respectability and its prospects for recognition and collective bargaining rights. These considerations appeared particularly pressing when periods of high unemployment strained union budgets and diminished the chances for victory.^[3] Even if strike calls eventually won approval, union rules required that local and then national officials (who might have little familiarity with conditions in individual shops) review grievances and attempt negotiations. By the time constitutional procedures for declaring a strike were complete, the grievance may have smoldered for some weeks.^[4]

But union leaders were not indiscriminately cautious. It was precisely over grievances concerning control issues that they exercised the greatest restraint. They had good reasons to do so. Compared to demands for shorter hours or higher wages, struggles against dilution or piecework represented a risky investment of strike benefits. A local demand for a wage increase would benefit all members; as a result, their loyalty during a strike was likely. Union representatives could more easily justify the demand in public on the grounds of an improved standard of living and more easily argue the case in negotiations on the grounds of equity with employees elsewhere and the employer's ability to pay. On the em-

ployer's side, wage demands did not threaten sacred management rights and could be offset with changes in work practices. At the least, compromise was always possible. The combination of broad member support and employers' willingness to compromise gave strikes for wages or hours a practical appeal to union leaders. By contrast, control issues tended to be sectional in appeal, illegitimate in employers' eyes, and harder to negotiate. When employers introduced piecework in one department, for example, some members welcomed the opportunity to increase their earnings. Members in other departments and plants, for the moment secure, might be reluctant to strike in sympathy. A ban on piecework challenged employers' freedom to manage their own business and pay their workers according to individual merit. They could generally count on favorable press coverage and public sympathy in their resistance to such trade union tyranny. On the question of whether or not to introduce piecework, finally, workers and managers could scarcely split the difference.^[5]

Union leaders did not make these calculations in a vacuum. They found them amply confirmed by the strike statistics of their own organizations. At the 1897 IAM convention, President O'Connell reviewed the most important strikes over the two previous years. Four strikes had concerned wages, and all had resulted in compromise. Four others had involved control issues—one for the closed shop, and one each against dilution, piecework, and an obnoxious superintendent. Of these four, three had been lost.^[6] The policy implications seemed clear to O'Connell. During the convention and throughout the late 1890s, he argued that the spread of piecework could not be halted; members should instead make concessions regarding piecework in exchange for the nine-hour day.^[7] ASE leaders reached similar conclusions respecting dilution, urging attention to the earnings rather than the specific qualifications of machine operatives.^[8]

Industrial Relations

In their defense of craft control, engineers and machinists received from national leaders only loose guidance and support that often was too little and too late. These weaknesses in union policy contributed to and reinforced a system of industrial relations charac-

terized by localism and informality. Even in well-organized American cities negotiations between machinists and employers took place on a firm-by-firm basis and concerned issues peculiar to the individual shop. By the 1890s, districtwide agreements on wages and hours were common in Britain, but questions of control remained subject to plant-level bargaining.^[9] Such bargaining was in both countries relatively informal. Custom rather than contracts ratified craft control. Employers who infringed on these standards would be visited not by a union official but by ad hoc deputations of their own employees.^[10] Failing an amicable settlement of the dispute, the workers might turn to guerrilla tactics (restriction of output, refusal to instruct dilutees, or simply quitting) to press their claims.^[11] Union officials, particularly national ones, entered the fray only as a last resort. And settlements, if any, usually took the form of verbal understandings or, especially in the United States, the posting of new work rules without explicit reference to any agreement between management and the men or their unions.^[12] The industrial relations of the 1890s, like union policies, thus made the battle against employer policies a responsibility of union locals and, above all, the workers in the shops.

Britain

Factory Politics up to the 1897–1898 Lockout

How did engineers respond to employer attacks on their customary prerogatives and standards? In terms of goals, most engineers directed their energies to defending the craft status quo. In terms of tactics, they relied on sectional action and local organization. Conservative goals and sectional action reflect the continued importance of skill in production. Craftsmen acting alone still had great bargaining power and, having little need for allies, they faced no pressure to compromise their exclusive interests. District and shop-floor organization filled gaps left by union policies and conformed to the local and informal nature of industrial relations. By reproducing the exclusiveness of trade unions at the shop level and enhancing the strength of skilled employees at work, rank-and-file organization also channeled protest toward defensive goals and sectional action.

Engineers countered management encroachments on craft standards with work rules specifying conditions under which unionists would accept employment. These rules varied from one district to another. In well-organized centers ASE district committees ruled that only apprenticed engineers could do work undertaken in the past by engineers. The rule banned not only unapprenticed workers but also members of other crafts from the performance of certain tasks (e.g., fitting) and operation of certain machines (e.g., lathes): disputes over demarcation were almost as common as those over "illegal men."^[13] Elsewhere district committees insisted that "skilled" work be done by men in receipt of the local craft rate rather than by those who had served their time; or they simply demanded that certain wage rates be attached to given classes of machines, regardless of who worked them.^[14] Although this strategy opened the door to unapprenticed workers, it still aimed to protect craft standards of pay against dilution. Other rules combating dilution included an outright ban on one worker running more than one machine and on instructing handymen or setting up jobs for less skilled operatives.^[15] If new semiautomatic machines did work formerly done by skilled men, engineers claimed those tools as their exclusive preserve, to be operated at the standard craft rates.^[16]

Strongly organized districts also refused to allow the introduction of payment by results. In weaker areas or in shops where piecework had existed for some time, rules stipulated the conditions under which unionists would accept the system. For example, rates had to be mutually agreed between foremen and workers and extra pay offered for overtime. Both to discourage the system and to prevent such abuses as rate cutting, some districts also limited members' earnings (e.g., to time and a third). Such restrictions were undoubtedly still more prevalent on an informal basis.^[17]

Engineers did not at first negotiate these defensive rules with employers. Although work rules often reflected traditional practices long accepted by management, they were unilateral in character, laying down the conditions under which ASE members would work. Bargaining and compromise played little role. An engineer asked to run two lathes was obligated to refuse, under penalty of fines or expulsion from the union. This was the first line of

defense against new shop practices. Many employers abandoned innovations rather than lose their skilled workers. Unrepentant employers faced strikes, picketing, and the blacking (boycott) of their work or products at other firms. By selecting one shop at a time for attack, district committees both conserved strike funds and discouraged the spread of objectionable practices to other companies.^[18]

These were exclusive policies aimed at defending engineers' status and jobs against all comers, including members of other trades and less skilled workers. Engineers opposed piecework for themselves, but rarely engaged in sympathetic action to aid other workers faced with the same evil. The tactics with which these goals were pursued were no less sectional. Engineers' refusal to accept or retain employment in offensive shops relied on the indispensability of skilled workers—a strategy not available to less skilled workers. When engineers resorted to strikes, they received union benefits, and the ASE often did its best to support skilled nonmembers who otherwise might have acted as blacklegs (scabs). In most cases no such support was extended to the laborers inevitably thrown out of work during strikes; because laborers could rarely replace craft workers, unions could safely ignore them.

Engineers relied on local and workshop organization to defend craft privileges, and it is here that the influence of unions and industrial relations is clearest. The basic cell of the ASE was the union branch, composed of members residing within a certain area. The key agency of local policy, however, was the district committee. In each industrial center a district committee, elected by local members, took responsibility for regulating the trade. Because the ASE Executive failed to initiate trade policy, district committees formulated work rules and disciplined members who ignored them. Because employers were organized at most on a local basis, and industrial conditions varied, work rules differed from one district to another. District committees laid down the work rules, but organization within the shops carried them out. Well before the 1890s, active unionists on the shop floor assumed the tasks of enrolling new hands, reporting on working conditions to their branches, and presenting shared grievances to management. Gradually, these functions were formalized in the shop stewards' duties. Workers would select (informally or by vote) one person in a department to represent them in dealings with man-

agement. District committees, for their part, recognized shop stewards as union representatives in the works charged with recruiting new members, making sure old members were paid up, and reporting any encroachments on local work rules—by employers or by other unions. The 1896 ASE convention formally conferred on district committees the power to appoint shop stewards.^[19]

Shop steward organization helped local engineers overcome some liabilities of trade union structure and policy. Residentially based branches were ill equipped to deal with grievances shared by workers in the same shop who belonged to separate union branches. In any case the workers were often unwilling to wait for the weekly meeting to air their concerns and decide on appropriate action. They could press their demands more promptly and present a united front to employers through shop meetings and work-place organization. The shop steward system was also adapted to industrial relations in which conditions varied from one firm to another and in which initial negotiations took place informally between managers and employees. Under these conditions grievances often had to be handled on the spot, and the shop steward system developed to defend the workers' interests from case to case.^[20]

Trade unionism and industrial relations thus favored local organization in the conduct of craft protest. The character of local organization, in turn, ensured that factory politics would be defensive in goals and exclusive in action. Skilled engineers dominated the district committees, which determined trade policy, making these bodies the "hard core of workers' resistance to technological change."^[21] Shop stewards carried district committee authority inside the factory gates, bearing responsibility for defending local work rules. District committees, moreover, were organized along the same sectional lines as national trade unions, and shop stewards represented only members of their own unions. Engineers also organized on the job far more effectively than did less skilled workers. This collective advantage reinforced as well as reflected engineers' strategic position in the labor process. Technical and organizational advantages together enabled them, acting alone, to pursue goals that divided their interests from workers of other skills.

Despite their autonomy in handling industrial conflict, district

committees and shop steward representation thus reproduced the defensive priorities and exclusive organization of craft unionism at the local level. This institutional context inhibited innovative demands and broader solidarities in the mobilization of craft grievances. Joint committees did occasionally appear at work, and members of different unions did cooperate locally. Yet these movements—for the nine-hour day on the Tyne in 1871, for wage increases in London and Lancashire during 1873–1875, and for the eight-hour day in the 1890s-did not concern issues of craft privilege or workshop control.^[22] For the latter purposes local struggles remained defensive and exclusive.

The absence of radical factory politics among engineers by no means ensured harmony between local unionists and the ASE national executive. The Executive Council had no wish to see union funds and reputation dissipated by ill-considered strikes and resented the reluctance of districts to submit trade decisions and strike plans for executive approval, as required by the constitution. For their part, rank-and-file engineers chafed under the executive's moderation and the protracted constitutional procedures for reviewing disputes and approving strikes. Improved local organization gave engineers the means to take matters into their own hands. The results were friction over unconstitutional strikes and frequent criticism of national leaders for their excessive caution and insufficient support of local struggles.^[23] A lack of union discipline created problems for employers as well as for ASE executive officials.

The 1897–1898 Lockout and the Terms of Settlement

In the spring of 1897, a joint committee of engineering trade unions in London initiated a movement for the eight-hour day, and by the end of May more than one hundred firms had conceded.^[24] On June 5, however, a number of London employers (hitherto unorganized) announced the formation of a London branch of the Employers' Federation of Engineering Associations (soon renamed the Engineering Employers' Federation [EEF]). The joint committee refused to recognize the London association and served notice on individual recalcitrants that if they did not grant the forty-

eight-hour week by July 3 members would be withdrawn from their shops. The EEF responded in kind: if a strike took place, 25 percent of union members in federated firms throughout the country would be locked out each week, beginning July 13. This threat was carried out, prompting about seventeen thousand ASE members, along with several thousand from other unions, to leave work. By October the EEF led 579 firms in the lockout, with forty-five thousand craftsmen (twenty-two thousand from the ASE) out of work. After government intervention to bring the two sides together, and the rejection in December of two successive employer proposals, union executives withdrew the demand for an eight-hour day and recommended that their members accept the Terms of Settlement worked out between the Federation and the unions. In January 1898, unionists approved that recommendation by a vote of 28,588 to 13,927. By this time more than seven hundred firms had enforced the lockout against 47,500 skilled engineers.

The lockout clearly expressed the basic tendencies of factory politics during the 1890s, except in one respect: the conflict took place on a national scale. Engineers aimed to preserve customary craft controls at the expense of less skilled workers, and during the lockout cooperation among craft unions remained fragile. The disputes over workshop control that led to the lockout were conducted locally, and only the organization and tactics of the EEF shifted the battle to a national terrain. Finally, the lockout revealed tensions between a relatively accommodating union leadership and members committed to the defense of craft control at the expense of managerial prerogatives.

The lockout was the culmination of a period of intensifying friction over work practices, particularly dilution and piecework. In February 1897, for example, the EEF complained to the ASE of strikes over the manning of machines in Sunderland, Barrow, and Elswick and restrictions on overtime in Sunderland, Hartlepools, Clydebank, Belfast, and Barrow.^[25] These disputes sought to prevent employers from substituting less skilled workers on certain machines or to secure the craft rate for new machines regardless of the skill required to run them. The conduct of the dispute was no less exclusive. At the outset the Boilermakers' and Patternmakers' unions refused to take part. In Barrow initial cooperation among unions of craftsmen and laborers broke down during the lockout.^[26]

Local autonomy also played an important role. The strikes that precipitated the EEF's formation did not reflect any coherent national policy. Instead, they aimed to defend local work rules, which varied from one district to another. These strikes, like the demands in London, also continued the strategy of concentrating union resources on individual districts or employers. Such local independence highlighted a divergence in the priorities of national officials and union members. From the start of the lockout, ASE leaders sought to focus the dispute on the question of the eight-hour day rather than on issues of craft control. They did so in part with an eye to public relations. Yet, as the lockout wore on, union officials appeared more willing than the rank and file to trade workshop control for a shorter workweek. Unionists rejected the first settlement proposed by the EEF (declaring the employers' unfettered right to manage their shops without union interference and limiting even the right of unions to bargain collectively over wages) by 68,966 votes to 752. Shortly thereafter, union leaders, on their own initiative, placed before their members two ballot proposals. One offered the original terms, slightly modified in the union's favor; the other added to these terms a fifty-one-hour week (something never offered by employers). The proposals imply that union leaders were willing to accept management control over production in exchange for concessions on working hours and hoped that the ballot results would persuade employers to make that exchange. The union rank and file, however, rejected both proposals by huge majorities—thus demonstrating their own priorities.

Similar conclusions emerge from considering the employers' side of the dispute. The formation of local employers' associations in the north from 1894 on, and of the national federation in 1896, represented responses to the union strategy of picking off employers one by one or district by district. With the consolidation of local bodies in 1896, the EEF national treasury could be pitted against that of the ASE in any single dispute. Through national organization employers also sought to replace local autonomy in work rule formulation with industrywide standards.

Employer policies also demonstrate that craft control, not hours of work, was at stake. From the beginning of negotiations to the conclusion of the lockout, the EEF insisted on a basic principle: "The Federated employers.... will admit no interference with the

management of their business, and reserve to themselves the right to introduce into any federated workshop, at the option of the employer concerned, any condition of labour under which any of the trade unions here represented were working at the commencement of the dispute in any of the workshops of the federated employers."^[27] Although employers hardly expected unionists to accept this principle gracefully, their strategy and demands imply a recognition that union leaders would be more accommodating than their constituents. Authority to remove craft restrictions on production practices was sought not on the shop or local level but through an agreement with national union officials. And the Terms of Settlement make it clear that the EEF expected union leaders to enforce this agreement on their members.

The EEF's objectives, then, were to establish management's right to change production practices as it saw fit; to do so on an industrywide basis, free from diverse local restrictions and unpredictable local action; and to commit national unions to enforcing these principles on their often unruly members. The Terms of Settlement (later renamed the Terms of Agreement) imposed on defeated unions in January 1898 were designed to realize these goals.

The agreement formally abrogated the union work rules developed to defend craft control. According to the Terms, "there shall be no limitation on the number of apprentices." As against work rules governing assignment of workers to machines, the Terms gave employers "full discretion to appoint the men they consider suitable to work [machines].... The employers consider it their duty to encourage ability wherever they find it, and shall have the right to select, train, and employ those whom they consider best adapted to the various operations carried on in their workshops, and will pay them according to their ability as workmen." The Terms also asserted that "the right to work piecework at present exercised by many of the federated employers shall be extended to all members of the Federation, and to all their union workmen."

For many unionists the lockout and settlement were nothing less than an effort to destroy trade unionism in the industry. Such a prospect did find some enthusiasts among employers, who regarded unions and restrictive practices as two sides of the same coin. "Trade unionism as now practised in the British engineering

trade has to go, or else the British engineering industry has to go. One or the other must be smashed.... It is not only to the interest but the duty of employers to 'smash the union.'"^[28] Yet the EEF did not lump craft unions and craft control together, and the Terms presumed national unions with effective control over their members in the shops. The Terms, after all, represented a national agreement between union leaders and the employers' association. The desired solution to control conflicts was a centralized one, to be enforced by union executives on recalcitrant members. More important, the existence of well-organized national unions was required by the industrywide "Provisions for Avoiding Disputes" set forth in the Terms. Under this procedure, management in any federated firm agreed to meet deputations from their workers if a grievance arose. If the dispute could not be settled, it passed into the hands of the local employers' association and the union district committee. If, finally, the local conference "failed to agree," a central conference between national officials took up the question. Without a union this procedure would be unworkable, as would another stipulation of the Terms: that no strikes occur until the entire procedure had been completed. Preventing workers from walking out unconstitutionally required unions to keep their members in check.

Employers hoped—ultimately in vain—that the Terms would counter the most obnoxious features of engineering factory politics during the 1890s. Conflict produced by management's offensive in the workplace, combined with the peculiarities of union policies and industrial relations, would be contained through a national agreement. In place of craft restrictions the Terms prescribed not just management discretion but also a union obligation to uphold those prerogatives. For local autonomy in the formulation of work rules, the Terms substituted national standards. Indeed, although wages and hours could safely be left to district negotiation, control issues were taken out of rank-and-file hands altogether. Finally, union-management grievance procedures would replace unofficial and unpredictable strikes led by shop stewards or local leaders—and make unions responsible for keeping engineers at work. In addition, employers expected that full-time union officials would be more accommodating than rank-and-file leaders, and the procedure, by centralizing negotiations, was intended to make their

moderate views prevail. A year after urging employers to smash the union, a leading trade journal argued that "it is immeasurably preferable in discussing matters of policy with the workmen to have a responsible agency to deal with, because, after all, the funds of a trade organisation have a balancing influence on its officers.... If a manager compromises with an irresponsible [unofficial] leader, the men may disown such a leader and replace him by a worse, if more talkative demagogue, whereas the unionist is more or less of a capitalist, with the steadying effect which financial expectations bring.^[29]

The United States

Factory Politics up to 1900

Factory politics among machinists of the 1890s were typically defensive in goals, sectional in tactics, and conducted on a basis of local self-reliance. The strategic role of skilled men at work, the nature of craft unionism, and the character of industrial relations favored these responses to management offensives. Differences between the United States and Britain are found largely in the specific organization and tactics of craft struggles and may be attributed to the weaker position of machinists in the typical American shop.

Machinists attempted to block dilution with work rules limiting the number of apprentices and reserving essential machine shop tasks exclusively for duly apprenticed men. In this way they hoped to monopolize jobs and preserve their bargaining power at the expense of less skilled employees. If necessary, machinists fell back on the stipulation that "skilled" work (in effect, work that once required craftsmen, even if it no longer did) be paid at the standard union rate, thus reducing the competition from, if not the presence of, handymen in the shops. To further protect skilled jobs and wages, the IAM prohibited its members from working more than one machine at the same time.^[30] The IAM's position on payment by results was just as uncompromising: the union's constitution simply forbade the introduction of piecework. Where piecework already existed, however, no national policy regulated its operation. Trade rules thus varied from one shop or district to another.

Strongly organized machinists often demanded that piecework systems be discontinued, with demands backed up by strikes and concessions written into formal agreements. But in most cases machinists had to rely on covert restriction of output to secure favorable piece prices, prevent cuts, and safeguard the earnings and effort levels of the average (or even less than average) machinist.^[31]

Unionists did not consider these policies suitable topics for negotiation with employers. The IAM censured, fined, or expelled members who violated work rules, and it boycotted firms that refused to observe union conditions. Such unilateral strategies, to be sure, often failed. American employers were in a better position than their British counterparts to survive union boycotts because they relied less heavily on skilled labor and could readily find nonunion machinists. As a result the IAM was far more eager to confirm craft rules through formal contracts between employers and union locals.^[32] Union contracts and formal work rules reflected the failure of traditional defenses. Before the 1890s, the strategic importance of machinists in metalworking, combined with a strong group ethos defining appropriate work behavior, usually sufficed to resist management encroachments. As employers became less dependent on craftsmen and piecework undermined work group solidarity, machinists had to resort to formal rules, formal organization, and formal controls over management.^[33]

Whatever the tactics, machinists aimed to preserve workshop privileges for themselves alone. Like British engineers, machinists defended their jurisdictions from rival unions as vigilantly as they resisted the upgrading of less skilled workers.^[34] Machinists backed up exclusive policies with sectional tactics. For example, they refused to instruct handymen or to give helpers the same guidance extended to bona fide apprentices. They took more drastic measures where needed. A Cleveland machinist reported the difficulties a handyman experienced in operating his machine: "The intense mental strain he underwent trying to locate the trouble caused him to stay away to-day and it will be to his interest to stay away for good.... I needn't add that the machine was o.k. before he took it and is o.k. now."^[35] Other strategies, such as restricting output or simply boycotting obnoxious employers, relied on resources available only to skilled workers: their superior knowledge of production techniques and their indispensability for essential

machine shop tasks. Striking machinists, moreover, received union benefits; the unskilled employees laid off during the dispute were left to shift for themselves.

An unsupportive national union and parochial industrial relations made machinists, like engineers, rely on local organization to defend craft interests. Responsibility for formulating and enforcing specific work rules lay primarily with local lodges, because in regulating piecework or shop discipline, for example, no national policies applied.^[36] The variation and localism of industrial relations also favored local autonomy. Machinists employed in railroad shops faced quite different problems than those working for northeastern machine tool firms or midwestern agricultural implement factories. In the latter settings more advanced production methods and anti-union employers might require concessions on work rules to preserve even a toehold in the plant. With metal trades employers as yet unorganized, machinists faced few pressures to coordinate policy on a national basis.

Workshop organization played an important role in enforcing union rules. In many shops one or more machinists took responsibility for checking membership cards, notifying lodge secretaries of vacancies, and monitoring working conditions on behalf of the union. Whether or not they were formally recognized as a shop committee, such workers would be regularly involved in deputations to management (the IAM's usual first step in dealing with grievances), in calling shop meetings, and in leading strike committees.^[37] It is less clear how tightly shop committees were linked to local lodges. Committeemen were union activists and often worked with local officials in formulating demands and planning shop meetings, but they were not usually formal union stewards. IAM lodges did, however, try to secure formal recognition for shop committees and protection against victimization of men doing committee work."^[38]

The role of shop committees in policing the works corresponded to an industrial relations system in which conditions varied from one firm to another, negotiations were in the first instance conducted informally, and many managers preferred to deal only with their own employees. Shop committees also overcame certain obstacles to the defense of craft rights created by formal union procedures. These bodies allowed machinists to act more quickly

against encroachments than IAM procedures allowed. Shop committees also offered a means for united action by members otherwise divided among different lodges. Committeemen, finally, were more familiar with specific plant conditions than were local officials, much less executives in Washington. Machinists, like engineers, called in union officials only after the failure of their first line of defense: direct action.

Local autonomy was less strongly developed in the United States than in Britain. The IAM had some consistent policies governing dilution and incentive pay (including a ban on the two-machine system and the introduction of piecework), and union lodges incorporated these rules into contracts with employers whenever possible. The narrower discretion of American locals may be traced in part to the delayed development of unionism relative to changes at work. Machinists, unlike engineers, could not rely on custom and strong workshop organization to support craft rights. More standardized work rules, backed by the IAM Executive, represented a second line of defense.

Machinists also were less effectively organized on the job. Outside of the best-organized districts, shop committees appear to have been ad hoc affairs, forming to present demands or prepare for strikes, vanishing (or being discharged) once the crisis had passed. The fragility of workshop organization among machinists, compared to engineers, reflects the lesser importance of skill in U.S. manufacturing and the weaker position of craft unions on the job. American workers were also more nomadic than British ones.^[39] Frequent job changes—in search of high wages, in response to grievances, or to evade blacklists—undercut work group cohesion and stable shop organization. Given the relative weakness of shop committees, demands and disputes that in Britain might have been initiated by shop stewards were instead the responsibilities of IAM lodge officials.

The weakness of workshop organization is also evident in the strategies adopted for defending craft standards. The IAM resorted more often than did the ASE to sympathy strikes and boycotts rather than unilateral action, and these strategies typically relied more on cooperation among local union leaders than on rank-and-file activism within the shops.^[40] The relative importance of official cooperation again suggests the weakness of union organization on

the shop floor and the inability of local work rules alone to hold the line. Indeed, Montgomery suggests that where work rules remained secure and union defenses intact craftsmen less often participated in sympathetic action.^[41]

Trade unions and industrial relations, then, favored local autonomy in struggles to preserve craft control. In comparative perspective the different character of American production methods and craft union development did weaken local organization, especially within the plants. Yet even if IAM shop committees and lodges were less effective in resisting employer offensives, their organizational features ensured that protest would be defensive and exclusive. These bodies, like their British counterparts, extended the sectionalism of national unions to the local and workshop levels: shop committees represented only IAM members, and they relayed national and local policies to the rank and file. Shop organization, whatever its comparative weaknesses, also enhanced craftsmen's power relative to less skilled employees. These factors combined to inhibit joint action on widely shared grievances and progressive demands. Machinists instead typically acted alone and in defense of their traditional privileges.

Machinists' local self-reliance had one further consequence: endemic friction between the IAM Executive (the Grand Lodge) and the members over unconstitutional strikes. Union leaders in the United States, as in Britain, were more cautious in resorting to strikes and more conciliatory in their defense of craft privileges than were rank-and-file machinists.^[42] Those in the shops had no wish to let encroachments stand or to see favorable opportunities to press demands slip by while constitutional procedures ran their course. Local organization provided an agency for prompt action. IAM lodges also retained the majority of each member's dues, lessening the need for Grand Lodge aid in conducting strikes.^[43] As a result, between June 1899 and June 1901, national officers sanctioned 66 strikes involving approximately 9,650 machinists; over the same period local lodge reports (which would not have been exhaustive) showed them engaged in 246 strikes involving 12,912 machinists.^[44] IAM leaders could only lament unofficial strikes and make ritualistic pleas for stronger Grand Lodge control over strikes and strike benefits. "Our Constitution prescribes the proper steps to be taken if there is a prospect of a strike resulting from an effort

to settle a grievance, but of what use is it? How many of our past strikes have been undertaken in accordance with the provisions of our law ... ? What is to prevent a dozen or more strikes at the same time?—each one of them inaugurated in the same way by the irresponsible authority of a Local or District Lodge?"^[45] In 1900, the answer—for the IAM Executive and leading employers—appeared to be a national trade agreement.

The 1900 Strikes, the Murray Hill Agreement, and the Open Shop

Early in 1900, IAM District Lodge 8, representing all Chicago locals, presented 150 employers with a number of demands, of which the ostensible centerpiece was the nine-hour day. Most firms refused even to meet with District 8 representatives. A strike began on March 1; at its peak it involved some five thousand machinists. Following Chicago's example, four hundred more struck in Columbus, Ohio; three hundred in Paterson, New Jersey; and similar numbers in Cleveland, Detroit, and Philadelphia.^[46]

When Chicago machinists first presented their demands, local employers were largely unorganized, and a number of smaller firms signed the union agreement. Larger employers soon banded together, however, under the auspices of the National Metal Trades Association (NMTA). Founded in New York City in 1899 to resist demands by patternmakers, the NMTA had eight members in Chicago by the beginning of 1900. The threat posed by Chicago machinists brought the NMTA additional members and inspired some seventy employers to form the Chicago Association of Machine Manufacturers. From the outset, local and national NMTA representatives took the lead in managing the strike and negotiating with union officials.^[47] After some unsuccessful efforts to break the strike,^[48] NMTA leaders informally pledged to accept the principle of collective bargaining, to grant the nine-hour day in installments over the coming year, and to refrain from discriminating against unionists or strikers. In exchange the IAM called off its strikes throughout the country, with all outstanding issues to be taken up in a national conference.^[49] The Chicago Agreement, approved by local employers and IAM members in March, settled the strike on these terms. It established a board of arbitration,

which, meeting in New York's Murray Hill Hotel, reached a national pact (the Murray Hill Agreement) on May 18.

The details of the dispute reveal basic characteristics of factory politics up to 1900, with conflict focused on craft privileges, local autonomy, and union discipline. The 1900 strikes continued struggles over craft control that had been accelerating since 1897. Between May 1897 and June 1899, for example, the IAM General Executive Board had received requests for assistance in fifteen wage disputes. Over the same period, locals sought support in twenty disputes over piecework, twelve over dilution, and eighteen over abuses in the hiring and firing of machinists.^[50] Although IAM leaders claimed a shorter workday as the machinists' key goal,^[51] proposed agreements put forward by local lodges emphasized instead issues of workshop control. IAM locals asked employers to recognize certain tasks as machinists' work, for which minimum wage rates must be paid. Machinists also demanded that employers observe the apprenticeship regulations laid down in the IAM constitution, and Columbus locals added a ban on the two-machine system. To check victimization and arbitrary employment policies, proposed agreements stipulated that in the event of layoffs employers give preference to those discharged when rehiring workers, seniority and proficiency to govern." Finally, in a more general check on arbitrary authority at work, machinists demanded that shop committees be recognized for the purpose of negotiating grievances, with unresolved disputes to be referred to joint conferences between local union officials and the company. Failing agreement, a board of arbitration would be established, with two members selected by the union local , two by the firm, and a fifth chosen by those four. This board's decision would be binding.

These were traditional, exclusive goals. Machinists sought to protect their monopoly of basic machine shop tasks. Apprenticeship restrictions aimed to block the promotion of handymen at the expense of machinists. Minimum wage rates assigned to machinists' work prevented employers from substituting cheaper labor. In each strike the workers demanded not only that employers accept their definition of a machinist, but also that only IAM members be employed for that work. Those in other trades as well as those of lesser skills were to be denied access to machine shop jobs. Nor in the actual conduct of the strike were machinists inclined to cooperate with other crafts or with semiskilled workers.

The 1900 strikes also demonstrate tensions between local autonomy and national leadership. For the IAM Executive the 1900 strikes were to achieve the nine-hour day; but local lodges took the initiative of adding their own demands, which varied from one city to another. Referring to Chicago, IAM's Vice-President Wilson lamented that, although the union's key goal was the shorter workday, "the men were in that frame of mind that they thought they could win anything, that they had only to make a demand and it would be granted."^[52] In the pursuit of local concerns, machinists also lacked respect for executive authority and IAM constitutional procedures. When President O'Connell and organizer Stuart Reid arrived in Chicago, three thousand members had already walked out. Convinced that this was enough, O'Connell and Reid urged the rest to remain on the job. "Our request was ignored, and 1,500 more walked out.... All our energies have been exhausted trying to keep men at work."^[53] Nor was O'Connell sure that the strikers would accept the Chicago Agreement and return to work. The decisive meeting was held at a Salvation Army hall, which conveniently contained a large bass drum. The drum was brought on stage and "played a very important part, for it was beaten with tremendous effect when speeches were being made favorable to the agreement and was always silent when they were unfavorable."^[54]

O'Connell had good reason for concern, for in his pursuit of the nine-hour day and formal bargaining status he had agreed to drop many of the control demands made by local machinists. The closed shop, the formal recognition of shop committees, and the restrictions on the two-machine system and on management prerogatives in hiring and firing were all abandoned by IAM leaders in the course of negotiations.^[55] Indeed, under the Murray Hill Agreement IAM officials promised that union members would "place no restrictions upon the management or production of the shop, and will give a fair day's work for a fair day's wage."^[56] This clause, together with the Agreement's guarantee of employers' freedom to hire nonunion labor, was apparently omitted from copies of the pact circulated to IAM locals.^[57] Clearly, IAM leaders were more willing than their constituents to trade control for a shorter workday and formal bargaining rights, and they were aware that these different priorities could create difficulties.

For employers, as for rank-and-file machinists, the problem of control dominated the 1900 strikes. Employers had no objection to the nine-hour day so long as it was widely granted and Chicago firms were not put at a competitive disadvantage. The minimum wages machinists requested caused no alarm because most Chicago employers already paid their workers at least that much.^[58] What manufacturers could not accept were restrictions on their rights to hire nonunion labor, to assign workers to machines as they pleased, and to judge for themselves whom to retain and whom to discharge. Nor would they formally recognize a shop committee to contest their decisions and enforce union work rules. So long as union leaders agreed with these priorities, the NMTA would accept them as responsible bargaining partners. Union recognition and the nine-hour day were prices employers were willing to pay to consolidate their control of the shops.

NMTA strategies during the 1900 strikes also demonstrate employers' concern with two other features of factory politics in the late 1890s: local autonomy and irregular strikes. NMTA negotiators refused to settle the strikes on a purely local basis. Although wages were a legitimate subject for local bargaining,^[59] local autonomy in the area of work rules and management practices was unacceptable. NMTA representatives in Chicago declined to deal with the local union because it "does not have a correct knowledge of conditions as they affect the industry at large.... But we will recognize your national union through our national association"^[60] and thus ensure that "practically the same conditions of labor shall prevail in all the different sections."^[61]

A national agreement would have the further virtue of protecting employers from irregular local disputes. The NMTA had been founded to present a united front against local labor demands. By recognizing national union leaders, employers hoped to enlist their aid in checking local (and often unofficial) strikes. The importance of this for NMTA leaders is indicated by their refusal to negotiate until IAM officers called off outstanding strikes. If President O'Connell could not do so, employers reasoned, there was no guarantee that he could enforce any subsequent agreement or prevent strikes pending arbitration.^[62] Union officials evidently shared the NMTA's hopes for a national solution to local unrest. "If this idea—the board of arbitration—is carried out in a spirit of fairness and

equity, mutual concessions being made ... there need never be any more strikes or lockouts, as far as the machinists are concerned."^[63]

The Murray Hill Agreement of May 1900 embodied the goals of NMTA and IAM leaders. Employers conceded the nine-hour day, to be introduced in two installments at six-month intervals, and granted enhanced pay rates for overtime work. NMTA members agreed to employ no more than one apprentice for every five machinists—a concession of little value to craftsmen because employers no longer relied on apprenticeship to train workers.^[64] Employers also largely accepted the union's definition of a machinist as a worker competent to perform specified tasks. But no minimum wages were guaranteed for those tasks, and employers explicitly retained the right to decide which men were "competent to perform the work and to determine the conditions under which it shall be prosecuted." In return the union abandoned local demands for the closed shop, seniority rights, and shop committee recognition and pledged not to place any "restrictions upon the management or production of the shop."

The Agreement also reaffirmed the procedure that had been established in Chicago for dealing with disputes. Any grievance not settled informally within the shops had to be referred to the national board of arbitration, consisting of three representatives each of the NMTA and IAM. In November both sides agreed to add an intermediate layer of negotiations, involving representatives chosen by the local branch of the NMTA and by the aggrieved employees. No strikes or lockouts were permitted until the completion of national conferences, and NMTA and IAM leaders promised to discipline unruly members.

Confronted with many of the same obstacles as their British counterparts, NMTA firms thus adopted a similar solution: a national trade agreement affirming management control and enlisting union leaders to check rank-and-file workers.^[65] IAM officials eager for recognition and organizing opportunities willingly obliged. Yet for all the apparent advantages of Murray Hill to employers and union executives, the Agreement collapsed within a year, with the breakdown in large measure engineered by the NMTA.

The reduction of hours to nine a day was to take place in two steps—the first half-hour in November and the second in May

1901. In November it became clear that employers did not assume that the cut in working hours would be compensated by an increase in wage rates. For the most part employers, with business ample and labor scarce, did raise wages. Machinists, however, demanded that their leaders secure a blanket pay hike corresponding to the nine-hour day, and here the NMTA balked.^[66] Sticking to the letter of the Murray Hill Agreement, employers insisted that wage negotiations were a local matter, subject to national arbitration only upon local failure to agree, with each case decided on its individual merits. Although NMTA leaders refused to deal with wages on a national basis, they also advised local members faced with IAM proposals "that no individual action on said document be taken."^[67] Having already committed itself to a national strike to win the nine-hour day at non-NMTA firms, the IAM chose to fight as well those NMTA firms that refused to grant wage increases. On May 20, 1901, some forty thousand machinists went out as planned. Strikes persisted at least into October, but except in Chicago and in small, non-NMTA shops elsewhere, the IAM failed to secure shorter hours, wage increases, or union agreements.^[68]

The strikes against NMTA firms were nominally a breach of the Murray Hill Agreement, from which the NMTA, deploring the IAM's irresponsible behavior, withdrew. But employers' disenchantment with Murray Hill had deeper roots. NMTA representatives complained that the union's pledge to leave management to employers alone was not being honored in the shops. "In numerous cases and through widely separated localities the union machinists have flagrantly restricted the production of the shops, denied flatly the right of the employer to govern his own affairs, and have interfered with the management and methods of our shops, and it is this point which we declare to be the issue which we defend in this contest."^[69] Nor did the Murray Hill Agreement succeed in preventing unauthorized strikes.^[70] These were two sides of a single problem: national IAM leaders could not, after all, control their own members. Indeed, for many employers the Murray Hill Agreement had effects quite contrary to those intended by the NMTA. Union recognition helped spur dramatic increases in IAM membership, from 22,500 in 1900 to 32,500 a year later.^[71] Rather than respecting the terms of agreement, machinists used their added power to repudiate those terms—refusing to work "harmo-

niously" with nonunion workmen, insisting on craft restrictions, and taking advantage of the business boom to stage quick strikes. Discussions among manufacturers at an NMTA conference late in May concluded that "the foothold gained in the shops by the union, under the operation of the New York agreement, had resulted in the introduction of practices by workmen which were subversive of discipline and detrimental to the interests of employers.... These facts being made clear at the conference, a determination to free themselves absolutely from union control grew with irresistible strength."^[72]

The employers articulated their determination in the NMTA's "Declaration of Principles," adopted May 28, 1901. Management's prerogatives were once more asserted, combined not with a national system of union-employer arbitration but with the open shop. "Since we, as employers, are fully responsible for the work turned out by our workmen, we must, therefore, have full discretion to designate the men we consider competent to perform the work and to determine the conditions under which that work shall be prosecuted.... The number of Apprentices, Helpers and Handymen to be employed will be determined solely by the employer," and "employees will be paid by the hourly rate, by premium system, piece work or contract as the employer may elect." In the event of a dispute over some other issue, firms were advised to meet with representatives from their own employees, but no negotiations would be entertained if the workers went on strike.^[73] This declaration brought the NMTA new recruits,^[74] and beginning with the 1901 strikes the association's full resources were committed to defending the principles of managerial rights and the open shop—principles now defined as synonymous.

Comparative Conclusions

Engineers and machinists of the 1890s aimed to preserve customary prerogatives and privileges at the expense of both management and other workers. Craft protest rarely involved cooperation with members of other trades or with less skilled workers. The continued strength of craftsmen in late-nineteenth-century manufacturing favored mobilization behind such conservative goals and along sectional lines. The institutional environment for protest similarly

promoted exclusive action. Unsupportive unions and localistic industrial relations led metalworkers to rely most heavily on district and shop organizations to defend their interests—a strategy that frequently put national leaders and the rank and file at odds. These same organizations, however, reproduced the exclusiveness of craft unions at the local level and made it unnecessary for machinists and engineers to seek the support and take into account the interests of other work groups.

The relatively tenuous position of skilled workers and craft unions in the United States did weaken workshop organization. Machinists often had to resort to sympathy strikes and were more dependent on local union officials to coordinate the defense of craft standards. Yet these differences in tactics and leadership had little effect on the goals and alliances of craft protest. Machinists still enjoyed a stronger bargaining position than less skilled workers, and they remained organized locally along sectional lines.

This picture of factory politics in the 1890s might suggest that engineers and machinists were uniformly conservative. They were not. Distinct minorities in both the ASE and the IAM endorsed socialism and supported the organization of less skilled metalworkers. Industrial unionist sentiments reflected the impact of Britain's "New Unionism" and America's Knights of Labor and the American Railway Union during the late 1880s and early 1890s, along with a growing recognition that changes in industrial organization required alternatives to traditional craft strategies.^[75] Socialist and industrial unionist commitments, moreover, had some effect in both unions: ASE members elected a socialist General Secretary (George Barnes) in 1896, and IAM delegates in 1895 and 1897 expanded direct membership control over national officers through the use of the referendum. In neither union, however, did radicals influence the formulation of trade policies or the conduct of disputes. The conditions that would enable progressives to gain broader support among engineers and machinists did not develop until after 1900. During the 1890s, practical factory politics remained defensive in goals and exclusive in strategy.

For employers, craft control meant impediments to their right to manage. Local autonomy, with work rules varying in character and stringency from one shop or district to another, placed manufacturers on an unequal competitive footing, and rank-and-file ini-

tiative exposed them to unpredictable disruptions in production. The 1897–1898 and 1900 disputes typified these dilemmas for metal trades employers and at the same time seemed to offer an opportunity to solve them once and for all. The solutions represented by the Terms of Settlement and the Murray Hill Agreement were the same: formal union guarantees to respect management rights and to discipline union members, and to do so on a uniform, national scale.

The key comparative problem is to explain the subsequent divergence in employer strategies. Why did employers abandon the Murray Hill Agreement in favor of the open shop, while the Terms endured? Certainly the Murray Hill Agreement did not live up to employer expectations. Craft restrictions continued, as did unofficial strikes in violation of the arbitration procedure. Montgomery concludes that this inability of IAM leaders to hold their constituents to the terms of Murray Hill accounts for employers' recourse to the open shop.^[76] But ASE national officials were perhaps even less successful in this regard, and still the EEF clung to the Terms. The most important reasons for this divergence involve, once again, contrasts in the modernization of the labor process and the timing of craft union development. Differences in production practices and the strength of trade unions had two important consequences for management strategy. First, NMTA employers had less to gain from collaborating with union leaders and less to lose in fighting them. Second, these differences in manufacturing techniques and craft unionism contributed to broader industrial traditions—traditions that made open shops especially attractive to U.S. employers and trade agreements congenial to British ones.

In the United States dilution had made greater headway, and had done so more rapidly, than in Britain. This progress made union restrictions more irksome to U.S. employers and simultaneously reduced their reliance on craft unionists to get the work done. NMTA members thus had both greater interest and greater resources to attack craft standards. Because the erosion of craft control was under way before the IAM's consolidation, U.S. employers had another advantage: relatively feeble union opposition. The NMTA faced a union much weaker than the ASE, and employers quickly found that the Murray Hill Agreement served to strengthen the IAM. The IAM enrolled only 11 percent of American machinists in 1900, compared to perhaps 50 percent of engi-

neers enrolled in the ASE and other craft unions.^[77] Fighting the IAM was a far more realistic option than taking on the ASE, and an option that might have been foreclosed by union organization under the protection of the Murray Hill Agreement. By the same token the IAM was thus of less potential value in controlling machine shop employees. In any case U.S. firms had less need for unions to standardize labor costs and conditions among shops and cities, for the concentration of capital and specialization of firms eased competitive pressures by other means.

Given these conditions, one might reasonably ask why the NMTA resorted to the Murray Hill Agreement for even a single year. Although some NMTA leaders sincerely supported an experiment in national arbitration, for many members in 1900 (and many more who accepted the NMTA's leadership in the dispute but joined only in 1901), the Murray Hill Agreement was probably a temporary expedient. The Agreement was affordable, and it provided some stability in labor costs when business was exceptionally prosperous and skilled workers scarce. It also offered breathing space for the fledgling NMTA to expand its organization and prepare for future battles. With increasing pressure on profit margins in 1901 and after, employers had fewer incentives to abide by the Murray Hill Agreement,^[78] and they were in a stronger position to insist instead on the open shop. It is also significant that the Agreement emerged from confrontations in Chicago, the nation's most strongly unionized city.^[79] Here collective bargaining survived even after the demise of Murray Hill. Elsewhere the IAM was rarely in a position to force collective bargaining on unwilling employers.

The British engineering industry, by contrast, experienced a major boom for several years following the conclusion of the 1897–1898 lockout, thanks to the Boer War. Although national collaboration rewarded the hopes of the EEF no more than those of the NMTA, under prosperous conditions British employers were reluctant to risk a general confrontation with engineering craft unions. For the remainder of the prewar years, there was no business slump so serious or prolonged as to tempt the EEF with the prospects of abandoning the Terms. Indeed, the industry's procedures may have helped delay and limit wage advances relative to prices, compared to what engineers might have secured if given a free hand.

Industrial traditions as well as immediate practical considera-

tions influenced employer policies. These traditions demonstrate the influence of the timing of changes at work relative to industrial relations. By 1897, British engineering had a tradition of collective bargaining between employers and skilled workers that stretched back over forty years. That tradition had been consolidated during a period of relative stability in manufacturing techniques and slow but steady growth in exports. Most progressive employers in the 1890s assumed that disputes should be resolved through negotiation and agreement.^[80] Although some managers were tempted by the prospects of a union-free environment in the midst of the 1897–1898 lockout, manufacturers' continued reliance on skilled workers made open-shop policies unrealistic. EEF leaders saw the crisis instead as an opportunity to secure formal union compliance with managerial initiatives and a formal union role in shop discipline. A representative of the General Electric Company in London could justifiably claim in 1901 that "our attitude towards the Trades Unions has always been to support them so long as they give us a guarantee that their members are efficient workers."^[81]

No such breathing space for the consolidation of collective bargaining existed in the U.S. metal trades. Even before the 1890s, U.S. employers had begun meeting new product markets and a scarcity of skilled labor with strategies of dilution and shop reorganization, rather than through a slow expansion of employment within the technical status quo. By the time the IAM gained its footing in the mid-1890s, rapid changes at work were already under way, and management had secured some freedom from dependence on machinists. Under these conditions union recognition and collective bargaining seemed to endanger employers' right to manage and appeared incompatible with business success.^[82] Particularly in the northeast—where union organization was weakest and production methods most progressive—industrial relations traditions involved informal negotiation with the firm's own employees, if possible on an individual basis. No legacy of collective bargaining existed to help the Murray Hill Agreement survive the 1901 dispute.

Broader national traditions reinforced the choices made in each case. There is some validity in the stereotyped contrast between turn-of-the-century British and American employers—one civil, tactful, and conservative; the other rough, unyielding, intolerant

of obstacles. The same conservatism that made most British engineering employers reluctant to seize new technical opportunities and revamp traditional workshop practices^[83] made them unwilling to break with customary industrial relations. Given traditional manufacturing practices, they were in any case unable to dispense with engineers and their unions. U.S. employers, eager to overcome technical obstacles to more efficient and profitable production, were similarly impatient with the human obstacles presented by craft unions. A more advanced labor process put them in a better position to rid their shops of union interference. Moreover, they had at their disposal alternatives to bargaining with union leaders that were less available to British manufacturers—particularly judges willing to virtually ban strikes and boycotts and state and local governments willing to apply force on their behalf. Heavy-handed repression of union activity, in turn, received greater support from the press and the middle class in the name of property rights and American individualism.^[84] Thus U.S. employers had not only the inclination to drive unions from their shops but also the means to do so.

The contrasting outcomes of the 1897–1898 and 1901 disputes should not be overdrawn. There remained British shops (particularly in new branches and centers of the industry) in which managers declined to deal with "outside" interference and employees risked discharge if they advertised their union affiliations or insisted on union conditions. Although most U.S. employers (and all NMTA members) refused to negotiate with union representatives, skilled machinists remained valuable. Unless they made trouble, their union ties were not too closely monitored. To keep the peace, employers met deputations to discuss grievances, even if these were in effect shop committees acting under local union instructions. Despite these qualifications, the conflicts of the 1890s had starkly different outcomes: in Britain the Terms of Agreement, in the United States the open shop. This contrast in particular accounts for the subsequent divergence in British and American factory politics.