PART TWO
THE MANAGERS' COMPANY

Three
The Decision Makers

Preoccupied with the governed classes, European social historians have not shown as much curiosity about the governing classes, not even about the new elite of businessmen and salaried managers emerging in the nineteenth century. Their career trajectories, responsibilities, compensations, family backgrounds, and outlooks await serious analysis.^[1] The PGC, as one of France's early corporations, provides an opportunity to study salaried managers just as they came to form a distinct socioprofessional category. Corporate management was not yet a familiar activity and lacked clear-cut norms and prescriptions when the firm was founded. Only railroad companies and a few other large enterprises existed to serve as models.^[2] In truth, their example was only marginally useful to decision makers at the PGC. Corporate forms and practices arose gradually within the gas company, partly through trial and error. The model of the bureaucratic state proved highly influential, too. Such were the forces molding a powerful, new economic elite.

The Administrative Hierarchy

The statutes of the PGC laid down once and for all and with Cartesian clarity the lines of authority that were supposed to direct the firm. Given

[1] For one rare study of these questions, see Maurice Lévy-Leboyer, "Hierarchical Structure, Rewards, and Incentives in a Large Corporation: The Early Managerial Experience of Saint-Gobain, 1872-1912," in Recht und Entwicklung der Grossun-ternehmer im 19. und frühen 20. Jahrhundert, ed. Norbert Horn and Jürgen Kocka (Göttingen, 1979), pp. 451-472.

[2] The importance of railroads in establishing patterns in corporate management is a well-known theme. See Alfred Chandler, "The Railroads: Pioneers in Modem Corporate Management," Business History Review 39 (1965): 16-40, and The Visible Hand: The Managerial Revolution in American Business (Cambridge, Mass., 1977). A few of the PGC's early executives had worked for railroads, but it is doubtful whether that experience was formative.

the weight of French tradition, one might expect those lines to have been highly centralized, and they were. The nature of the executive at the center, however, changed several times over the life of the firm. Furthermore, the concentration of authority proved far greater in principle than in reality, the result of both choice and circumstance. The PGC was not so inflexible as to have faced all its challenges without structural innovation. It responded to the environment with new managerial approaches and tolerated even a good deal of ambiguity about lines of authority. Of course, such flexibility did not in itself guarantee wise or appropriate decision making.

Sovereignty over the firm was vested in the stockholders, but they immediately delegated their authority to a conseil d'administration, or board of directors, composed of twenty elected shareholders. Meeting faithfully at least once a month for the fifty-year life of the firm, board members envisioned for themselves a supervisory role.^[3] No decision would have force without their approval, but they declined to be the managers of the corporation. Instead, the very first board, led by the owners of the merging gas companies, decided on a collective executive. The board created a five-member executive committee (comité d'exécution ) made up of rotating board members. Each member was to oversee a set of operational departments. Though the committee's name implied the passive role of carrying out the wishes of the board, the committee was actually the governing force—until the need for further change became evident.^[4]

The owners of the new gas firm quickly found a collective executive too cumbersome. The first year of business brought vast new opportunities, the beginning of daytime consumption, shortages, confusion, and the need to plan for expansion. An executive committee composed of parttime managers preoccupied by interests outside the firm could not deal seriously with pressing matters. Besides, they received compensation for their tasks essentially as stockholders, through dividends. The need for a full-time chief executive who would receive ample compensation for his services to the gas company alone was manifest. In November 1856 Louis Margueritte, an owner of one of the merging firms and an influential board member, proposed the creation of the office of director. The incum-

[3] The deliberations of the board are in AP V 8 O nos. 723-737.

[4] Ibid., no. 723, deliberations of January 11, 1856. The records of the executive committee are in nos. 665-700.

bent would provide leadership and implement all the decisions approved by the board and executive committee. These two bodies would continue to meet (in the case of the committee, as often as once a week), with the director presiding. He would also have the power to choose the committee members.^[5] Thus the office became the central one in the firm even though the director was not absolute master. As the PGC grew larger and more complicated and as board members grew more distant from the gas industry, the director assumed the role of the company's animating force. On matters of potential controversy it is likely that the director was careful to consult with influential board members and build a consensus in the executive committee. In any case, minutes of the committee meetings show virtually automatic confirmation of the director's proposals.

With the appointment of a director before the first year of operation was over, the PGC would seem to have joined the ranks of the new breed of firm, still rare in France and Europe, run by a salaried manager rather than by the owner.^[6] Yet this was not quite the case, at least for a while. The board of directors offered the post of director to Margueritte, but he turned it down. Another board member, Vincent Dubochet, took the position. He, like the board's first choice, had been an owner of one of the pre-merger gas companies and a founder of the PGC. Both were among the largest stockholders. Whether the board fully intended the choice to set a precedent for future appointments is unclear, but the selection of Dubochet as director was only half a step away from the owner-manager. Accentuating the incomplete transition to a new form of management was the board's failure to establish a salary for the new director until a year after his appointment.^[7] The definitive move to a salaried director came with Dubochet's retirement in 1858, when the board entrusted the fate of the firm to an officer of the prestigious Corps des Ponts et Chaussées (Corps of Bridges and Roads). An annual salary of twenty thousand francs, supplemented by a generous profit-sharing bonus, was established immediately.^[8] The chief executive in the PGC was no longer a collective entity nor an owner-founder but rather a graduate of France's most illustrious technical institutions and a member of a national administrative elite. Though salaried directors of corporations were soon to become com-

[5] Ibid., no. 723, deliberations of November 20, 27, 1856.

[6] For a historical survey of large corporations in Europe, see Herman Daems and Herman Van Der Wee, eds., The Rise of Managerial Capitalism (Louvain and The Hague, 1974), and Alfred Chandler and Herman Daems, eds., Managerial Hierarchies: Comparative Perspectives on the Rise of the Modern Industrial Enterprise (Cambridge, Mass., 1980).

[7] AP V 8 O , no. 723, deliberations of April 30, 1857.

[8] Ibid., deliberations of May 1, 1858.

monplace, appointing one was not yet routine when the PGC was founded.^[9] The firm had spent three years groping toward such an office as a solution to its executive needs.

The central office retained essentially this form throughout the rest of the company's history, but there were some adjustments. The post of assistant director was added in 1860 to help carry the growing burden of administration. The director's position also suffered a temporary demotion between 1892 and 1901. In an effort to deal with labor demands from city hall and with the threat of unions, the board appointed a well-known advocate of industrial reform to the directorship. It seems clear, though, that the person he replaced, Emile Camus, remained in command. Immediately upon resigning as director, Camus was named delegated board member, charged with "a superior control over all services." Camus received higher pay than the new director and was present at most of the crucial meetings.^[10] The exposed political situation of the firm after 1890 apparently forced complications on the executive hierarchy but did not change it fundamentally.

The corporate statutes were inevitably hazy about the staff that would assist the director and executive committee. The principle of indivisible sovereignty discouraged a priori concern with the delegation of decision making. Nonetheless, the statutes did recognize thirteen "officers" of the firm; eight of them were to supervise bookkeeping departments, and the rest were entitled "engineer" or "assistant engineer" and were to oversee production operations.^[11] What is noteworthy is how much the list diverged from the positions that really came to matter. By no means were the officers and the effective managers of the PGC the same. The company operated with a nebulous concept of management. Even as crucial a figure as the superintendent of a factory was not listed as an officer. By contrast, the statutes made provisions for such relative nonentities as "chief of correspondence" or "assistant cashier." Moreover, personnel records never distinguished clearly between mere clerks and managers. The conceptual framework organizing the personnel charts was salary level—above or be-

[9] On the graduates of elite engineering schools heading important French corporations, see André Thépot, "Les Ingénieurs du Corps des Mines, le patronat, et la seconde industrialisation," in Le Patronat de la seconde industrialisation , ed. Maurice Lévy-Leboyer (Paris, 1979), pp. 237-246.

[10] AP, V 8 O , no. 733, deliberations of April 14, 1892. The post of administrateur-délégué had always existed (indeed, it was required by law) but had not been filled until this awkward moment for the PGC. Camus did not move out of the director's residential suite within the corporate headquarters until 1898 (see no. 735, deliberations of August 25, 1898).

[11] Ibid., no. 723, appendix to deliberations of December 26, 1855.

low three thousand francs a year—not level of responsibility n Beginning managers were on the same salary ladder as office clerks; both started at eighteen hundred francs. The only difference was that the managers advanced faster. Even without a clear notion of what "management" could and should be, the PGC did develop an executive corps that served its needs well.

The new gas company quickly assembled a staff of twenty-three people, who were primarily responsible for ordinary operations (table 4). Seventeen of these headed technical departments or divisions and were graduates of state engineering institutes; the rest headed clerical departments. The managers below the director did not simply constitute a mechanism for passing on his orders. In fact, it is possible to apply the classic labels of top, middle, and lower management to the hierarchy—though admittedly the language is entirely anachronistic. Nineteen officers supervised directly the production and distribution of gas. Three others oversaw, evaluated, and perfected the work of those nineteen, so they might be considered middle management. The director and his immediate assistants coordinated the efforts of their subordinates and were chiefly responsible for general policy. These distinctions do need some qualification, but they hold true in a broad sense. Between the founding of the firm and the end of its golden age, the size of the managerial staff nearly doubled, with most of the growth at the lower level. There was hardly any further expansion at any level during the era of adjustment. The crisis through which the PGC passed did not spark a reconsideration of administrative organization, and the reduced pace of growth obviated the need for an expanded executive corps.

What sort of relations prevailed among the different managerial levels? The corporate statutes as well as national traditions would lead one to expect thorough centralization.^[13] In theoretical terms, the director made all decisions, even trivial ones. In practice, his oversight did cover many matters of detail, with middle and lower managers merely supplying ad-

[12] For examples of personnel charts, such as they existed within the PGC, see ibid., no. 153, "Cadre du personnel," and no. 162, "Assimilation." Michael Crozier, The Bureaucratic Phenomenon (Chicago, 1964), p. 273, writes of a "lag" in the development of French managerial functions because firms were overcentral-ized and managers gloried in being in control of everything. As we shall see, this stereotype does not quite hold for the PGC.

[13] Octave Gelinier, Le Secret des structures compétitives (Paris, 1968); Crozier, Bureaucratic Phenomenon ; Eugene Burgess, "Management in France," in Management in the Industrial World, ed. Frederick Harbison and Charles Myers (New York, 1959), pp. 207-231; James Laux, "Managerial Structures in France," in The Evolution of International Management Structures , ed. Harold Williamson (Newark, Del., 1975), pp. 95-113.

vice or information. Yet centralization was not comprehensive. The outstanding characteristics of the PGC's authority structure were the dispersal of decision making and the unclear lines of power An uneven pattern of centralization and decentralization quickly evolved. Both principle and pragmatism gave it a peculiar shapelessness that partook of none of the Cartesian clarity the statutes described.

Managers informally expanded their authority into areas that superiors left for them, and the latter did not guard all their powers with a ferocious jealousy. The director reserved for himself the most noble activities,

appointing personnel, allocating rewards and sanctions, and authorizing expenditures. In these areas there was hardly any delegation of authority to relieve top management from involvement even in petty details. It hardly pays to conceive of a middle or lower management in these matters except as a transmission belt for decisions made at the top. Each December the director personally reviewed the records of every clerical worker and decided on annual bonuses and promotion.^[14] Heads of departments participated in these decisions only by providing written comments about employees. The chief of the lighting department lacked the right to grant even a simple leave of absence for a clerk in his office. Usurping that power earned the engineer a formal rebuke.^[15] Middle management did not select their own assistants or immediate subordinates; the decisions came from above. Only by restricting information available to a busy director or making strong recommendations could division chiefs assemble staffs of their choice.^[16] The power to spend was likewise highly centralized. The director ruled on the type of hand towels that would be purchased for the bathrooms of the headquarters. Not until 1872 did factory superintendents gain the right to spend twenty francs without prior approval from above.^[17]

By contrast, there soon came to be important spheres into which centralization did not extend. The director was actually quite removed from the daily operations of major departments. He appeared to know little about products and the ways they were manufactured and did not seek to learn about them. The company had no absolutely uniform wage or labor policies. By the 1860s factory superintendents had ceased reporting to the directors on a regular basis. In these circumstances centralization failed, not by default, but rather by conscious design. The director welcomed the benefits of delegated authority, at least in some spheres. Thus, when the opportunity to install telephones connecting the factories to headquarters appeared in 1879, the director vetoed the proposal on the grounds that the new device would undermine the responsibility belonging to his subordi-

[14] The decisions were formalized at one of the last annual sessions of the conseil d'administration and of the comité d'exécution.

[15] AP V 8 O no. 93, notification of January 9, 1869 (fol. 71).

[16] This informal input may explain how division heads were able to have graduates of their own schools appointed under them.

[17] AP V 8 O , no. 1081, ordre de service no. 20; no. 751, "Secrétariat." Heads of departments and divisions were forbidden to correspond directly with suppliers (ordre de service no. 98). Corporate rules made them get the director's approval for all maintenance projects or for any changes on approved projects (ordre de service no. 93).

nates. Not until twenty years later did the PGC install telephone service, and then only because strike threats imposed the need for immediate communication.^[18]

However amorphous the patterns of authority, they were not entirely the product of capriciousness on the part of successive directors. There was a logic, though not a consistent one. The centralization of budgetary and personnel decisions was in a sense ideological, reflecting the directors' conception of their sovereignty. Yet their aloofness from nearly all concerns at the factory level probably arose from pragmatic calculations about control over the labor force.^[19] Management did not want workers at each of their plants uniting against the company, and the best way to encourage fragmentation was to treat each factory as an autonomous unit. Furthermore, the director wished to strengthen the hand of superintendents over the labor force by granting his managers the status of masters in their own houses. This mixture of principle and practical calculation generated much corporate policy.

Inevitably, the untidy dispersal of authority gave middle and lower managers considerable responsibility.^[20] No document explicitly defined their duties. Their functions evolved as a result of experience, shared understandings, and personal initiative. Strategic, tactical, and operational decisions gravitated to executives at each level. All managers had to be active in four areas—daily operations, planning, financial oversight, and research. Even top management supervised or intervened in the details of daily routine. Most of the burdens fell on lower managers, however, and the PGC was probably understaffed at that level. The chief of the by-products department ran by himself an operation that involved revenues of some two million to three million francs and 250 workers.^[21] Several superintendents directed plants with nearly a thousand laborers and millions of francs in equipment with the assistance of at most one other engineer.

Since the PGC never centralized planning procedures, every manager had to divert his eyes from current concerns to consider the future. The director depended on department heads to recommend areas of expansion and new markets. The systematic preplanning of new factories was a task

[18] Ibid., no. 680, deliberations of October 31, 1879, and no. 695, deliberations of April 26, 1899. Superintendents sometimes disregarded formal directives from the chief executive. See, for example, no. 90, "Précautions prises au point de rue hy-giènique pour la boisson des ouvriers."

[19] See below, chapters 6 and 8.

[20] The thousands of reports scattered through the archives of the PGC shed light on their functions. Especially rich are cartons 709-719, 746-770.

[21] AP, V 8 O , no. 1060, "Usines-Traitement des goudrons et eaux ammonicales."

born of the industrial revolution, and it imposed a marathon of toil on the engineers. Any large capital improvement project involved corporate managers in inventing, testing, and perfecting equipment. When the PGC renovated its coal-handling machinery, nearly all new devices were designed by its engineers and constructed in its workshop.^[22] Moreover, every department chief acquired a thick file of correspondence with inventors, for they continually explored innovations that could be useful to their departments and examined the more promising ones in their laboratories.

Oversight of bookkeeping was an activity managers dared not neglect. It was a duty on which top management insisted more than any other, issuing sharply worded reprimands when necessary. The factory managers filed fifty-six reports each day on expenditures. Scrutinizing the books was required because honesty was a problem within the accounting staff. When an embezzlement scheme in the factory accounting office was exposed in 1873, it hurt the reputation of a plant superintendent, for the director charged him with insufficient supervision.^[23]

Apart from the duties imposed by current and future operations, managers undertook individual research and improvement projects. The engineers rarely left existing procedures as they found them; they devoted many of the workdays to tinkering with details of production. Although the PGC employed a consulting expert to oversee research (for many years Henri Regnault, a distinguished scientist and professor at the Collège de France), the endeavors lacked genuine coordination. Managers' projects were usually self-directed and not always intimately related to their immediate responsibilities at the firm. Nonetheless, the results could sometimes prove useful. The assistant chief of the factory division developed a crucial pressure-regulating device; the entire gas industry used the tables on gas flow calculated by the head of the factory division; a chief of the coke department was responsible for a new type of street lamp.^[24] The

[22] On the preplanning of factories, see Sidney Pollard, The Genesis of Modem Management (Cambridge, Mass., 1965), p. 262. J. Laverchère, Manutention mécanique du charbon et du coke dans les usines de la Compagnie parisienne du gaz (Paris, 1900), illustrates the work of the engineers in designing and building their own equipment.

[23] AP, V 8 O no. 161, "Etat des pièces. . ."; no. 675, deliberations of April 24, 1872; no. 676, deliberations of May 3, 1873, and January 17, 1874; Rapport, March 28, 1874, pp. 39-41.

[24] Bulletin de l'Association amicale des anciens élèves de l'Ecole Centrale 35 (1903-1904): 232; AP V 8 O , no. 677, deliberations of October 7, 1874; no. 672, deliberations of June 26, 1867; no. 1060, report of October 10, 1859; Philippe Delahaye, L'Eclairage dans la ville et dans la maison (Paris, n.d.), pp. 151-152; Alphonse Salanson, "Ecoulement du gaz en longues conduites," Société technique de l'industrie du gaz en France. Neuvième congrès, 1882, pp. 157-164.

managers' research agendas demonstrated a disinclination to specialize, which was all well and good given their multifarious duties.

The responsibilities assigned to each post and the career paths of the managers did not allow them to concentrate exclusively on a narrow set of technical matters. Engineers commonly moved from one department or division to another in the course of their careers. Eugene de Montserrat began as an assistant factory superintendent after graduating from the Ecole Centrale des Arts et Manufactures; he then became, successively, head of the secretariat, chief of the coke department, and chief of the distribution division. Plant superintendent Leroy had behind him several years in by-products production and in the machine shop. Louis Dhombres became an assistant factory superintendent after finishing the Ecole des Mines, moved to the coal-testing laboratory, and was promoted to the head of the mounted main office.^[25] Even when serving in one post, these managers could not confine their purview to engineering questions. They had to deal with personnel problems and marketing matters, among others. The workday for the head of the by-products department might have included drafting a report on the proper way to clean a coal-tar filter and then a memorandum on the sale of ammonium sulfate to farmers in Picardy. The director did not hesitate to ask the chief of the distribution division for advice on legal strategies regarding a gas explosion, and the latter did not hesitate to respond. This engineer also had to make business calculations about which streets would be profitable to service.^[26] The analysis of budgets or capital improvement options was an integral part of the managers' duties, as they usually lacked the staff to handle the financial questions for them. Being a part of a large-scale organization—and only a handful of French corporations were larger than the PGC—imposed only a limited degree of specialization on the managers.

Such duties left little room for leisure and obscured the distinction between free time and work. Managers did task-oriented labor, and there were no fixed hours on the job. Factory superintendents and their assistants lived at the work site and were on duty at all times. The director, the head of the factory division, and the chief construction engineer met on Saturday mornings to examine new projects. Sunday labor, at least on

[25] In the absence of personnel dossiers for managers, one must follow their careers through the deliberations of the conseil d'administration , AP, V 8 O , nos. 723 -737.

[26] Ibid., no. 626, "Sulfate d'ammoniaque"; no. 92 (fols. 126-130); no. 828, "Accidents." The head of the secretariat wrote reports on the marketing of coke; see no. 751, report of June 18, 1858.

personal projects, may have been normal. When the owner of the Rossini Theater bestirred himself early one Sunday morning to have gas restored to his establishment, he found the chief of the lighting department at his office. The chief sent the impresario to see the director, who was also at work and assured him that both executives would be on the job all day.^[27] As far as can be discerned from the incomplete records, vacations were not part of the annual calendar. Starting in the late 1880s there were some requests for a week's leave during the slow summer season, usually disguised as health-related absences, but these came from a minority of engineers. It was not only for wage earners that management conceived of hard work seven days a week, fifty-two weeks a year, as natural and inevitable.

Beyond constant application the company also demanded a rare combination of moral qualities from its executives. They had to be eager to accept new challenges and grow on the job. The self-confidence to make decisions on matters that were not susceptible to technical analysis was a basic requirement of their work, and it was a virtue they did not seem to lack. But self-confidence had to be tempered by self-effacement when necessary. The company was not especially sensitive to the egos of its executives and assumed their pride was expendable. The director once sent the assistant factory chief, Edouard Servier, to examine an electric pressure regulator that competed with his own invention. The mission was still more delicate in that his competitor had accused Servier of copying. Yet his superior found that Servier had nonetheless carried out the task with "the most perfect impartiality."^[28] Clearly the company expected much from its engineer-managers, and these men, as a result of their backgrounds, expected much from themselves.

The Social Composition of Management

Through its early boards of directors the PGC grew under the casual supervision of some of France's most original and innovative entrepreneurs. Among these were the fourteen principle owners, partners, and directors of the gas firms that had fused to form the company. These pioneers of the gas industry in France had not been members of the conservative Orléan-ist business aristocracy that safely dominated the court and banking circles. Some were clearly outsiders who won great wealth despite, or perhaps as a consequence of, breaking rules. Jacques and Vincent Dubochet,

[27] Ibid., no. 766, report of May 4, 1870; no. 1081, ordre de service no. 180.

[28] Ibid., no. 1070, report of September 25, 1861.

owners of the former Parisian Gas Lighting Company, were born in Switzerland (Vevey) and were well known for their republican principles. They had participated in Carbonari conspiracies in their youth and became wealthy patrons of Léon Gambetta during the Government of Moral Order. Gambetta received financial support from the Dubochets for newspapers and campaigns, most notably during the May 16 (1877) crisis. Vincent's death in 1877 was marked by two distinctions: Leading Opportunist republicans were at his funeral, and he died one of the wealthiest men in France, with an estate worth more than thirty-five million francs.^[29]

Louis Margueritte, the principle architect behind the merger forming the PGC and the owner of the largest pre-merger firm, was another outsider. Born into a family of well-off Rouennais merchants in 1790, Mar-gueritte developed a serious interest in both industry and theater. He claimed to have written a tragedy that the Comédie-Française produced in 1824. Though he entered the gas industry the very next year, his ties to the arts did not dissolve. In fact, he married one Mademoiselle Minette, a former actress at the Vaudeville Theater. He eventually came to own huge estates outside Paris and died one of the largest landlords in the Seine-et-Oise. His fortune was rumored to be eighty million francs.^[30]

Thomas Brunton, another board member and a partner to the merger, was born to English parents in 1793. His father had made a fortune by taking British methods of cotton spinning to Normandy. The Revolution ruined the business, and his father went to prison during the Terror. Brunton's status as an alien was reinforced by his habit—inexcusable to some proper-thinking Frenchmen—of calling himself an engineer even though he lacked a diploma from an appropriate French school. The Imperial administration declined to recommend Brunton for the Légion d'honneur because he did not "enjoy much consideration among industrialists."^[31]

The pioneers of the gas industry were joined on the board by some of the men most responsible for the French "industrial revolution" of the Second Empire. The largest investors in the PGC were Emile and Isaac Pereire, the "best representatives of Saint-Simonian dynamism in service

[29] I. P. T. Bury, Gambetta and the Making of the Third Republic (London, 1973), pp. 111, 307-308, 414, 441. On Dubochet's wealth, see AP, D Q⁷ , nos. 12387-12389, 12396. Jean-Marie Mayeur, Les Débuts de la III République (Paris, 1979), p. 50, refers to Dubochet as the "Mécène des républicains."

[30] Archives nationales, F¹² 5201. On hearsay regarding Margueritte's fortune, see Maurice Charanay, "Le Gaz à Paris," La Revue socialiste 36 (1902): 433.

[31] Archives nationales, F¹² 5098.

of the Imperial economy" according to Guy Palmade.^[32] Though the gas company was less pathbreaking and historically significant than the Pe-reires' Crédit mobilier, one of Europe's first industrial banks, it proved more enduring and profitable. The Pereires brought their circle of business associates and fellow investors to the PGC's board.^[33] Hippolyte Biesta, the director of the Comptoir d'escompte and a collaborator of the Pereires in creating the Crédit mobilier, was on the first board. Alexandre Bixio, who served on the board from 1855 to 1865, also sat on the boards of the Credit mobilier, the Railroad of Northern Spain, and the General Transportation Company, all Pereire projects. Emile's son-in-law, Charles Rhoné, was also a board member. These associations emphasized the ties of the PGC to the men who were shaking up the French economy during its most dynamic era and to other pillars of the new corporate capitalism.^[34]

The Pereire circle continued to encompass most of the leading nationally connected entrepreneurs on the PGC's board even after the failure of the Credit mobilier (in 1867) and the humbling of the family.^[35] In spite of its size and profitability the company never succeeded in forging links to other great names of French capitalism, like Paulin Talabot, Henri Ger-main (of the Crédit lyonnais), Paul-Henri Schneider (of Le Creusot) or the Rothschilds. Only one representative of France's financial aristocray, the Haute Banque, sat on its board—André Dassier.^[36] As the significance of the Pereire group faded on the national scene, the PGC's board lost its entrepreneurial luminaries. In 1864 at least thirteen of the twenty board members served on the boards of other large and important corporations.

[32] Guy Palmade, French Capitalism in the Nineteenth Century , trans. Graeme Holmes (London, 1972), p. 130.

[33] On this circle, see Robert Locke, "A Method for Identifying French Corporate Businessmen," French Historical Studies 10 (1977): 261-292, and lean Autin, Les Frères Peteire (Paris, 1984).

[34] Charles-Joseph-Auguste Vitu, Guide financier: Répertoire général des valeurs financières et industrielles (Paris, 1864), lists the boards of directors of large firms. Note that reference works on French entrepreneurs, even important ones, hardly exist.

[35] AP, V 8 O , no. 726, deliberations of April 2, 1868. Emile and Isaac Pereire resigned from the board in April 1868, but members of the younger generation remained.

[36] Pierre Dupont-Ferrier, Le Marché financier de Paris sous le Second Empire (Paris, n.d.), p. 70. David Landes, Bankers and Pashas (Cambridge, Mass., 1958), p. 13, notes that "there was hardly a corporation of any importance [in France]— canal, railroad, or public utility—that did not feature among its founders and on its board the names of one or more of these few firms who formed . . . the Haute Banque." The PGC fit his description, but just barely and not for its entire life.

By 1878 only five of the PGC's administrateurs sat on other boards, and these were largely a legacy of the Pereire connections.^[37]

The businessmen who replaced the Pereire group or sat alongside its remaining members were well-off Parisians, grands bourgeois, but not captains of industry or finance. The banker Charles Mussard had an estate worth 690,000 francs at his death; he was neither a Pereire nor a Du-bochet. Jules Doazan, a stockbroker (agent de change ), possessed just under a million francs. Athanase Loubet was an important merchant and a former president of the Parisian Chamber of Commerce, but his fortune of 1.8 million francs did not give him the stature of a department-store magnate.^[38] Under the Third Republic the PGC increasingly lost its ties to other great corporate enterprises.

Instead, the PGC recruited to its board ever-larger numbers of distinguished scientists, state officials, and administrators to replace business leaders. The representation of scientific expertise became rather formidable. One of France's leading chemists, Henri Sainte-Claire Deville, joined the board in 1874.^[39] Louis Troost, professor of chemistry at the Sorbonne, became president of the corporation and one of the more active hoard members. Eugène Pelouse, an applied chemist, came to the board after developing a widely used condenser for coal-gas production and finding new ways to use by-products.^[40] There were also a member of the In-stitut de France and a vice president of the French Geological Society. Whereas these notables were eminently qualified to examine the firm's technical procedures, several of the men who joined them on the board had occupied important positions in state administration. The general inspector of mines, Meugy, came to the board in 1880. A former director of the postal service, Baron, served with him as did two former councillors of state. The stockholders confirmed the trend toward reduced ties with other corporations by placing on the board the retired director of the PGC and the current director.^[41]

The shifting profile of the board of directors, from economic movers and shakers to administrators, mirrored the declining vigor and aggressiveness of the PGC's entrepreneurial policies (see chapter 4). Yet we

[37] Vitu, Guide financier ; Alphonse Courtois, Manuel des fonds publics et des sociétés par actions (Paris, 1878).

[38] AP, D Q⁷ , nos. 12371, 10710, 10714, 12342.

[39] Harry Paul, The Sorcerer's Apprentice: The French Scientist's Image of German Science, 1840-1919 (Gainesville, Fla., 1972), pp. 77-78 ; L. F. Haber, The Chemical Industry during the Nineteenth Century (Oxford, 1958), p. 77.

[40] Archives nationales, F¹² 5231.

[41] Appointments to the board were announced in the deliberations of the conseil d'administration. In most cases information available on these people was slender.

should not make too much of the parallel. There were more important reasons for the growing caution of the corporation, not the least of which was the approaching end of its charter (giving heavy, new investments less opportunity to pay off). Moreover, patterns of aggressiveness and caution in business policies were not clearly confined to a distinct phase of the firm's life, and it is far from certain that the board had more than a passive impact on decision making.^[42] The changing composition of the board may well have reflected wider trends in the French economy rather than changes within the corporation. After all, Paris had ceased to be the center of commercial and financial innovation that it had been during the 1850s. Great names in business were fewer and farther between. The depression of the 1880s hit the economy of France harder and longer than those of other countries. When France resumed vigorous economic growth at the beginning of the new century, the innovative leaders were specialists, like Louis Renault, who confined their activities to one firm.^[43] Such figures would not have considered the gas industry as marked for special growth in any case and might have turned to the infant electrical industry. The last boards of the PGC were well suited to the task at hand—finding a technological niche and adapting to new corporate responsibilities as a public service and as a model employer.

<><><><><><><><><><><><>

The central figure and animating force in the PGC was the director. After owner-entrepreneur Vincent Dubochet took the post for its first eighteen months, it went to a salaried manager. The PGC was one of the early firms to initiate a practice that was to become a distinctive mark of French capitalism. It sought its chief executive not among the subordinate officers already in the firm, nor among the heads of comparable firms, but rather in the civil service. After Dubochet the directors of the PGC were all engineers of the Corps des Ponts et Chaussées.

The corps was charged with overseeing and improving the nation's infrastructure, and its engineers were indisputably public servants of high rank. They had diplomas from France's most distinguished and exclusive school, the Ecole Polytechnique. Furthermore, the corps accepted only the polytechniciens who had graduated at the top of their classes. The corps in effect comprised a post-Napoleonic aristocracy. Recruited through rig-

[42] Only two board members had enough of an attachment to the PGC to leave substantial legacies to its personnel. Those members were Germain Hervé, an early entrepreneur in the gas industry, and Raoul-Duval, an engineer-entre-preneur and a polytechnicien.

[43] Maurice Lévy-Leboyer and Francois Bourguignon, L'Economie française au XIX? siècle: Analyse macro-économique (Paris, 1985), pp. 78-84; Palmade, French Capitalism, pp. 187-216.

orously competitive examinations, the engineers who were admitted came overwhelmingly from haut-bourgeois families. The purpose of their long years of preparation and demanding careers was service to the state. Moreover, members were bound by powerful codes of collegial loyalty and personal honor. Wearing the distinctive military uniform at all times was obligatory for the engineers of the corps. The dignity of the corps was an ideal around which the members had to organize their lives. Aristocrats of the Old Regime had had to accept an occasional mésalliance; there was nothing they could have done about the waywardness of individual blue bloods. However, the engineers of the Corps des Ponts et Chaussées had to submit their marriage plans for approval to their director general.^[44]

Preparation for the corps marked the engineers for life. Each officer was the product of about fifteen years of cloistering—first nine years at the lycée, then six years at the Ecole Polytechnique and the Ecole des Ponts et Chaussées. The latter two were run on a military model, and minute regulations governed the details of the students' lives. Vacations were short and infrequent; students spent little time outside the school. Hazing and deeply rooted customs inculcated a strong corporate identity.^[45] At the same time, the rigors of the selection process and of the training bred a sense of elitism and authority. Contemporaries noted in the officers that emerged from this formative experience a distinctive comportment and even a distinctive way of thinking. No wonder the chief executives of the PGC always identified themselves first as "engineer of bridges and roads" and only then as director of the firm.^[46]

The PGC was one of the earliest private enterprises to take engineers out of state service and place them in corporate management. The path by which these post-Napoleonic aristocrats came to accept—even welcome— the new career opportunities is worth examining. Members of the corps enjoyed great prestige; they were admired for their learning, for their expertise, and for the weighty matters they handled. There was hardly a family in France that would not have taken pride in having a member in the corps. The daily existence of an officer, however, was usually mundane. The pay was modest, barely enough to sustain bourgeois standards

[44] On the corps, see E Fichet-Poitrey, Le Corps des Ponts et Chaussées: Du Génie civil à l'aménagement du territoire (Paris, 1982), and A. Brunot and R. Coquand, Le Corps des Ponts et Chaussées (Paris, 1982).

[45] Fichet-Poitrey, Corps, pp. 33-34; John Weiss, "Bridges and Barriers: Narrowing Access and Changing Structure in the French Engineering Profession, 1800-1850," in Professions and the French State , 1700-1900, ed. Gerald Geison (Philadelphia, 1984), pp. 30-40.

[46] Brunot and Coquand, Ponts et Chaussées , p. 133. The directors turned over 5 percent of their salaries to the state for their pensions as engineers "on leave."

unless subsidized by private wealth. The engineers' duties could often be routine or at least subject to bureaucratic roadblocks. It is easy to imagine that there were discontents, and Honoré de Balzac, that preeminent interpreter of the bourgeois soul, left a memorable literary portrait of a troubled engineer in Gérard of Le Curé du village (based perhaps on his own brother-in-law). Gérard emerged from the Ecole Polytechnique with aspirations for a brilliant career and a yearning for la gloire. Instead, his career, though honorable, brought him mainly hard work, routine, and low pay. He came to dread a future consisting of "counting pavement stones for the state" and waiting for a small promotion every few years.^[47] Balzac's speculations on the psychology of the state engineer were dramatic, to be sure, but not necessarily accurate. There was little evidence of a crisis of morale within the corps. Balzac surely underestimated the engineers' commitment to hierarchy, discipline, and service to the state.^[48] Individual officers may have despaired, but they did so privately. Outwardly the corps projected a sustained attachment to its responsibilities and to its acquired status. Defections from state service were rare. Not until 1851 did the war minister find it necessary to issue a decree regulating permanent leaves of absence, allowing engineers to take outside positions. Even so, officers did not often take advantage of the regulation until the 1880s.^[49] The lack of opportunities may have had much to do with the hesitations. The directorship of the small firms characteristic of the early industrial era was not suitable for men of their talent and standing even if the material rewards were attractive. But the emergence of large-scale enterprise—mines, railroads, machinery construction, and utilities—offered them a lucrative alternative that they might accept as appropriate. The appearance of firms like the PGC allowed this administrative elite to reach out and capture new positions entailing considerable economic power.

Dubochet's successor as director, Bridges and Roads Engineer Joseph de Gayffier, may not have married an actress—the corps would never have allowed that—but he was a trailblazer in his own way. De Gayffier was one of the early members of the corps to leave state service and take a position in private industry. In doing so, he helped create a model that

[47] Ibid., pp. 138-141.

[48] Terry Shinn, L'Ecole Polytechnique , 1794-1914 (Paris, 1980), p. 181, on the mentality of the graduates of the institute. For an assessment of the bourgeois "soul" that differs substantially from Balzac's, see Theodore Zeldin, France , 1848-1945, 2 vols. (Oxford, 1973-1977), 1:11-130. Zeldin stresses the limited ambitions of most French bourgeois.

[49] Brunot and Coquand, Ponts et Chaussées, p. 257; Shinn, Ecole Polytechnique, pp. 94, 167.

would eventually become commonplace in French large-scale industry. In 1858 he was still an exception and perhaps had something of Gérard in him. Born to a wealthy Auvergnat family in 1806, Joseph became a student at the Ecole Polytechnique and at the Ecole des Ponts et Chaussées by dint of hard work and uncommon intelligence. When he finished the rigorous training, he began the long, slow climb up the official rungs of the corps. By the age of forty he was still only a second-class engineer earning forty-five hundred francs a year and had already served in the departments of Indre, Somme, Oise, and Côtes-du-Nord. The modest pay and difficult work may have interested de Gayffier in more lucrative endeavors outside the corps. His subsequent employment showed how both the engineers and the corps would have to adapt to new career patterns.^[50]

De Gayffier's first attempt to serve private enterprise went smoothly. In 1845 he left the corps briefly to be director of a public-works firm in Portugal and readily received a leave of absence for that purpose; he was back in the corps within two years. The next involvement with a private firm proved more complicated. De Gayffier asked for another leave in 1856 so that he could take a post with the Grand Central Railroad. His superiors at Bridges and Roads were reluctant to grant the request on the grounds that the position involved directing operations that had been subcon-tracted to another firm. The officials insisted that "the reputation of the corps requires that the position of engineers who take leaves from its ranks . . . must be perfectly defined and must present nothing untoward in the eyes of the public." Clearly the corps was trying to set lofty standards for outside employment, whereas de Gayffier sought to extend those limits· He finally worked out an acceptable definition of the post and received the authorized leave. During his employment with the railroad he earned six times as much as the corps would have paid him.^[51]

The merging of the Grand Central Railroad with a larger line placed de Gayffier in a new predicament. He lost his position but was no longer willing to return to the corps· He was reprimanded for not reporting to his assigned post when the leave ended. Soon he suffered the further humiliation of being passed over for promotion to first-class engineer. He apparently spent two years in this ambiguous situation before the offer from the PGC arrived. Perhaps an exceptionally disgruntled state engineer was the only sort that private enterprise could attract at that time.

Why board members of the PGC selected de Gayffier to direct the firm

[50] Archives nationales, F¹⁴ 2233¹ .

[51] Ibid. See especially Conseil général des Ponts et Chaussées, deliberations of August 7, 1856.

is not clear. His assignments as an engineer had prepared him for work with railroads. He had planned canals, dredged harbors, and supervised the laying of track, but he had no experience whatsoever with the gas industry. Moreover, his reputation within the corps had its blemishes; internal reports described his character as "inconstant" and the quality of his work as "mediocre." Nonetheless, like most of his colleagues at Bridges and Roads, de Gayffier had had experience administering large projects. It is possible that de Gayffier had made important contacts when he worked with the Grand Central Railroad, for that was a Pereire undertaking.^[52]

Still working out the terms by which state engineers would serve private enterprise, the directors of the corps did not approve de Gayffier's new post without some clarification. The officials were concerned that this position would associate him too closely with commercial operations, considered beneath the dignity of the corps. They had to receive assurances that de Gayffier's functions would entail the oversight of a large corporation serving a useful public purpose and that baser matters, such as purchasing coal, would be the responsibility of his subordinates. With those assurances de Gayffier was allowed to begin his thirteen-year career with the PGC (1858-1871). Though he was undoubtedly pleased to earn many times the salary of a Bridges and Roads engineer, his status in the corps did not cease to weigh on his mind. Apparently self-conscious about being only a second-class officer, he campaigned for a promotion even as he assumed the directorship of the PGC.^[53]

Emile Camus succeeded de Gayffier as director in 1871 and remained at the helm of the firm for the next twenty years. The son of a prosperous notary from Charleville (Ardennes), Emile had entered the Ecole des Ponts et Chaussées after graduating fourteenth in his class from the Ecole Polytechnique. Like his predecessor, he had held several positions in private firms before his employment with the PGC, and none had been in the gas industry. In 1858 he had taken leave from the corps to head a firm that transformed more than three thousand hectares of marshland around Mont-Saint-Michel in Brittany into arable land. This demanding project won high praise for its technical accomplishments. In 1860 Camus was named assistant director of the PGC. How Camus came to the attention of the gas company is not, in this instance, a mystery: he had married a

[52] Ibid., report of prefect to minister of commerce, January 3, 1855; undated fiche from Corps des Ponts et Chaussées.

[53] Ibid., report of March 24, 1859; prefect to minister of commerce, July 30, 1860.

relative of de Gayffier in 1852 and had worked under his direction at the Grand Central Railroad in 1856.^[54] In fact, the last director of the PGC, Léon Bertrand (appointed in 1901), was also a relative of Camus and de Gayffier.^[55] The company was thus under the control of one "dynasty" for thirty-eight of its fifty years, showing the narrowness of the search for a chief executive officer. De Gayffier's successors also owed him gratitude for helping create the precedent of leaving the corps to serve a profit-making enterprise. Camus did not face questions from his commanders in the corps about the appropriateness of his position; he received a leave of absence as a matter of course.

Stéphane Godot, who became director in 1892 (serving until 1901), was the only one of the PGC's chief officers to have displayed any personal rebelliousness or an inclination to think critically about wider social questions. As a twenty-year-old student at the Ecole Polytechnique he committed an (unknown) offense that resulted in his expulsion. Receiving indulgence from the war minister, he was readmitted but dropped from ninth to twenty-fifth place in the class. Perhaps as a result of the chastising experience, he redoubled his efforts as a student at the Ecole des Ponts et Chaussées and graduated second.^[56] The rest of his life was not a model of passive conformity, however. He became interested in the school of thought inspired by Frédéric Le Play and developed a reputation as an industrial paternalist. When the PGC faced a new era of industrial relations in the early 1890s, Camus turned the post of director over to Godot. The union leaders took the change as a conciliatory measure. He resigned in 1901, ostensibly for reasons of health, when a dialogue about industrial reform was no longer necessary.^[57]

Godot's successor, Léon Bertrand, showed no such signs of rebelliousness or questioning. He was the son of a professor at the Ecole Polytech-nique who was also an immortal of the Académie française. Léon obtained only enthusiastic praise from teachers and superiors: his character was "excellent," his work habits "irreproachable."^[58] In a sense Bertrand marked the completion of an evolution among the PGC's leaders, from nonconforming entrepreneurs to unconventional state engineers to model officers of France's most distinguished technical corps.

Each of these engineer-directors was part of a national elite, not only

[54] Archives nationales, F¹⁴ 2185¹ ; F¹² 5101.

[55] I have been unable to discover the precise relation among all three engineers, but an elegant tomb in the Père-Lachaise cemetery attests to the alliance of their families.

[56] Archives nationales, F¹⁴ 11481.

[57] See below, chapter 4, on Godot's relation to the gas personnel.

[58] Archives nationales, F¹² 8516; F¹⁴ 11520.

as a result of his own membership in the corps but also through family membership.^[59] A generation or two earlier, their progenitors had been successful professionals. Camus's father had been a notary, Bertrand's grandfather a physician. The path of further ascent had led through the grandes écoles and from there to high state posts. The directors' family trees contained numerous maîtres de requêtes and auditeurs of the Council of State, judges, and professors. Relatives in banking or business were fewer. Among the last group were those who had also "parachuted" from a state corps into industry. Official documents were always able to categorize the directors' parents as well-off, but huge fortunes were rare. De Gayffier's personal estate of 632,000 francs was probably typical. Though substantial, it was not the wealth of a Dubochet or even of a board member like the banker Dassier, who died with a fortune of more than four million francs. By moving into the PGC, engineer-directors added sizable income to the prestige and power their relatives enjoyed as hauts fonctionnaires .^[60] Perhaps the lure of lucre was not so powerful among most of their cob leagues who remained in the corps, or perhaps desirable opportunities were lacking. As we have seen, the PGC kept its search for chief executives within narrow limits.

Was the Corps des Ponts et Chaussées the most appropriate source of leadership for the PGC? It is hard to identify specific qualities and training that made these engineers essential to the firm. Their education, highly abstract and oriented toward mathematics, did not ensure a technical grasp of industrial problems. The most valuable asset they possessed was the ability to deal as equals with the other graduates of the grandes écoles they were likely to meet in the course of doing business. As polytechni-ciens, they could address the prefect or municipal engineer as mon cama-fade and use the informal tu .^[61] Such standing was worth something, but the deepest reason that the PGC's board turned to the corps to find its chief executive was no doubt a conventional, uncritical respect for hierarchy. State and society defined these men as the nation's administrative elite.

<><><><><><><><><><><><>

A clear sign that the Ecole Polytechnique and the state's technical corps did not produce men with precisely the necessary preparation to direct gas production was that the PGC recruited the engineers in charge of opera-

[59] Ezra Suleiman, Elites in French Society (Princeton, 1978); John Armstrong, The European Administrative Elite (Princeton, 1973).

[60] AP, D Q⁷ , 12341 (fols. 56-57); Shinn, Ecole Polytechnique , p. 90.

[61] For an interesting illustration of the right to use informal forms of address as well as of the inevitable contacts made at the Ecole Polytechnique, see Ibid., no. 155, Fontaine to Godot, April 26, 1902.

tional departments from other schools. Heads of divisions, departments, and factories were graduates of the Ecole Centrale des Arts et Manufactures (founded in 1829) or the external classes of the Ecole des Mines.^[62] Both of these institutes had assumed the task of training industrial leaders. Though they lacked the supreme prestige of the Ecole Polytechnique, they stood just beneath it. The bourgeois family that placed a son in one of these state schools had reason to be pleased. Industrialists often chose one of these alternatives over the Ecole Polytechnique for their heir if he was serious about continuing in business.^[63]

The Ecole Centrale and the external classes of the Ecole des Mines, with their industrial mission, were products of the emerging factory era. So were their graduates. Just as the early directors of the PGC departed from convention by leaving state service and creating a new role for themselves at the helm of large firms, lower managers were creating their own social persona. The PGC was born at the moment when engineer-managers of industrial firms were appearing. Enrollment figures at the two schools attest to the rapid expansion of the milieu (table 5). No more than the PGC's director did the heads of departments or plants build their identity around business administration. Their professional designation was "industrial engineer," a new occupational title that came into common use around the time the PGC was founded. Earlier all who used the title "engineer" had been members of a state corps. The birth of civil or industrial engineering as an established occupational category can perhaps be dated from 1848, when the Society of Civil Engineers was founded.^[64] Managers at the PGC

[62] The term "external classes" needs explanation. Technically, the Ecole des Mines was an extension of the Polytechnique. It enrolled two sorts of students. The "student-engineers" were graduates of the Polytechnique, usually top-ranking ones. These students were destined for the Corps des Mines, which, like the Corps des Ponts et Chaussées, was a highly prestigious body of state engineers. The Ecole des Mines also admitted "external" students, men who had generally not attended the Polytechnique and had gained admission through an examination. These students were not eligible for the Corps des Mines and were being trained for industry.

[63] The early history of the Ecole Centrale is well served by John Weiss, The Making of Technological Man: The Social Origins of French Engineering Education (Cambridge, Mass., 1982). See also Louis Guillet, Cent ans de la vie de l'Ecole Centrale des arts et manufactures, 1829-1929 (Paris, 1929). On the Ecole des Mines, see Louis Aguillon, L'Ecole des Mines de Paris: Notice historique (Paris, 1889), and Gabriel Chesneau, Notre école: Histoire de l'Ecole des Mines (Paris, 1932). The external classes of the Ecole des Mines were not at first so exclusive nor so rigorous as classes at the Ecole Centrale but eventually became much more demanding. Not until its later years did the PGC recruit heavily from Mines.

[64] Terry Shinn, "From 'Crops' to 'Profession': The Emergence and Definition of Industrial Engineering in Modern France," in The Organization of Science and Technology in France , 1808-1914, eds. Robert Fox and George Weisz (Cambridge, 1980), pp. 184-210; John Weiss, "Les Changements de structure dans la profession de l'ingénieur en France de 1800 à 1850," in L'Ingénieur dans la société fran-çaise, ed. André Thépot (Paris, 1985), pp. 19-38.

benefited from its growing social acceptability. Already in 1856 Gustave Flaubert was able to have Charles Bovary's assertive mother entertain aspirations that her son would be either a judge or a civil engineer

The sorts of firsts that PGC engineers could achieve were well illustrated by the career of Louis Arson, the chief of the factory division for thirty-eight years and, as such, the principal operational manager Arson graduated from the Ecole Centrale in 1841 with the first diploma awarded in mechanical engineering. One of his early jobs was with a machine-construction firm that made the first French locomotives and the engines for the first transatlantic steamships. He was the first graduate of the Ecole Centrale to sit on its advisory board.^[65] Arson and his colleagues at the PGC were members of the generation that helped establish the social persona for France's industrial managers. It was certainly not the case, however, that the first generation imagined itself without governing norms and models. Only by adapting some of the formal structures

[65] Archives nationales, F¹² 5082; "Discours de Mont-Serrat," Bulletin de l'Association amicale des anciens élèves de l'Ecole Centrale 35 (1903-1904): 98-102. On Arson's private life and family, there is a carton of interesting documents: AP, D E¹ , Fonds Lestringuez.

of the traditional professions did industrial managers gain social acceptance.^[66]

The distinctive features of the French engineering profession, as it had developed under the tutelage of the state, deeply marked the careers and work culture of the PGC's managers. One manifestation was the reconstruction of the professional hierarchy within the company in the form of closed castes. As we have seen, top managers were graduates of the most exclusive schools, the Ecole Polytechnique and the Ecole des Ponts et Chaussées.^[67] Middle and lower managers were recruited from the technical institutes of the next rank, the Ecole Centrale and Ecole des Mines. These graduates were excluded from top management regardless of their record of achievement on the job, but they had their own privileges. They monopolized the administration of divisions, departments, and factories. By contrast, men with diplomas from the least exalted of the engineering institutes, the Ecoles d'Arts et Métiers (gazarts as they were known), were relegated to modest posts. It was not that these schools failed to provide rigorous training and produce capable students; but the social level of recruitment was lower and the prestige less resounding. The officers of the PGC accepted the gazarts mainly for staff positions.^[68] Thus the PGC allocated managerial posts on the basis of criteria external to the firm and replicated the hierarchy of the engineering profession.

Notions about the ways managers should do their work came from the engineering profession as well. Managers inevitably looked to the engineers of the state corps to define their responsibilities and work culture.^[69] State engineers provided guidance on how a technically trained elite would function within a bureaucratic setting. They inspired ideals of au-thoritativeness, independence, and bureaucratic loyalty. The managers of the PGC, following the officers of the corps, eschewed specialization and readily delved into all aspects of administration, including nontechnical ones. Far from regarding mundane details as beneath them, they wel-

[66] Robert Anderson, "Secondary Education in Mid-Nineteenth-Century France: Some Social Aspects," Past and Present , no. 53 (1971): 125.

[67] The one exception proves the rule. Eugène de Montserrat, a graduate of the Ecole Centrale, was named assistant director in 1901. The PGC was liquidated before he could have moved up to the top post, if indeed that was a genuine possibility for him.

[68] C. R. Day, "The Making of Mechanical Engineers in France: The Ecoles d'Arts et Metiers, 1803-1914," French Historical Studies 10 (1978): 439-460. Not all large corporations relegated the gazarts to minor posts. See Claude Beaud, "Les Ingénieurs du Creusot à travers quelques destins du milieu du XIX? siècle au milieu du XX?," in Thepot, L'Ingénieur, pp. 51-59.

[69] On the functioning of the engineers of the corps, see Jean-Claude Thoenig, L'Ere des technocrats: Le Cas des Ponts et Chaussées (Paris, 1973), pp. 165-214.

comed involvement in every aspect of supervision. Yet even when they were directly involved in decision making, they posed as impersonal authorities who could evaluate a matter with detachment. The PGC's managers also emulated the state engineers by asserting a degree of independence from their immediate employer and identifying with a larger scientific community. They did so chiefly through their personal research projects, which were related only tangentially to their current position in the enterprise. The factory superintendent, Paul Biju-Duval, and the chief of gas production, Albert Euchène, published findings on the specific heat of iron and nickel. The assistant factory head, Edouard Servier, studied the chemistry of coal tar and made noteworthy empirical observations. His superior, Arson, turned out a steady stream of gadgets from his laboratory.^[70] Though there was a good deal of tinkering and research in progress, it was not usually coordinated or directed by the firm toward established goals. The projects reflected the personal interests of the managers, and they considered the materials, laboratories, and personnel of the company at their disposal to pursue their work. The company accepted such independence and even expressed pride in the accomplishments of its engineers when they won scholarly recognition. The only limit it sought to place on the independence of its personnel was prohibiting paid consultation for other firms.^[71]

Though there was this independent aspect to the managers' work culture, the impact of the French engineering tradition was to reinforce loyalties to the organization. Nurtured under the aegis of a powerful state while capitalism was still a weak motor of change, French engineers hardly had a chance to form an autonomous professional group with individual careers as the focus of professional life. Instead, state engineers imparted a sense of comfort with bureaucratic procedures, lifelong commitments to the organization, and ambiguity about the morality of the marketplace.^[72] The acceptance of hard work and modest rewards coupled with a respect for hierarchical authority that characterized the corps d'état certainly influenced the managers of the PGC. Thus, entitling themselves civil engineers was not a superficial affectation: they were making a statement about their cast of mind and their expectations on the job.

By the late nineteenth century some contemporaries viewed it as a pe-

[70] Bulletin de l'Association amicale des anciens élèves de l'Ecole Centrale 35 (1903-1904): 232; AP, V 8 O , no. 677, deliberations of October 7, 1874; no. 672, deliberations of June 26, 1867; no. 1060, report of October 10, 1859.

[71] AP, V 8 O , no. 666, deliberations of June 26, 1858.

[72] This analysis follows the thinking of sociologist Dietrich Rueschemeyer, Power and the Division of Labor (Stanford, 1986), pp. 122-124.

culiarity of French industry that engineers so dominated managerial posts in private enterprise. Though the monopoly of the profession was especially complete in France, the situation was not unique.^[73] In any case, as France's industrial prowess retreated before that of its neighbors, the dominance of engineers became a source of regret and anticipated reform. Industrial critics denigrated the training that French engineers received. The attacks began with the fossilized education at the apex, the Ecole Polytech-nique, on the grounds that its instruction was too mathematical, too theoretical, and contemptuous of the practical problems posed by industry. The PGC, like most firms, implicitly accepted such criticisms, for it drew its personnel from schools that were more specifically oriented toward the application of science. However, observers familiar with German technical training argued that the Ecole des Ponts et Chaussées, the Ecole Centrale, and the Ecole des Mines ultimately resembled the Ecole Polytechnique more than the German schools. All of the French institutes placed far too much emphasis on theoretical approaches and mathematical training; all relied heavily on lectures even though laboratory exercises were necessary for good instruction in the rising fields of organic chemistry and electricity; all led French engineers to develop general knowledge and eschew the specialization that would have been more beneficial to industry.^[74] The critics made some worthy points, but they postulated for the engineers a narrow, technical role in production. The responsibilities that managers held in a firm like the PGC did not justify the sort of specialization and laboratory training that the critics desired. Indeed, the partial concentration that French technical institutes did permit often proved superfluous because the graduates' assignments were unrelated to their academic specialties. The ideal of the elite schools, theoretical training aimed at rapid assimilation of new concepts, was not as outmoded as the critics maintained.

[73] Max Leclerc, La Formation des ingénieurs à l'étranger et en France (Paris, 1917). The dominance of engineers in France was not as unusual as Leclerc believed. See Heinz Hartmann, Authority and Organization in German Management (Princeton, 1959), p. 162; Jürgen Kocka, "Entrepreneurs and Managers in German Industry," Cambridge Economic History of Europe (Cambridge, 1978), 7(1):492-589; Chandler, Visible Hand, p. 95.

[74] Louis Bergeron, Les Capitalistes en France (1780-1914) (Paris, 1978), p. 70; Leclerc, Formation des ingénieurs; André Pelletan, "Les Ecoles techniques alle-mandes," Revue de métallurgie 3 (1906): 589-620; Antoine Prost, Histoire de l'enseignement en France, 1800-1967 (Paris, 1968), p. 303. The presumed deficiencies of French engineering education have been summed up and forcefully reasserted in Robert Locke, The End of Practical Man: Entrepreneurship and Higher Education in Germany, France, and Great Britain , 1880-1940 (Greenwich, Conn., 1984).

It is no wonder that the Ecole Centrale was proud of the versatility of its students.^[75] Generalized training for careers modeled on the state engineers was what managers wanted and needed.

A more pertinent criticism came in the early twentieth century from Henri Fayol, the founder of managerial science in France and one of the most respected engineers of his day.^[76] Fayol readily acknowledged the multiplicity of disciplines that managers had to master. For him, the deficiency in training arose from its exclusively technical character. Though engineers faced responsibilities for personnel, marketing, accounting, and financing, they received no formal instruction whatsoever in administration. Fayol refused to regard a capacity for mathematical reasoning as a sound basis for judgment in these areas. For decisions in these nontechn-ical fields the PGC followed the British model of empirical training on the job. Perhaps Fayol underestimated another quality that French technical training succeeded in imparting, a sense of confidence in handling the disparate responsibilities of business management. The engineer-managers of the PGC appeared comfortable with their multiple duties: their decisions may not have been especially keen, but they were consistent and conscientious. Such comfort with nontechnical fields may explain why the engineering profession ignored Fayol's challenge for so long.

The influence of the state institutes extended to staffing. In countries like Britain, where such institutes did not exist, filling managerial positions was a great problem. Owners refused to trust salaried executives to use their capital honestly and efficiently, so kinship and friendship played a large role in hiring.^[77] The engineering schools of France provided the guarantees of probity and expertise that personal familiarity did across the Channel. The PGC relied on the recommendations of the school directors more than any other source of recruitment. Faith in the excellence of the alma mater, reassurances provided by shared experiences, and established contacts made managers draw their new colleagues from their own schools. Factory chief Arson recommended one engineer after another from the Ecole Centrale. As his influence waned and Paul Gigot, his assistant, took charge, appointments gravitated to the Ecole des Mines, Gigot's alma mater.

More than excellence of training, the engineering institutes offered the

[75] Weiss, Making of Technological Man , p. 225.

[76] General and Industrial Management , trans. Constance Storrs (London, 1949). Fayol first published his influential tract in 1916, but it was based on talks given before the war.

[77] Pollard, Genesis of Modern Management , pp. 11-13.

PGC a cadre of managers with the proper attitudes and firm character. Despite minor variations in curriculum, all the institutes sought to impart an ideology of sobriety discipline, and assiduousness. Standards of accomplishment were very high, and the schools worked the students mercilessly. Consciousness of class rank, continuously reassessed, ritualized the view of life as a constant struggle. Any student who did not embrace hard work and persistent application left the program. Those who remained developed a sense of intellectual and moral elitism. For good reason one historian of the Ecole Centrale has labeled it a "factory of the bourgeoisie."^[78] Although not all graduates attained the schools' ideals, the PGC found among them a group of men with ample aptitude for science, a passion to apply it, and a commitment to the work ethic. The PGC took full advantage of the discipline inculcated at the Ecole Centrale and Ecole des Mines.

Satisfied with the elite schools and loyal to them, the PGC regarded the Third Republic's efforts to create new institutes of applied technology in the last two decades of the century with complete indifference. The company gave no support whatsoever to the creation of the Ecole de Physique et de Chimie Industrielle when it was founded in Paris in 1882 and hired only a few of its graduates for humble laboratory posts.^[79] The firm did not wish to make a place for specialists and technicians lacking elite diplomas in its managerial structure.

Careers and Rewards

The privileged family background of the young men who received diplomas from the grandes écoles is well documented. Not only the Ecole Po-lytechnique and its schools of application (Mines and Ponts et Chaussées) but also the Ecole Centrale and the external classes of the Ecole des Mines were preserves of the well-to-do bourgeoisie.^[80] The social antecedents of the PGC managers who were centraux conform to the rule. Their fathers

[78] Weiss, Making of Technological Man , p. 234; Weiss, "Bridges and Barriers," pp. 39-41.

[79] Ville de Paris, Cinquantième anniversaire de la fondation de l'Ecole de physique et de chimie industrielle de la ville de Paris (Paris, 1932). On another set of new technical schools, see Harry Paul, "Apollo Courts the Vulcans: The Applied Science Institutes in Nineteenth-Century French Science Faculties," in Fox and Weisz, Organization of Science, pp. 155-181.

[80] Maurice Lévy-Leboyer, "Innovation and Business Strategies in Nineteenth-and Twentieth-Century France," in Enterprise and Entrepreneurs in Nineteenth-and Twentieth-Century France, ed. Edward Carter, et al. (Baltimore, 1976), p. 108; Weiss, Making of Technological Man, p. 205; Shinn, Ecole Polytechnique , pp. 101,142.

were businessmen (often retired), liberal professionals, and gentlemen property owners (table 6). Though not all were extremely rich by the standards of the day, many were, and others offered good connections. Albert Ellison was the son of a property owner who lived on the genteel rue Madeleine (eighth arrondissement). The head of the meter department was the offspring of a Parisian banker. The uncle of another engineer was the director the the major newspaper, Le Siècle. Alexandre Arson's mother left a fortune of more than 350,000 francs to her children. The only conspicuously humble manager to emerge from the Ecole Centrale was a son of an employee at the firm; the company had provided a fellowship and hired him when his studies were completed.^[81] Thus, it seems that the emergence of large-scale industry diversified career patterns for the children of solidly established families but did not offer a source of advancement for the hardworking children of the common people. The question remains whether firms like the PGC gave these sons of the well-off the kinds of careers that allowed them to perpetuate their family status.

The PGC compensated its chief executive officer generously enough to have gratified Balzac's Gèrard. The directors of the PGC had annual salaries of twenty thousand to twenty-five thousand francs, and profit-sharing

[81] See sources for table 6. On Arson, see AP, D E¹ , Fonds Lestringuez, contrat de mariage.

bonuses doubled or even tripled the compensation.^[82] Camus, who presided over the firm during its acme of profitability, earned as much as eighty-two thousand francs. His average pay was just under seventy thousand francs, thirty-eight times the earnings of one of the firm's young office clerks. By contrast, when engineers of the Corps of Bridges and Roads achieved a long-awaited raise in 1906, their incomes were still modest. The state paid ordinary engineers (a position officers were likely to have after ten years of service) five thousand to seven thousand francs. The decree of 1906 gave chief engineers ten thousand to twelve thousand francs.^[83] Falling profits inevitably lowered the compensation of the directors who headed the PGC during the era of adjustment, but Godot (who presided from 1892 to 1901) still averaged fifty thousand francs a year, and Bertrand earned an average of forty-six thousand francs between 1901 and 1905. Unlike the chief engineers in the esteemed state corps, the PGC's directors received incomes to which grands notables were accustomed.^[84]

Most middle and lower managers had to accept less—usually much less. Their level of prosperity depended on whether they received profit-sharing bonuses as well. The PGC singled out six to eight middle and lower managers for the benefit. In 1880 the favored posts were the head of the factory division and his assistant, the chief of the distribution division, the engineer of the by-products department, the head of the coke service, the chief of the coal department, the senior accountant, and the head of the customer accounts office. The bonuses, ranging from twenty-five hundred to fifteen thousand francs, completed salaries comparable to those of chief engineers of the Corps of Bridges and Roads. Thus, the managers who received the bonuses were relatively well paid, though they may have had to wait a while to attain such compensation. Two different sorts of career pattern obtained (figure 8). A few managers were fortunate enough to head major departments for most or all of their careers and to benefit

[82] The annual compensation offered to each director (announced in the deliberations of the conseil d'administration ) was as follows: to de Gayffier, twenty thousand francs in salary plus an unspecified bonus based on profits; to Camus, twenty-five thousand francs in salary plus one thousand francs for each franc of dividends over twenty-five; to Godot, twenty-four thousand francs in salary plus one thousand francs for each million francs in profits; to Bertrand, twenty thousand francs in salary plus five hundred francs for each franc of dividends over twenty.

[83] Brunot and Coquand, Ponts et Chaussées ,p. 714; Weiss, "Changements de structure," p. 26-28.

[84] In 1869, more than half the deputies in a National Assembly dominated by notables had incomes of more than thirty thousand francs a year. The chief officers of the PGC were in their range. See Robert McGraw, France , 1815-1914: The Bourgeois Century (Oxford, 1986), p. 167.

Fig. 8. Annual Income of Managers, by Length of
Career (in Francs). From AP, V 8 O¹ nos. 665-700.

from large bonuses. During Arson's long incumbency as chief of the factory division he earned a salary of fifteen thousand francs (which remained constant for thirty-eight years) and bonuses that often came to that much or more. Paul Audouin became chief of the by-products department within ten years after leaving the Ecole Centrale. In 1870, as sales to the dyestuffs industry were about soar, the director granted him a profit-sharing bonus. For twenty-five of his forty-four years with the PGC he earned at least twenty thousand francs annually and as much as twenty-six thousand francs, far more than Bridges and Roads engineers could ever hope to earn. This son of a professor at the Jardin des Plantes had no material reason to regret he had not become a state engineer.^[85]

A more typical career pattern for the engineer-managers at the PGC entailed long years of moderate earnings before benefiting from high salaries and generous annual bonuses. After graduating from the Ecole des Mines, Paul Gigot began his career humbly, as a subinspector of construction at eighteen hundred francs a year. Seven years later, when his talents were recognized, he was a superintendent of the Vaugirard gas factory, but

[85] On Audouin's background, see Archives nationales, Légion d'Honneur 72, dossier 38.

he earned only five thousand francs. His promotion to assistant chief of the factory division in 1868 did not bring important material gains, though it did guarantee he would some day succeed Arson. In 1883 the company finally awarded him a small annual bonus, and ten years later he began to earn more than twenty thousand francs as head of the division. The long wait for a sizable salary also characterized the career of Charles Boissière. This graduate of the Ecole Polytechnique (1863) and the Ecole des Mines (1868) slowly worked his way up to head of the coke department after twenty-seven years. At that time his salary rose dramatically.

In spite of the need to wait long for high pay, Gigot and Boissière were among the fortunate minority of gas managers. Most of their colleagues did not do nearly so well, despite their hard work and their loyalty to the firm. Managerial salaries at the PGC were probably in line with those of other large firms, but the pay was by no means munificent.^[86] We may take as a benchmark the salaries paid to Bridges and Roads engineers in 1906 as well as the pay scale established by the city of Paris for its supervisory personnel in 1898. Paris paid the administrators who were roughly equivalent to lower-level managers at the PGC seven thousand to ten thousand francs a year. The city's departmental and division heads could earn sixteen thousand to eighteen thousand francs.^[87] The PGC compensated its managerial personnel comparably. Their income, then, did not at all reflect the company's notorious profitability. They were paid like civil servants.

Engineers' emoluments at the PGC were slow to diverge markedly from those of the office workers (table 7). This was partly because managers began at the same modest pay as clerks, eighteen hundred francs. The salaries also reflected the humble positions in which engineers began their careers despite their elite schooling—as draftsmen, secretaries to middle managers, assistants to assistant plant superintendents—for a few years at the very least. Even Boissière, a graduate of the Ecole Polytech-nique as well as the Ecole des Mines, spent three years as a draftsman with the company. When he had become chief of the coal department and was recommended for the Légion d'honneur, his file did not even mention that phase of his career. The centralien Théodore Bouffé began as nothing more than a chief foreman in the gas distillation plant.^[88] Engineers did not usually remain at such lowly posts for more than two or three years, but the

[86] For a comparison with Saint-Gobain Chemical Company, see Lévy-Leboyer, "Hierarchical Structure," p. 468; with Le Creusot, see Beaud, "Ingénieurs du Creusot," p. 54.

[87] AP, V 8 O , no. 153, "Arrêté préfectoral concernant l'organisation du cadre du personnel supérieur des services techniques."

[88] Ibid., no. 162, Pernolet to Minister of Interior, July 8, 1905; no. 678, deliberations of August 28, 1876.

position of assistant factory superintendent—none too glorious nor remunerative in itself—could be more enduring. Several managers were stuck as assistants for ten years or more. One graduate of the Ecole Cen-trale entered the PGC in 1879 and was still assistant superintendent of the Passy plant in 1902. Moreover, such unassuming starts were not usually purgatories inevitably leading to grander positions. For every Gigot, who eventually achieved a handsome compensation, there were three or four others who did not. Julien Vallet, also from the Ecole Centrale, entered the PGC a year after Gigot but did not have the same success. He retired as engineer of the construction department after thirty-one years, having attained a salary of twelve thousand francs. Still another centralien, Alfred Vialay, reached an income of only nine thousand francs as assistant head of the distribution division after thirty years with the PGC. Engineers of Bridges and Roads or of the city of Parris had little to envy these two; yet their situations were the typical ones. As table 8 suggests, promotions in the PGC were a slow, orderly affair at best. A few engineers never advanced out of technical-staff positions.

The gas company did not raise general salary levels significantly for managers except in compensation for heavier responsibilities. The average pay of assistant departmental and division chiefs was only 16 percent higher in 1902 (fifty-six hundred francs) than in 1858 (forty-eight hundred francs). Factory superintendents did receive significant increases (earning an average of fifty-one hundred francs in 1858 and ten thousand seven hundred francs in 1902), but in return the corporation asked more of them. Their raises reflected the burgeoning responsibilities entailed in an eightfold increase in gas production. The PGC apparently could escape having to raise salaries for its routine engineers because there was no

shortage of them. The state's engineering institutes grew faster than the mediocre pace of France's industrial development required. When the company needed a machinery inspector, the head of the service could produce a list of fourteen engineers who solicited the post. The contention that France's industrial growth was slowed by a lack of technical personnel is implausible.^[89] Of course, the dearth of opportunities did not necessarily reconcile executives to their modest pay. It is easy to suspect that some may have been disconcerted by the doubling of wage levels during the life of the firm while salary levels remained rather flat.^[90]

Director Godot learned to his surprise and embarrassment from an irate stockholder that some factory superintendents lived well at the company's expense. The PGC gave them a house on the factory grounds, and it came with a coachman, a gardener, and a servant. Forced to justify paying for the help (about which top management had been kept ignorant, no doubt through a conspiracy of silence), production engineer Albert Euch-ène argued that superintendents were graduates of "les Ecoles" and therefore required such amenities.^[91] Euchène may have been correct about engineers being accustomed to a genteel domesticity as a result of their family background. But most managers had to live soberly, indeed, if they existed purely on their own incomes. Prudent marriages were almost a necessity if they desired luxuries or needed to give their sons the same

[89] Ibid., no. 777, report of May 21, 1872 (fol. 20); Paul, "Apollo Courts the Vulcans," p. 180.

[90] See chapter 7 on wage levels and on the push for higher wages within the PGC.

[91] AP, V 8 O , no. 162, report of Euchène, January 25, 1896.

sort of expensive schooling as they had received.^[92] The head of the coke department, Emile Brissac, resided in an unpretentious apartment on the unfashionable fourth floor. His building was not even covered in the sculptured stone that denoted a luxurious residence. For all its modesty, the dwelling cost Brissac a fourth of his salary at the time. To have rented in one of the elegant edifices that lined Baron Haussmann's new boulevards might easily have cost half his earnings.^[93] Careers as engineer-managers proved uncertain routes to building an estate. Factory superintendent Louis Cheron died after twenty years of employment in the PGC with a fortune of sixty-three thousand francs. Of that sum, fifty-one thousand had come from an inheritance and from a wife's dowry.^[94] Only the fortunate minority of managers who attained large emoluments and profit-sharing bonuses could hope to build handsome estates. Arson had received only twelve thousand francs from his parents at his marriage but was able to give his daughter a dowry of sixty-one thousand francs.^[95] Director de Gayffier died with a fortune of 632,000 francs, ten times the value of Cheron's. We can be sure that he did not earn the wealth in the Corps des Ponts et Chaussées.^[96]

Not only was access to the highly paid posts limited, but most managers knew relatively early in their careers that they would never leave lower management. The PGC reflected French society at large in channeling a small core of candidates into the elite track at a young age and selecting them by criteria other than proven accomplishment on the job. Servier became assistant chief of the factory division, the chosen successor to the powerful Arson, within three years of graduating from the Ecole Cen-trale. His rapid elevation was not due to extraordinary promise as a student (he graduated thirty-first of 127 students) nor to the brilliance of his first years with the company. It was a matter of having the confidence of an influential member of the board of directors.^[97] Similarly, Euchène became the third-ranking engineer of the factory division only three years after leaving the Ecole des Mines because Arson saw special qualities in

[92] Shinn, Ecole Polytechnique , p. 52, estimates that the schooling of a serious candidate for the Polytechnique cost about five thousand francs a year. The cost of preparing for the Ecole Centrale or Ecole des Mines could not have been much less.

[93] AP, D 1 P⁴ (Cadastre for 7, Cité Malesherbes). At the time Brissac was earning six thousand francs and paid fifteen hundred francs in rent.

[94] Ibid., D Q⁷ 12666.

[95] Ibid., D E¹ , Fonds Lestringuez, contrat de mariage, Héloise Arson-Eugène Les-tringuez.

[96] Ibid., D Q⁷ , no. 12341 (fols. 56-57).

[97] "Notice nécrologique sur Edouard Servier," Bulletin de l'Association amicale des anciens élèves de l'Ecole Centrale 16 (1884-1885): 205-209.

him. Factory superintendents with twenty years of seniority had to watch this newcomer shoot past them. At the same time, successors to the chief executive of the PGC "parachuted" in from the Corps des Ponts et Chaus-sées without prior experience at the company or, for that matter, in the gas industry. In truth, there were no long, hard roads to the top at the PGC; most long, hard roads led to the middle under the best of circumstances.

It may be that engineers willingly traded off some pay and career opportunity for stable, lifelong positions within one corporate bureaucracy. In this way, too, they emulated the state engineers. Managerial careers at the gas company showed few signs of the hurly-burly of an era undergoing structural economic transformation. Seeking out new opportunities, rising and falling with the fate of recently launched enterprise, jumping from one firm to another as an attractive opening appeared—such comportment did not characterize the personnel of the PGC. Most would have been discouraged if their careers had resembled the turbulent one of a colleague, Paul Desmazes. He graduated from the Ecole Centrale in 1866 and took a position with the Charente Railroad. When ownership of the line changed, he left and became director of a glass factory. It was soon sold, however, and he became an engineer at the larger glassworks of Vier-zon. He entered the PGC in 1881 but stayed for only eight years.^[98] His varied career was exceptional, almost unique. Ordinarily engineers entered the PGC at an early age, and this was often their initial job in the profession. Of the fifty-one engineers whose careers we can follow, twenty-nine (57 percent) came to the company directly after graduation. The minority that did not do so spent only a short time elsewhere; the average delay between finishing school and employment with the PGC was 3.6 years. Hiring experienced managers who had proved their worth in another job was obviously not a practice in which the PGC engaged. It wanted executives who would mature entirely within the firm. Only two of the fifty-one engineers joined the PGC after ten years of work elsewhere. The PGC meant employment to be lifelong and created a training program for its recruits soon after beginning operations.^[99] The company wished the new personnel to learn the ways of the firm and advance within it.

In addition to serving as the first employer, the PGC was almost as often the final employer as well. Less than a quarter of the engineers in the sample resigned their positions. Even when they did so, on average only 5.4 years after entering the PGC, the ideal of lifelong employment

[98] Bulletin de l'Association amicale des anciens élèves de l'Ecole Centrale 32 (1900-1901): 80-81.

[99] AP, V 8 O , no. 666, deliberations of June 11, 1859.

was being honored in the breach. The minority who departed had apparently concluded early that they could not commit themselves to the PGC, so they left in time to establish a career in another enterprise.^[100]

Both the managers and the company appeared to agree on the value of career-long employment. Though the firm did not hesitate to pile responsibilities on its executives, it never fired or demoted except in rare cases of dishonesty. The directors apparently regarded an appointment that did not work out well as their fault and silently endured the consequences. The strongest sanction used to discipline managers was a discreet withholding of a bonus or a raise. Incompetence was tolerated at high levels, as demonstrated by the career of Alexandre Arson. This head of the factory division had been one of the brightest and most productive graduates of the Ecole Centrale. He had served the PGC well for many years; but as he entered his sixties, he lost touch with the latest advances in industry, and his mental powers may have diminished as well. His reports betrayed a distinct absence of clarity and loss of logical rigor. Arson's stubborn commitment to outmoded technology hurt the firm. Yet he remained in his post until his retirement in 1893. The company reduced neither his generous salary nor his large bonuses from a profit-sharing plan. When he retired after thirty-eight years, the PGC feted him and gave him a supplementary pension.^[101] The response to his diminished usefulness was to shift subtly decision making to his assistant engineer. How different this practice was from that of the American steel industry of the nineteenth century, which, according to David Brody, cashiered managers when they were no longer effective.^[102]

The assumption of career-long employment played a role in setting managerial salaries. Seniority was an important determinant of pay. It was not unusual for two managers with analogous posts to earn salaries that differed by as much as 30 percent if one had been with the firm much longer than the other. Engineers found their salaries rising only in small, periodical increments. This was a bureaucratic conception of pay, and it placed the focus on the long run. The engineers seemed quite comfortable

[100] Among the minority of engineers who resigned their posts were those who had never intended to stay with the PGC. Their families had important interests in provincial or foreign gas firms, and they had worked for the PGC only to gain experience. Thus Dominique Favette became head of the Ottoman Gas Company, Edouard Melon director of the Lille Gas Company, and Charles Foucart owner of the Poissy Gas Company.

[101] AP, V 8 O , no. 690, deliberations of December 27, 1893. For an example of Arson's declining grasp of his work, see no. 168, report of January 16, 1884. 102. David Brody, Steelworkers in America: The Nonunion Era (New York, 1960), p. 25.

with the underlying assumptions; indeed, they tried to force career-based pay on the subordinate personnel whenever possible. One immediate consequence of such an approach to compensations was to keep salaries low and slow to rise. This situation prevailed in the state bureaucracy as well.

A certain degree of paternalism accompanied the managers' lifelong careers. If one died on the job, the PGC assumed responsibility for the education of his children. It welcomed engineers' sons into the firm and, when the children were competent, delighted in having them succeed their fathers.^[103] Corporate celebrations centered on long service to the firm— perhaps above all other achievements. When Superintendent Cury retired after fifty-eight years with the PGC and with one of the pre-merger gas companies, an elaborate banquet was held in his honor, and he received his full salary as a pension.^[104] Thus the corporate culture of paternalism encompassed executives as well as the subordinate personnel, but with this difference: the company would always regard its workers as children needing stern guidance, whereas it insisted on initiative from its managers.

Secure careers and paternalism compensated somewhat for the modest salaries and limited horizons. So too did the responsibilities managers received, often at an early age, and the opportunities to exercise their talents. Managers could have no trouble seeing parallels with their professional models, the members of the grands corps d'état, and the similarities must have helped them accept their situation. Nonetheless, it is easy to imagine that some engineers had a sense of disappointment, even distress, in the face of their narrow material rewards. If so, the usual response was quiet acceptance or, occasionally, an effort to improve the situation by finding a new post. Rarely did managers draw enough attention to their discontents to make the corporation reconsider its policies. Yet one such instance occurred during the late 1870s, a booming period for the Parisian economy, when general pay levels rose substantially. The dynamic economic climate allowed and encouraged managers to seek alternatives to their modest situation within the PGC, and it was hit by a small but unprecedented flurry of resignations. Four lower-level managers left in 1879. Such circumstances brought the PGC to take its first—and only— general step to improve the pay for young engineers. That action, no

[103] AP, V 8 O , no. 688, deliberations of June 3, 1891. When the gas main engineer suffered an untimely death, the PGC undertook to pay for the upbringing of his two children. Though the company was always prepared to give employment to the sons of one of the managers, it never guaranteed that the child would succeed his father in an important post. The PGC started sons off as assistant plant supervisors, and they would have to work their way up on their own. Most failed to rio so.

[104] Ibid., no. 693, deliberations of January 22, 1898.

doubt in proportion to the threat, was a measured one: the firm simply began to start recruits at tweny-four thousand francs rather than eighteen hundred francs (the pay it continued to give entering clerks).^[105]

Another decision made by the director a few months later hinted at a more organized protest on the part of factory superintendents. In August 1880 the director suddenly created a series of pay steps (or "classes") for them and promised a rapid promotion of worthy individuals from one step to next. At the same time, Arson proposed giving them incentive pay.^[106] The possibility that this flurry of policy reformulation was a response to collective protest by superintendents about their level of compensation must be taken seriously—though there is no direct evidence to support this speculation—for the measures mirrored precisely the ways the company answered collective demands by workers, clerks, and foremen for higher pay.^[107] Moreover, the moment would have been ripe for the superintendents to take action. It was a time of soaring profits, and the company was adding workers and productive capacity as never before. Factory managers might have cited in support of their demands their growing responsibilities, inflation, and the need to maintain earnings differentials over workers, who had just received a substantial raise. In any case, nothing came of the protest, if indeed it ever materialized. The decision to initiate pay steps soon became a dead letter. Arson's proposal was ignored. Slow, discretionary advances resumed. After 1880 managers found that the time had passed to make demands on their firm. Its earnings record, brilliant up to then, soon became much less so. Plans for growth were scaled back, and gas consumption stagnated until after the turn of the century. Managers now had to contend with growing insubordination from workers and clerks. By no means were the engineers inclined to make their own discontents part of the wider challenge to authority within the firm.

The Social Role of Engineers

The superintendent of the Passy factory in 1868, Théophile Léopold, complained about his company-supplied housing. Léopold lamented that he

[105] Characteristically, the PGC did not commit itself formally to the new policy. It simply put it into practice with an obvious preference for the old way.

[106] AP, V 8 O , no. 680, deliberations of August 6, 1880; no. 716, report no. 65, "Salaire des chauffeurs." Arson's proposal called for salaries determined by a complex formula based on the yield of gas per ton of coal, the amount of coal distilled, the load per retort, and the seniority of each superintendent.

[107] The foremen launched one of their two collective protests over pay at this moment. See Préfecture, B/a 176, reports of May 18, 19, 1880. See chapters 4 and 8 for responses to workers' demands for higher pay.

was forced to live in close proximity to his asistant superintendent. He resented "this cohabitation, which is so intimate for two colleagues," and affirmed that his discomfort would be "infinitely less" if he "had for a neighbor a subaltern employee, with whom no occasion for intimacy could possibly present itself."^[108] The sense of caste this manager obliquely expressed gives credence to the psychological portrait that Charles Kindle-berger has drawn of French managers, stressing elitism, arrogance, and the incapacity for human relations.^[109] To be sure, the engineers of the PGC did uphold the formality of a bourgeois social code and were well aware of constituting an elite. Nonetheless, their acceptance of modest horizons and the loyalty they accorded an organization that demanded so much and rewarded so sparingly reveals other, more deeply rooted characteristics. Duty and hard work were the most prominent features the engineers used to represent themselves—and perhaps they did so with good reason. A useful point of departure for exploring their personalities and attitudes is the successful completion of training that exalted effort, discipline, and conformity to hierarchical rules. Whereas many bright students could not endure the strains of the scientific institutes, and many more dropped out than received degrees, the managers of the PGC had thrived in this demanding environment. All but three of forty-eight graduates of the Ecole Centrale and the Ecole des Mines had ranked in the top half of their classes; 47 percent were in the top quarter. As students, they had nearly always received favorable reports on conduct and attitude; making noise in study hall was about the worst offense any of them had committed.^[110]

Ultimately, the sources are almost silent on the engineers' lives outside their managerial roles.^[111] Yet the eulogy describing one former superintendent of the machinery department as "cold, feeling all the weight of the responsibilities he bore," has the ring of truth. Similarly, factory man-

[108] AP, V 8 O , no. 768, "Usine Ternes," report of June 14, 1868.

[109] Charles Kindleberger, "Technical Education and the French Entrepreneur," in Carter et al., Enterprise and Entrepreneurs , p. 25.

[110] Archives de l'Ecole Centrale, registres de promotion; Archives de l'Ecole des Mines de Paris, dossiers des élèves. I was not permitted to examine the dossiers at the Ecole des Mines directly, but Madame Maisonneuve was kind enough to gather the information I solicited. I am most grateful for her assistance.

[111] One reason for the silence, though undoubtedly a minor one, was the discretion of the managers about their colleagues' private lives. See, for example, AP, V 8 O , no. 162, Lalubié to director, January 21, 1899. Former superintendent Au-brun had an affair with the daughter of a minor employee at his plant. The current superintendent, Bodin, knew about the affair, but the director had to learn about it from the employee. Evidently, the engineers' identification with the firm did not preclude acknowledging that privacy was an absolute right.

ager Biju-Duval left a reputation for having a serious and reserved demeanor: family and music were reported to be his only refuges from work. Charles Monard of the coal department was said to be a "stranger to worldly distractions," without interests aside from his work.^[112] Circumstances gave such testimony a note of verisimilitude. The engineer-managers, from wealthy families, had been raised in Paris or sent there at a young age to pursue serious study. They had resisted all the temptations of that modern Babylon and had fulfilled high parental ambitions by passing through a dearly charted course based on rigorous selection. Deferred gratification and self-control allowed them to attain the qualifications necessary for employment at the PGC. Their acceptance of the limited prospects within the corporation was an exercise of habits they had developed early in life.

Whereas several of the PGC's founders or board members had pronounced political views, managers seemed to eschew political engagement.^[113] Whether their silence was the result of prudence or lack of commitment is impossible to say. Nonetheless, the managers could be mobilized to defend the industrial order they were perpetuating (and helping to transform) against challenges from the left. Their defense of the status quo derived as much or more from moral sensibilities as from class interest. The engineers, self-consciously devoted to the work ethic, thrift, deferred gratification, and hierarchical authority, imagined that workers shared none of these values. They took it as a given that among their duties was implanting their own values in their subordinates. This task was far from being a preoccupation, but it did surface when workers seemed in danger of imbibing more threatening views. So thoroughly pessimistic was the engineers' assessment of workers that it would not be hyperbolic to see the managers' perception of industrial relations as a clash of cultures, a struggle for the triumph of civilizing forces.

These moral and class concerns made managers into inveterate "social engineers." They sought opportunities to teach their inferiors thrift and

[112] Bulletin de l'Association amicale des anciens élèves de l'Ecole Centrale 14 (1892-1893): 191; 35 (1903-1904): 230-233; 27 (1895-1896): 146.

[113] One exception to this rule—and an extreme one—concerns the case of Robert Louzon. A graduate of the Ecole des Mines, he came to the PGC around the turn of the century. He sympathized with revolutionary syndicalists and even loaned Victor Griffuelhes money to pay for a headquarters for the Confédération générale du travail. It was this act that apparently caused his firing; the reformist head of the gas workers' union, Louis Lajarrigue, informed on Louzon to management. See Jean Maitron, Dictionnaire biographique du mouvement ouvrier français , 16 vols. (Paris, 1972), 13: 316. I am grateful to Professor Bruce Vandervort for telling me about this exceptional case.

good work habits without, in the end, having much hope they could learn the lessons. A strike or a demand for a raise set managers to planning various forms of postponed compensation. Servier argued that any wage increase granted during the 1865 labor shortage should go into a fund on which workers could draw once they were beyond the age of fifty and after having served twenty-five years with the company. Euchène, the chief of production thirty years after Servier, perpetuated the moralizing scheme. He advocated placing a raise, if there had to be one, in a pension fund.^[114] The effort to impose on the labor force the career-based pay that was a part of the managers' own work culture was a major thrust of their social engineering, but not the only one. The shock of the Paris Commune impelled Arson to consider ways of preserving the social order. He arrived at the notion of creating an apprenticeship school for worthy sons of worthy workers. Having the fathers contribute two hundred francs to the tuition may not have been realistic, but Arson contended that participation would teach thrift and remind the laborer of his paternal duties. The reform was aimed at strengthening traditional sources of social cohesion, craft, and family, which Arson feared industrial development—of the very kind he directed—was undermining.^[115] One engineer reasoned that even the creation of a factory commissary presented opportunities for moralization. Run by the Sisters of Charity, it would inject chaste and uplifting qualities into the plant.^[116]

Contemporary social observers as well as historians frequently cast engineers in the role of mediators between industrialists and labor.^[117] Their expertise, salaried status, and familiarity with the realities of production supposedly positioned them strategically to make industry more humane. The case of the PGC does nothing to sustain the claim. Its managers were unable to transcend the perceived conflict of cultures between themselves and the workers. Their paternalistic gestures, modest in any case, were enacted in response to immediate labor unrest, and the urgency to reform disappeared the moment the danger passed. Though class barriers may explain some of these limits, the legacy of the engineers' own experiences quite possibly contributed to their rigidity. Schooled (literally) in deferred gratification, they learned that life was a struggle from which even their privileged births did not preserve them. They were intent on living by the

[114] AP, V 8 O no. 148, report of June 20, 1865; no. 1520, "Mission en Angle-terre du mois de mai 1893."

[115] Ibid., no. 712, "Création d'institution humanitaire à l'usine Clichy."

[116] Ibid., no. 717, "Avant projet: Usine du Landit."

[117] Emile Cheysson, Le Rôle social d'ingénieur (Paris, 1897); Shinn, Ecole Polytechnique , p. 210; Beaud, "Ingénieurs du Creusot," p. 58.

principle that meeting obligations was its own reward. Resigned to mediocre compensation, working most of their waking hours, and living fru-gally, engineers could have had little sympathy for demands from below for shorter hours and higher pay. The workers' refusal to see things their way only confirmed their worst fears and reinforced their prejudices. If such stern precepts were in fact the outcome of the managers' training and career experiences, then the engineers' culture was one more reason for troubled industrial relations in France.

<><><><><><><><><><><><>

A manager's angry outburst, one of the very few to have found its way into the corporate record, helps us understand what mattered deeply to this group. Paul Gigot, the chief of the factory division, penned a bitter letter to director Bertrand on the eve of retirement, asking to be relieved of his duties at once. Gigot complained of being "discouraged because the board [of directors] has not accorded me the same marks of esteem that other faithful servants have received." He noted that the PGC's president had sent Superintendent Hadamar a cordial letter at his retirement; the company had given Cury a plaque with a "very flattering" inscription; and by-products chief Audouin left the firm with a decoration "bearing witness to the esteem he enjoyed here." Gigot resentfully noted that his retirement notice, "which will remain in our archives," was a dry, formal document that did not express praise for his service.^[118] Pride in a hard job done well and loyalty to a bureaucracy were important to these managers.

As the new elite of corporate managers carved out their social roles, the bureaucratic and hierarchical practices of the Napoleonic state weighed heavily on them. By contrast, the flourishing of the "new capitalism" in the mid-nineteenth century failed to create a freewheeling situation that rendered old rules obsolete. Conformist graduates of the grandes écoles eventually replaced the iconoclastic entrepreneurs who had founded the gas industry. The celebrated materialism and new money of the Second Empire did not produce a breakdown of standards, a loss of familiar yardsticks of achievement and expected compensation. The hierarchies established by the state engineers continued to set standards and reinforce traditional measures of achievement. The principal way the administration of the PGC diverged from public administration was in rewarding a small portion of executives with generous salaries and profit-sharing bonuses.

The managers also emulated state engineers—and decisively separated themselves from the other social groups created by industrialization—in accepting the right of their superiors to select the individuals who would

[118] AP, V 8 O , no. 162, Gigot to director, November 16, 1905.

enjoy exceptional rewards. As we shall see in subsequent chapters, workers and white-collar employees chafed under hierarchical authority and sought to curtail it; the managers, however, accorded a legitimacy to decisions about their well-being made on high, in secret, without their input, and without reference to generalized rules. Whatever the engineers' private thoughts about the justness of the decisions touching them, they asserted and vigorously defended the right to control the personnel under their direction. At least the managers were not so hypocritical as to deny their superiors discretionary power while insisting on it for themselves. Such equity did not characterize all their stances.

Four
Operating the Company

The PGC represented a relatively recent departure in European economic life. Its size and complexity called for sophisticated cost controls, greater attention to marketing, the allocation of resources among distinct operating units, and strategies for dealing with a large, nontraditional labor force. This "modern enterprise" was run by salaried decision makers who proudly wore the new occupational title of industrial engineer, though they accepted for themselves many of the norms associated with the careers of high civil servants.^[1] It remains to be seen whether the new type of manager imparted a distinctive direction to decisions that had to be made regarding business practices and personnel.^[2]

Business Decisions

With Parisian consumers paying thirty centimes per cubic meter, good management was almost irrelevant to the profitability of the PGC. Slov-

[1] For definitions of the "modern enterprise," see Roger Price, An Economic History of Modern France , 1730-1914 (London, 1975), pp. 118, 147; Alfred Chandler, "The Railroads: Pioneers in Modern Corporate Management," Business History Review 39 (1965): 16-40; Alfred Chandler, The Visible Hand: The Managerial Revolution in American Business (Cambridge, Mass., 1977), pp. 1-3. For a general view of French business history, see Maurice Lévy-Leboyer, "The Large Corporation in Modern France," in Managerial Hierarchies: Comparative Perspectives on the Rise of the Modern Industrial Enterprise, ed. Alfred Chandler and Herman Daems (Cambridge, Mass., 1980), pp. 117-160.

[2] Among the relatively small number of studies dealing with large French enterprises are Jean Bouvier, Le Crédit lyonnais de1863 à 1882 , 2 vols. (Paris, 1961); François Caron, Histoire de l'exploitation d'un grand réseau: La Compagnie du chemin de fer du Nord 1846-1937 (Paris, 1973); Reed Geiger, The Anzin Coal Company , 1800-1833 (Newark, Del., 1974); Pierre Guillaume, La Compagnie des mines de la Loire (1846-1854) (Paris, 1966); Robert Locke, Les Fonderies et forges d'Alais: La Creation d'une entreprise moderne (Paris, 1978).

enly administration, irresponsibly generous settlements with labor, and technological backwardness would not have prevented the company from returning respectable yields. Nonetheless, the earnest and hardworking engineers who ran the firm expected better of themselves, and they gave a great deal more effort to their jobs than mere competence would have required. Balance sheets, capable of exposing waste and inefficiency when examined comparatively, show that by international standards the PGC was a well-run firm. The sense of inferiority that some French industrialists felt toward their British and German (and sometimes toward their Belgian) counterparts was unwarranted in the case of the gas industry (table 9). The PGC's above-average gross production costs were largely the result of the high price of coal—inevitable given the lack of large deposits of the right grades in France. Its net production costs, however, were enviably modest owing to substantial revenues from by-products.

The comparison invites a reevaluation of French entrepreneurship, examined from the distinctive perspective of this case study. Historians have developed some strong stereotypes about French business practices— heavily committed to the production of high-quality goods, disinclined toward mass marketing, oblivious to commercial considerations, valuing stability over growth. The term "Malthusian" inevitably arises, by which

is meant a determination to avoid opportunities for expansion. Proponents of these views often root such behavior in the cultural predispositions of the French bourgeoisie. One grand interpretation of the Third Republic posits an implicit contract between economic elites, on the one hand, and peasants, on the other. In return for conservative economic leadership, the rural masses safeguarded society from the demands of the proletariat. In contrast to such assertions, some leading economic historians have argued that there was nothing pathological about French business practices, that on the whole, entrepreneurs responded rationally to the market situations they faced.^[3]

Whereas the debate is usually framed in terms of the owner-entrepreneur, the case of the PGC reflects on decision making in a different and increasingly important milieu, that of salaried managers. The officers of the gas firm were representative of men who were soon to fill a large number of industrial posts and hold disproportionate economic power. They gained influence at a time when mass marketing emerged as a possibility and sometimes a necessity. There is reason to take special interest in the entrepreneurial spirit guiding the PGC.

It is well known that respect for entrepreneurship and for the morality of the marketplace was more deeply embedded in the social fabric of Anglo-Saxon countries than of France.^[4] Nonetheless, the managers of the PGC came to their jobs armed with a set of values that helped them to be mercenary and calculating about business matters. Their earnestness about making the PGC a well-run firm displayed itself most clearly in the drive to cut expenditures. Unnecessary spending was a deeply felt affront

[3] The vast literature on the subject includes David Landes, "French Entrepreneur-ship and Industrial Growth in the Nineteenth Century," Journal of Economic History 9 (1949): 45-61; François Crouzet, "French Economic Growth in the Nineteenth Century Reconsidered," History 59 (1974): 167-179; Claude Fohlen, "Entrepreneurship and Management in France in the Nineteenth Century," in Cambridge Economic History of Europe (Cambridge, 1978), 7(1): 347-381; Maurice Lévy-Leboyer, "Le Patronat français a-t-il été malthusien?" Le Mouvement social, no. 88 (1974): 3-49; James Laux, "Some Notes on Entrepreneurship in the Early French Automobile Industry," French Historical Studies 3 (1963): 129-134; Robert Aldrich, "Late Comer or Early Starter: New Views on French Economic History" Journal of European Economic History 16 (1987): 89-100. For an investigation of the industrial-agricultural ties, see Herman Lebovics, The Alliance of Iron and Wheat in the Third French Republic, 1860-1914: Origins of the New Conservatism (Baton Rouge, La., 1988).

[4] For an interesting study of American businessmen's attitudes, difficult as they are to pin down, and a source of comparisons with French business leaders, see Thomas Cochran, Railroad Leaders , 1845-1890: The Business Mind in Action (Cambridge, Mass., 1953).

to their professionalism as engineers and to their class values. They did not relax their concern about waste simply because profit levels were so high as to excite the jealousy of the public. The minor annual expenditure of 5,026 francs for omnibus rides by agents whose jobs took them across the city created consternation on the executive committee and stern demands for strict accounting. Managers, from the director to factory superintendents, treated water as an expensive commodity that had to be husbanded. Worn-out equipment and plants with leaky roofs signified the inclination to postpone capital improvements until absolutely necessary. For the same reason, the chief clerk of the customer service department had to receive clients in a room that he claimed lacked dignity.^[5] Sometimes the drive for thrift could seem counterproductive. The chief of the factory division warned that a large corporation could not be run "with the mentality of a greengrocer" and demanded larger maintenance budgets. Yet he was equally an advocate of cutting corners whenever possible.^[6] Managers accepted frugality as a virtue for its own sake and as a duty to the stockholders.

Though top managers were trained in state institutes and belonged to a corps that inculcated dedication to public service, they had no difficulty transferring their loyalty from the public interest to the private firm when they accepted their posts. They displayed no impulse to apologize for the exploitative features of the corporate charter; that the concession was legal and profitable seemed to answer all objections. They even had the ability to put aside ethical niceties when the firm stood to gain. A conspicuous example of the managers' amoral approach to business dealings came during the Franco-Prussian War. The executives of the PGC did make small patriotic gestures, undoubtedly in complete sincerity. But when the government, in desperate need of horses for the army, sought to purchase them from the firm's team, managers consciously selected the oldest and weakest animals and sold them at a high price. An internal memorandum on the matter, far from expressing shame, boastfully concluded that the sale was "not at all a bad piece of business."^[7] If policymakers at the PGC

[5] The executive committee asked factory superintendents to file monthly reports on their efforts to economize: AP, V 8 O , no. 665, deliberations of May 23, 1857. On the economy-consciousness of managers, see no. 751, "Rapport annuel de 1869 (2 janvier 1870)"; no. 768, "Usines de Batignolles et de Saint-Denis," report of April 14, 1861; no. 668, deliberations of April 27, 1861; no. 671, deliberations of June 6, 1866. The phrase l'esprit de l'économie resounded continually and with reverence throughout the superintendents' reports.

[6] Ibid., no. 1062, report of March 10, 1859; no. 715, "Voyage en Angleterre, année 1875."

[7] Ibid., no. 753, "Inspection de la cavalerie," report of December 23, 1870.

did not behave as model entrepreneurs, it was not because they had scruples that led them to eschew maximizing profits.

The energetic pursuit of revenues and thrift brought managers to rationalize most aspects of gas production. Those steps were partly responsible for the good showing of the PGC in comparison with Europe's other major gas firms. The production methods inherited from the pre-merger operations had not been up to the highest standards of efficiency at the time. Alexandre Arson, the head of the factory division in 1856, sought to change them radically and immediately. He replaced the iron retorts used by the early companies with earthenware ones because they were less expensive and conducted heat better. In the plant on the rue de la Tour, which the PGC inherited and continued to operate for a few years, each furnace heated only one retort, a situation requiring lavish use of fuel and manpower. Once in Arson's hands, that plant quickly received new furnaces that heated five to seven retorts. Purchasing coal was the major production cost, and the pre-merger companies did not even have scales to measure how much of it they used.^[8] Efficient use of coal became a preoccupation for Arson and his staff. Their initiatives in this area led to a significant innovation in the gas industry, the use of Siemens furnaces, which recirculated heated air. The PGC was the first to apply the more efficient furnace to gas production in France and realized savings of 36 percent in fuel.^[9] The basic change did not fully satisfy Arson, however, and he continued his search for still less costly methods of roasting coal. After spending thirty-six thousand francs on experiments, he had not yet found the best possible equipment, but the director overruled further expenditures.^[10]

The engineers managed to reduce the production cost of gas by about a third between 1857 and 1868 (figure 9). The essential breakthrough permitting the reduction entailed a new approach to distillation.^[11] Previously,

[8] Ibid., no. 665, deliberations of May 13, 1856; no. 615, report of Arson to director, February 15, 1884.

[9] Ibid., no. 715, "Etude sur les résultats comparatifs de la Compagnie de 1856 à 1870 (14 décembre 1871)"; no. 712, "Fours Siemens et fours à générateur (7 avril 1891)"; no. 671, deliberations of June 11, 1864.

[10] Ibid., no. 676, deliberations of May 24, 1873.

[11] On the new distillation process, see ibid., no. 715, "Etude sur les résultats"; Procès-verbaux et rapport de la commission nominee [e 23 janvier 1880 en exécu-tion de l'article 48 du traité intervenu le 7 janvier 1870 entre la ville de Paris et la Compagnie parisienne de l'éclairage et du chauffage par le gaz (Paris, 1880), pp. 30-38. On some of the results of the new process, see AP, V 8 O , no. 714, report of November 7, 1874; report of June 24, 1875; no. 716, "Salaires des chauffeurs (16 mars 1880)"; no. 711, "Etude sur les réductions qu'il est possible d'introduire dans les prix de revient."

Fig. 9. Cost of Producing One Hundred Cubic Meters
of Gas (in francs). From AP, V 8 O¹ , nos. 718, 1016.

gas works roasted coal at relatively low temperatures, both to save fuel and to keep the lighting power of the gas high. The production staff of the PGC learned (whether through experimentation or imitation of industry leaders is not clear) that dramatic improvements in efficiency were possible by distilling at very high temperatures (1,150-1,200 degrees Centigrade).^[12] Without compromising the lighting power, hot roasting raised the yield of gas from almost all grades of coal. Moreover, the larger the retort, and the more tightly packed it was, the better the results. And with larger retorts the company could save on labor and fuel. In 1855 the average retort received 60 kilograms of coal per charge; in 1869 charges had grown to 127 kilograms.^[13] Arson once announced his guiding principle to be "pitilessly exploiting labor, fuel, and maintenance expenses to the complete utilization of capital."^[14] He made that principle a reality during the 1860s.

Other departments could not easily replicate such success in reducing costs, but most realized significant savings. The cost of producing retorts dropped from about one hundred francs to thirty-eight francs even as they became bigger and lasted longer. The distribution division found new technologies and materials that cut the cost of laying and installing gas

[12] The superintendent of the Passy plant was given a bonus of ten thousand francs for raising productivity in 1863. He may have been responsible for the use of the new method of distillation at the PGC. AP, V 8 O , no. 725, deliberations of March 12, 1863.

[13] Ibid., no. 711, "Rapport no. 32 (8 mars 1868)."

[14] Ibid., no. 716, "Salaire des chauffeurs (16 mars 1880)."

mains. Whereas being near consumers was once the central consideration in locating plants, improvements in distribution technology gave the company more freedom to use inexpensive tracts of land and allowed it to deal effectively with the greatest logistical problem, handling coal. The PGC was able to build its newer plants (eventually numbering nine) on railroad lines or near canals.^[15] One of the largest expenses in increasing output was building facilities for storing gas. Production managers searched all of Europe for the best sorts of storage tanks and finally found the solution in Britain's expanding tanks, which were relatively inexpensive to build and required less land.^[16] Isaac Pereire had once criticized the business sense of engineers, bemoaning their indifference to costs in the pursuit of technical perfection, but his own PGC was hardly a victim of that malady.^[17]

The PGC's excellent comparative performance owed even more to raising revenues than to reducing costs. The director of the Lille Gas Company once remarked that the price of coal and the value of by-products were the keys to profitability in his industry.^[18] That comment goes far to account for the advantages that the PGC held over gas producers in London or Berlin. Aside from the enviable results of receiving thirty centimes for each cubic meter of gas, the PGC obtained great profits from the derivatives of coal distillation. Indeed, management displayed its most creative handling of emerging technologies and its most masterful entrepreneurial decisions in producing and marketing by-products.^[19]

The coal tar (goudrons ) that condensed when gas cooled had been regarded as a nuisance rather than a source of revenue by most gas companies through the 1850s. The pre-merger firms of Paris sold some of the tar to distillers, who refined it for a limited number of established uses, but threw much away. The newly formed PGC continued the practice in its early days. In 1857 it had no by-products department; it subcontracted the treatment of tar to an outside firm and burned some as fuel. Yet this unwelcomed residue was about to become an essential raw material for a

[15] Ibid., no. 715, "Etude sur les résultats," summarizes the improvements. See also no. 669, deliberations of April 19, 1862; no. 673, deliberations of February 3, 1869; no. 1067, "Service des usines: Briqueterie"; no. 25, report of Charles Singly to Paul Lependry, February 6, 1884; report of Servier to director, June 19, 1865.

[16] Ibid., no. 715, "Voyage en Angleterre, année 1875."

[17] Pereire's remarks are cited in John Armstrong, The European Administrative Elite (Princeton, 1973), p. 186.

[18] Compte rendu des travaux de la commission nominee par Monsieur le Ministre de l'Intérieur le 31 janvier 1890 en exécution de l'article 48 . . . (Paris, 1890), p. 31.

[19] For an overview, see the brochure prepared for the Exposition of 1878, "Produits dérivés de la houille," AP, V 8 O , no. 1290.

major new industry, organic dyes.^[20] That industry was born about the same time as the PGC. The young gas firm took advantage of the market opportunities, at first stumbling a bit; but eventually it became an innovator and a model of efficiency.

The artificial-dye industry arose with the development of aniline purple in the mid-1850s. Aniline was a product derived from refining coal tar. Within three years numerous aniline dyes were appearing on the market. The PGC was not prepared to leap into extensive production of the chemical, though; its lack of administrative structure and its subcontracting arrangement stood in the way. Still, top management was aware of the opportunities and soon appointed Paul Audouin to head a new department and undertake the treatment of coal tar internally.^[21] This was not a decision management would regret, for Audouin would make his operation exceedingly profitable through good fortune and through attention to market opportunities.

Lack of technical expertise in industrial chemistry prevented the company from rushing to profit from the enhanced value of aniline. Not until 1864 was the PGC able to market the product, and then only after some trouble. The proud graduates of France's finest technical institutes were completely dependent on temperamental British workers to show them how to manufacture aniline. The workers mounted one delaying tactic after another, and the engineers experienced problems in mastering the secrets of production on their own.^[22] The incident verified the criticisms of engineering education in France, but the PGC adapted much more smoothly to a second phase of the dye industry that started with the perfection of synthetic alizarin at the end of the 1860s.

The new development in dye production required anthracene, another residue of coal distillation, and transformed it into a valuable—often exceedingly valuable—commodity. Audouin was able to lead the PGC smoothly into its production as its price soared.^[23] His accomplishments were impressive enough to justify foreign manufacturers touring the

[20] On the development of organic chemicals, see Paul Hohenberg, Chemicals in Western Europe , 1850-1914: An Economic Study of Technical Change (Chicago, 1967), and L. E Haber, The Chemical Industry during the Nineteenth Century (Oxford, 1958).

[21] AP, V 8 O , no. 723, deliberations of June 13, 1857. Before Audouin was named to head the department, a subcontractor of coal-tar distillation was named "directeur du service de l'exploitation des goudrons." The new department was soon integrated completely into the firm under Audouin's management.

[22] Ibid., no. 670, deliberations of March 30, August 1, and November 26, 1864.

[23] On the development of the anthracene market for the PGC, see Procès-verbaux et rapport de la commission nominée le 23 janvier 1880 , pp. 10-14. The firm's scientific consultant, Henri Regnault, recommended caution in the face of the new industry, but Audouin's energetic policies prevailed. See AP, V 8 O , no. 675, deliberations of January 31, 1872.

PGC's by-product operation. So much had the company improved its handling of organic chemicals over the false steps with aniline that the Gas Light and Coke Company of London used the Parisian plant as a model for its own.^[24] The PGC produced anthracene of high-enough quality to satisfy the exacting Badische Anilin- und Soda-Fabrik, its principal customer, and continued to improve production methods. In 1875 the PGC patented its admired system for conserving heat during the distillation process.^[25]

Revenues from refining coal tar contributed substantially to the gains of the stockholders and peaked just as critics were growing irate over the prosperity of the firm. The income derived from a hundred kilograms of coal tar rose from 2.3 francs in 1865 to 6.9 francs in 1880 and had been as high as eight francs in 1876, when anthracene was still a new product and especially scarce. The PGC found itself in the enviable position of having an ever-growing stock of tar as the result of increasing gas consumption, and the value of the tar appreciated rapidly. Moreover, Audouin had reduced production costs by almost a third.^[26]

The ammonia water produced in purifying coal gas was another commodity that early gas companies had trouble utilizing profitably Some sold it to alum manufacturers, but much found its way into the Seine River.^[27] The newly formed PGC decided to produce ammonium sulfate and wasted little time moving into the market for chemical fertilizers, which was expanding with capitalistic agriculture. It strove to develop a loyal clientele in the heartland of commercial cultivation, the Beauce and Picardy. The sale of ammonium sulfate grew from five hundred thousand kilograms in 1869 to 2.1 million eight years later. The company could well tout its contribution to French agricultural progress. Until then, cultivators had had great difficulty procuring reliable chemicals. Fraud was rampant, and farmers had frequently renounced the use of fertilizers because of deceptions, widely fluctuating prices, or unreliable supplies. The PGC sought to encourage the use of ammonium sulfate by offering a high-quality product, keeping the price stable, and ensuring supplies to regular customers. It largely succeeded in this task, but it was also true that the

[24] Sterling Everard, The History of the Gas Lighting and Coke Company, 1812-1949 (London, 1949), p. 263.

[25] AP, V 8 O no. 677, deliberations of April 7, 1875.

[26] Ibid., no. 1060, "Rapport des sous-produits, 1882"; report of February 20, 1877.

[27] Rapport , March 23, 1871, p. 29. For early subcontracting arrangements, see AP, V 8 O , no. 720, "Marché avec MM. Mallet et Laming."

PGC did not resist the temptation of using its monopolistic power to charge prices not at all justified by production costs.^[28]

Led by the imaginative Audouin, the PGC actively sought outlets for other by-products as well. As advances in printing machines made mass-circulation newspapers more common, the company was able to sell some residues for the manufacture of ink. The new photography industry opened an additional market for ammonium sulfate.^[29] The success of the PGC in exploiting such opportunities was all the more noteworthy in that the firm did not lavish resources on Audouin's department. The byproducts chief was the sole engineer, and he had to handle commercial arrangements, direct production, and oversee personnel matters. The company had one chemist, and poor pay led to frequent turnover in that post.^[30] In the end, the company profited handsomely from a department it treated as marginal.

In many ways the development of coke sales challenged the commercial acumen of managers more than the chemical by-products operation did. The market for coke was small in 1855, and there were competing commodities. Parisian households had traditionally used wood for heating. Coke had had some limited uses, mainly in smelting. Expanding industrial consumption seemed at first the only viable option, for management did not anticipate residential customers. The PGC had begun with the intention of creating a vast new market for its gas as a source of heating in homes and industry. The name of the firm, the Parisian Company of Lighting and Heating by Gas, proclaimed the ambition, but it soon proved beyond its reach: within a year after the company began, it experienced difficulties in applying gas heat to domestic uses.^[31] Although it did not abandon research for proper heating equipment, the company left developments dormant for another two decades. In the meantime, coke was piling up on the distillation floor faster than it could be used for fuel. Getting coke accepted into homes seemed advisable from every point of view. In 1859 the firm created a coke department, which quickly became a large and lucrative unit.

Selling coke for domestic heating engaged managers in some moder-

[28] AP, V 8 O , no. 666, deliberations of June 5, 1858. The PGC began to sell ammonium sulfate in 1858. On commercial policies, see no. 626, "Sulfate d'ammoniaque"; no. 749, "Résultats obtenus du traitement des eaux ammonicales en 1868"; no. 1060, "Résultat du traitement des eaux ammonicales en 1871."

[29] Ibid., no. 669, deliberations of January 14, 1863; no. 671, deliberations of February 25, 1865.

[30] Ibid., no. 156, "Produits chimiques." On the poor pay for chemists, see no. 161, "Charles Steiner."

[31] Ibid., no. 723, deliberations of November 27, 1856.

ately ambitious technological and (especially) commercial activities. When the PGC tried to encourage the use of coke in homes, the practice was so uncommon that effective heaters did not exist. The company's engineers had to work with inventors and develop their own appliances that burned coke cleanly and distributed heat as evenly as possible. Having accomplished that task, it had to persuade residents to use the apparatus. The PGC turned to demonstrations and showrooms to interest clients. It put coke stoves in conspicuous places, such as omnibus stations. It opened a showroom on a major thoroughfare, the rue du 4-Septembre. On display were a range of coke stoves, simple ones for modest clients and elegant ones that might suit the drawing rooms of well-appointed residences. By 1885 the company had sold nearly fifty-five thousand stoves on its own. The coke market was greatly expanded by the Franco-Prussian War, during which other sources of energy were hardly available at any price.^[32]

Delivering the coke also had to evolve into a large-scale operation. The company at first transported coke only to its industrial customers; but in 1860 it created a domestic delivery service. It eventually employed more than two hundred carters. The wagons and uniformed deliverymen quickly became familiar sights of Parisian street life. The firm began to build its own wagons in 1867. Inevitably the new trade that the PGC had engendered became an essential public activity. During the winter of 1879-1880 especially cold weather and snows created shortages of coke and delivery problems. The result was public panic and rioting outside the PGC's coke yards.^[33]

The company was rewarded handsomely for its effective commercial ventures entailing by-products. As Table 9 makes dear, its superior performance over other European firms resulted from the sale of byproducts. That trade brought in about twice as much revenue as the Brussels or London firms received. The income more than compensated for a somewhat less efficient use of labor, higher administrative costs, and the higher costs of coal. The ability of the PGC's managers to lower certain production expenses and raise revenues offers a most flattering portrait of their entrepreneurial role. However, this picture requires a broader focus. The activities analyzed so far show the engineers taking advantage of opportunities that for the most part came to them without serious risk and

[32] Le Gaz: Journal des producteurs et des consommateurs , no. 15 (October 31, 1861): 184; Rapport , March 26, 1885, p. 27; March 28, 1872, p. 30. The annual reports to the stockholders provided figures on the sale of coke stoves.

[33] AP, V 8 O , no. 667, deliberations of August 8, 1860; no. 672, deliberations of May 25, 1867; no. 768, "Usine de Belleville,' reports of November 6 and December 11, 1870; Rapport , March 25, 1880, p. 33. Customers waited in line all night for coke during the Franco-Prussian War.

without exceptional effort. The successes did not test their will to struggle for customers and to develop markets methodically in a competitive environment. In truth, the gas managers were not entrepreneurial in this way.

The energetic defense of a monopoly over gas sold in Paris at high, fixed prices was not an exception to an otherwise competitive spirit among the PGC's managers. In fact, restricted market conditions were a model and an ideal they strove to emulate for the other products the firm sold. That managers were so often able to realize this goal says much about the compartmentalized and localized state of the French economy of the Second Empire, during which more open, national markets ostensibly emerged.

In 1857 the executive committee approved a report that called for the construction of a plant to purify ammonia water at an expense of 415,000 francs. The projected annual profits were 315,000 francs, a return of more than 75 percent—and from the very first year.^[34] It is not difficult to understand why the company pursued the marketing of by-products aggressively: it was able to anticipate such profits because the residues of purification would not be sold in a competitive market. Under the direction of board member Louis Margueritte, the corporation entered into price-fixing arrangements for ammonia products. It had similar terms for marketing alkalies and pitch.^[35] Engineer Audouin's strategy for selling ammonium sulfate was aimed at preserving control over prices. The PGC was the largest producer of the chemical in France, and Audouin ruled out selling any to fertilizer manufacturers, thereby cutting off competitors' supply of the essential raw material. The company effectively made fertilizer into a "luxury" commodity sold on the basis of quality rather than price. When a firm that made alkalies from cesspool refuse threatened to enter that market, Audouin counseled lowering prices temporarily in order to drive the competitor out of business. In most instances, though, lower prices were not part of his commercial strategy not even temporary. Perfected equipment had brought huge increases in the yield of ammonium sulfate, from 74 liters per metric ton of coal to 106 liters. Yet the price of the chemical rose steadily from twenty-eight francs in 1867 to forty-five francs in 1875 and forty-eight francs in 1877. Thus customers did not benefit at all from the PGC's technical prowess.^[36] Similarly a cartel

[34] AP, V 8 O , no. 665, deliberations of January 24, 1857.

[35] Ibid., no. 672, deliberations of January 16, 1867; no. 1063, "Note sur la Société du charbon de Paris." Pernolet, the owner of the Société du charbon de Paris, referred to the PGC as "master of the market" for pitch and asserted that it derived a profit of 25 percent on its invested capital as such.

[36] Ibid., no. 1060; no. 626, "Sulfate d'ammoniaque."

of coke producers from the department of the Nord opted to invade the PGC's "fief" by selling at lower prices, and the gas company responded with an agreement to restore the prices it traditionally charged.^[37] The managers of the PGC fitted perfectly Adam Smith's description of businessmen: they were ready to pursue the invisible hand of profit—and to collude in restraint of trade whenever possible.

The efforts of the PGC to launch and develop new businesses were distinctly limited. Management eschewed patiently building markets, competing with other firms, and anticipating initial losses. The company deided to enter a venture if and when it could set high prices and extract large profits immediately. Management's aggressiveness in creating the coke trade represented the pursuit not just of profits but especially of monopolistic profits. A commission appointed by the interior minister noted, in pointedly euphemistic language, the "elevated prices at which [the PGC] has had the talent to make its coke accepted by the Parisian consumer"^[38] Engineer François Hallopeau, the head of the coke department, contrasted two sorts of clientele. Manufacturers in the suburbs of Paris found the PGC's coke far too costly; if his salesmen were too aggressive, "they were shown the guard dog." But Hallopeau found urban residents to be pliant customers: 'Accustomed to paying a good deal for wood or oil, they will not hesitate to pay a few francs more for coke." In fact, the company sold coke to the public at four and eventually five times the cost of production and distribution. It chose to maintain these prices rather than reduce the stock when sales were slow.^[39] The company's behavior in the retort trade was little different. For a while, the PGC manufactured retorts and sold them to other gas companies, but at a markup of about 100 percent. When a smaller firm mechanized the operation and undersold the PGC, the latter quickly withdrew from the business.^[40] Thus it was not at all remarkable that a newly appointed coke engineer aspired to create a monopoly over the sale of all combustible energy sources in Paris by using

[37] Ibid., no. 25, reports of Brissac to director, January 7, 1887, August 13, 1888, and February 16, 1897.

[38] Compte rendu des travaux de la commission nominee le 4 février 1885 en exécution de l'article 48 du traité intervenu le 7 février 1870 entre la ville de Paris et la Compagnie parisienne d'éclairage et de chauffage par le gaz (Paris, 1886), p. 47.

[39] AP, V 8 O , no. 752, "Rapport sur la vente de coke (7 juillet 1866)." Lower coal prices were slow to be reflected in coke prices. See no. 752, "Chauffage," and no. 25, report of February 16, 1897.

[40] Ibid., no. 1067, report of Audouin to director, March 9, 1882. The firm be-hayed in a similar manner regarding gas motors and meter rentals. See no. 1257, "Rapport sur la machine à gaz (22 mai 1885)"; no. 24, "Service extßrieur. Renseignements divers," report of November 18, 1891.

the firm's influence over suppliers.^[41] Ultimately, management regarded the exceptional level of returns from gas sales as a situation they might hope to duplicate in all other operations.

The PGC was usually able to sustain high prices and profits because markets were imperfect. In the case of coke, customers had no initial notion of what a fair price would be since the PGC practically originated the trade and accustomed the customers to high prices at once. Competition failed to whittle down prices because most potential competitors conspired with the PGC to keep prices high. The smaller merchants in the coke trade charged as much as the PGC or more. Similarly, the markup on retorts was sustainable as long as competitors took the same profits.^[42] As we have seen, the PGC readily found partners in price-fixing arrangements for byproducts or used its market power to drive holdouts from the scene.

The final result of such entrepreneurship was that the PGC, as a pioneer of large-scale enterprise, did little to strengthen market forces or renew business practices. The company fit into—and even reinforced—the localized, stratified, and anticompetitive features of French economic life. Although the corporation pioneered new markets, its high-price, high-profit policies condemned those markets to limited size. The PGC may have represented an aggressive sort of management, but it was aggressive in limited ways that solidified the weaknesses of national economic life.

The types of market the PGC entered or constructed had clear implications for its technological decisions. The modest size of the markets dampened pressures for innovation, with predictable results.^[43] The company did not take full advantage of its breakthrough in distilling techniques; in 1905, 40 percent of the gas was still produced in ordinary furnaces rather than recuperative ones.^[44] The mechanization of labor-intensive operations was also exceptionally slow. As we shall soon see, the absence of market pressures allowed the PGC to make decisions in this area almost entirely on the basis of the need to control its work force. The ministerial commissions that evaluated the corporation in 1880, 1885, and 1890 were careful not to criticize the firm for technological backwardness but still left the

[41] Ibid., no. 25, report of Hallopeau to director, December 30, 1859.

[42] Ibid., no. 671, deliberations of August 10 and 24, 1864; no. 672, deliberations of January 16, 1867; no. 626, report of Audouin to director, April 22, 1874; no. 1062, "Note sur la Société du charbon de Paris"; no. 1257, "Rapport sur la machine à gaz."

[43] Conversely, management was most effective in promoting efficiency when it was under pressure for increased output. The rationalization of gas production occurred while immediate demand remained unfulfilled (that is, up to 1866). See ibid., no. 726, deliberations of June 14, 1866.

[44] Ibid., no. 620, "Etat dormant le nombre et le type des fours de distillation . . . (19 juin 1905)." Of 6,746 retorts, 2,856 were heated by ordinary furnaces.

impression that it was exceedingly cautious and complacent. As the corporate charter neared its term, the goal of avoiding capital expenditures reinforced the caution. No wonder the new Société du gaz de Paris, which succeeded the PGC in 1907, found the assets it had inherited decidedly old-fashioned and immediately undertook an ambitious modernization program.^[45]

<><><><><><><><><><><><>

The practices that management developed for its subsidiary products applied equally to the main business, selling gas. The director repeatedly asserted his desire to expand the clientele as much as possible, to make gas a part of daily life, "like air and water." Moreover, the corporate charter imposed on the PGC the obligation to do what was necessary to facilitate the use of gas.^[46] But the director did not really mean what he said; or if he did, the policies he pursued were inappropriate. The market for gas, it will be recalled, was limited in important ways.^[47] The PGC had virtually saturated the commercial sector, but the domestic market was largely untapped. By 1885, at the end of the golden era only one in six apartment houses had mounted mains; only 15 percent of apartments were adjacent to the mains; only 5 percent of households actually used gas. Such results raise questions about how the company went about marketing its principal product.

The superficial penetration of the residential clientele was generally the result of the managerial policies pursued during the golden age. The PGC treated gas much as it did by-products. It recognized a large and profitable "natural" clientele, created for the most part by beneficent forces beyond its control. These customers—commercial enterprises, in the case of gas—would pay dearly for gas, either because they had to use it or did not care about the cost. As for the potential users that might have been won over with some judicious enticements, the managers were indifferent to them. Not even the marginal profit of twenty centimes on each additional cubic meter sold interested officers in pursuing the customers who needed to be convinced. As a result, the clientele remained relatively small throughout the golden era.

Corporate policies regarding the ancillary costs entailed in using gas illustrated a cynical realism about markets—the view that some customers would employ gas at almost any price and the rest of the population was not relevant. Rather than supplying meters, connector pipes, and

[45] Parisian officials were well aware that they would soon inherit antiquated and in some cases unusable plants. See Frederic Sauton, Etude sur la question du gaz à Paris (Paris, 1902), p. 3.

[46] Rapport, March 7, 1857, p. 4; March 12, 1862, p. 24; March 28, 1874, p. 6.

[47] See chapter 1.

valves at nominal cost to encourage gas use, the company took a heavy profit on each. It earned a 40 percent return on the rental of the pipes connecting residences to the mains. Yields on the sale or rental of meters were between 20 and 30 percent.^[48] So high were such charges that the PGC earned three to four times the income that the London or Berlin gas companies derived from their accessory charges. The Parisian firm even included a payment for the "upkeep of connector pipes" that was fictitious; the fee involved no work and was simply a supplementary cost of using gas.^[49]

In addition to charging customers a steep installation fee, the PGC failed to make gas easily accessible. Mounted mains (first used in 1860) were the central means for bringing gas to residences. The firm elaborated a threefold package of incentives for lighting with gas between 1866 and 1872, but the policies did not receive enough thought or financial commitment for success. Management sought to stimulate owners of apartment buildings to install mounted mains and prepare at least three apartments for gas use (with connector pipes and fixtures) by offering a hundred francs. The payment just about covered the cost of installation, so the real encouragement was presumably the increased value of apartments with gas. Tenants in lodgings accessible to a mounted main were eligible to receive thirty francs for installing the necessary equipment. Again, that sum barely covered the cost of the work. The company also promised forty-franc incentives to gas fitters (appareilleurs ) for finding tenants who would take gas and giving them the means to do so. The payment yielded a profit of about ten francs for the fitters.^[50]

The marketing effort fell far short of achieving the announced goal of diffusing gas lighting broadly. Its provisions were poorly publicized. The company relied on the gas fitters to find interested landlords and tenants but failed to provide enough incentive for them to make a real effort. Managers explicitly recognized that ten francs did not excite much activity but declined to raise the bonus. In one important way the company provided the fitters with a disincentive: it had convinced them to construct mounted mains at rates below cost on the grounds that the mains would soon bring additional business from tenants who would order connector pipes and fixtures. In effect, the company had found an indirect means of

[48] AP, V 8 O , no. 25, report of Mayniel to director, July 31, 1858; no. 1064, "Album de divers tableaux graphiques."

[49] Ibid., no. 709; Pierre Mongéaud, La Question du gaz à Paris et le régime nouveau (Paris, 1908), p. 103.

[50] AP, V 8 O , no. 671, deliberations of April 21 and June 13, 1866; Compagnie parisienne du gaz, Instruction à l'usage du personnel du service extérieur concernant la fourniture du gaz (Paris, 1879), p. 16.

pushing some of the costs of the mains on the customer again. Yet fitters failed to benefit as promised because tenants usually did not respond to the opportunity that the mounted main offered.^[51] Moreover, landlords had to agree to so many provisions to receive the mains that they were confused or suspicious rather than eager.^[52] Ultimately, management refused to regard the mounted-main program as an investment that could readily yield a return of 30 to 40 percent with some nurturing and careful planning. Instead, it treated the program as an expense that had to pay off at once or be cut.

Concerned about the high proportion of mains that were underutilized, managers decided to curtail the incentive program rather than pursue it with renewed vigor. In 1874 the director announced that the PGC would no longer install them simply at the request of landlords and now would favor only the buildings that were likely to be the most lucrative. Soon the company formalized the restrictions by limiting the buildings that would receive mains to those with apartments renting for at least eight hundred francs annually. The PGC thereby excluded all but the wealthiest 5 or 6 percent of Parisian tenants from gas use. In 1880 the company canceled the hundred-franc bonus to landlords; now it would install mains at its own expense, but only after the owner deposited a hundred francs, which would be returned if and when three tenants became customers.^[53] Thus, the mounted-main program—the principal means for reaching residential customers—became a faithful reflection of management's cynical fatalism. The company drew the line at investments returning less than 30 percent, the yield of mains on the most luxurious buildings.^[54]

The policy of withholding gas from all but the wealthiest Parisians is revealing on a number of counts. Corporate engineers calculated in 1876 that mounted mains on apartment buildings where rents were below eight hundred francs brought a 9.6 percent return on the investment.^[55] That yield was of course substantial by most standards, and management's disdain underscored the inflated expectations within the firm. Furthermore, the abandonment of less wealthy tenants (not to mention the common people) announced the PGC's indifference to the marketing of gas. Profits

[51] AP, V 8 O , no. 671, deliberations of April 21, 1866.

[52] Ibid., no. 678, deliberations of February 14, 1877.

[53] Ibid., no. 677, deliberations of April 14, 1875; no. 680, deliberations of December 24, 1880; no. 30, report of Forqueray to director, February 7, 1876; Rapport , March 23, 1875, p. 8.

[54] AP, V 8 O , no. 30, "Statistique des conduites montantes au point de vue du prix des loyers (2 avril 1875)." For evidence that profits might have been even higher, see no. 1064, tableau no. 11.

[55] Ibid., no. 30.

from the mains were "only" 9.6 percent because so few tenants saw fit to install gas fixtures, yet the company took no useful steps to make gas more attractive to the hesitant. The disappointing returns from the mounted mains were the firm's own doing. The free-installation program implemented at the end of the 1880s was an admission that earlier incentive plans had failed.

The same refusal to consider long-term growth and the maximization of profit was evident in the company's policies toward gas motors. The managers began by predicting a brilliant future for gas motors in a city with so many small workshops. Noiseless, clean, safe, and free from the requirements of police inspection, they had advantages over steam engines. The company declared its intention of supplying motors at a nominal charge in order to reap huge profits from enhanced gas consumption. Quickly, and perhaps without consciously confronting the matter, managers displaced the goal of the policy from expanding consumption to gaining immediate profits from the sale of motors. The company placed a markup of at least 20 percent on the motors it sold.^[56] Once again the officers were attempting to exploit the "natural" customers who would use gas motors at any price. The trouble in this case was that the natural clientele remained tiny. The company discontinued the production of motors and bemoaned the poor performance of this operation but took no imaginative steps to build a market, which happened to be very large in capitals like Berlin.

In spite of its commercial policies, the PGC prospered. Managers were able to serve customers that needed their products regardless of cost. Otherwise, their attitude toward markets was arrogant, exploitive, and restrictive in all but rhetoric. The managers could well afford to sell gas as a "luxury" commodity during the golden age, and the brilliant corporate balance sheet raised few questions about the quality of policy-making. The passing of this era and the curtailment of a natural clientele offered a test of the managers' entrepreneurial mettle.

<><><><><><><><><><><><>

Management probably did not foresee the loss of its most lucrative gas customers and the crisis of consumption of the 1890s. If it did, it took no steps to anticipate the necessary adjustments.^[57] Nonetheless, when circumstances forced the executives to act, they analyzed the options coher-

[56] Ibid., no. 1257, "Machines à gaz"; no. 776, "Analyse du prix de vente des moteurs Lenoir' (fols. 88-94); no. 775, "Rapport sur la situation de l'exploitation" (fols. 476-486).

[57] Ibid., no. 1062, report of Lefebvre to director, February 3, 1876. More common than underestimating the potential of electricity was ignoring the competitor.

ently and took logical steps to create a niche for gas in the era of electricity. Management's performance was not especially innovative or imaginative. It merely became open—in a conservative fashion—to reasonable options it had not considered before.

The loss of large business customers to electrical lighting renewed management's interest in gaining residential customers and in promoting gas appliances. The shifts in marketing strategy found symbolic expression in the nature of firm's participation in the universal expositions, held every decade during the second half of the century. Industrial prowess was the theme of the corporate exhibit at the Universal Exposition of 1878. Visitors were shown vials of coal tar, piles of coke, displays of gas valves, and the latest sorts of mains.^[58] Probably few tourists found this showcase to be the highlight of their visit; nor did it convince them that the PGC was any closer to the way they lived their lives than was a steel mill. The motifs of the exhibit showed that the PGC was not fundamentally interested in winning customers.

The harsh decade that followed that exhibit forced the firm to reexamine its position in basic ways. The theme of the corporate pavilion at the Exposition of 1889 symbolized the direction and the extent of the reas-sessment. The pavilion took the form of a home furnished with all the latest gas appliances. Visitors could marvel at gas stoves, water heaters, room furnaces, and toasters. Gas motors in the kitchen scraped vegetables, rinsed bottles, and cleaned utensils. Below the domestic quarters was a restaurant with as much gas equipment as the company could invent. Despite the unkind cut by the Gazette nationale that the pavilion was like a museum of medieval artifacts, visitors could at least see some ways that gas would make their homes more comfortable and agreeable.^[59]

The strategy for marketing gas in an age of electricity was implicit in the exhibit. The company would pursue domestic and commercial cusomers and do so by accentuating gas appliances as much as lighting.^[60] To this end, the director charged the mechanical service department with devising and perfecting domestic appliances. Lefebvre, the engineer who headed the service, identified cooking as the most fruitful area of expansion. Not only did gas have technological advantages over electricity for that purpose, but also the strategy was recommended by deeply rooted mores. As the engineer pointed out, Frenchmen knew that it was "better

[58] Ibid., nos. 1290-1291, "Exposition de 1878."

[59] Ibid., no. 1292, "Exposition de 1889."

[60] The PGC finally created a sophisticated lighting laboratory in 1897. Auguste Lévy, Communication faite au nora de la Compagnie parisienne du gaz par Monsieur Auguste Lévy (Paris, 1900), pp. 3-4.

A corporate handbill illustrating the appliance showrooms. Courtesy Archives de Paris.

to eat by poor candlelight and have great food than to have brilliant light and poor food."^[61] His department also perfected gas heaters and found manufacturers to produce them on a large scale. The company used its showrooms, which had specialized in coke appliances for the previous decades, to publicize and sell the ranges and furnaces. To stimulate interest in cooking with gas, the company engaged a Madame Mees to give public lessons and to write a cookbook. Between February and December 1892 nearly three thousand people attended her talks, held at the various showrooms. The company also arranged with the Dufayel department store, one of the largest in the city, to sell stoves and heaters manufactured under the corporate label in 1897.^[62] The new quest for gas-powered devices sometimes brought engineers to a dead end, as well. They tinkered with gas-driven elevators, tramways, and automobiles.^[63] At least these failed efforts demonstrated a determination to make gas relevant to the machine age. Such a determination had not been evident before competition from electricity intensified. The PGC also tried to recapture some of its large commercial clients, this time for the use of appliances rather than lighting. It willingly gave subsidies to well-known restaurants for the installation of stoves or grills.^[64] Management may have hoped that, beyond selling more gas to cafes, it would also be familiarizing the public with meals prepared by gas.

In order for the promotion of appliances to be effective, the PGC had to find ways of convincing Parisians to install gas lines in their homes. With sixty-seven thousand apartments ready to be served by existing mounted mains, there was a large potential market close at hand but previously neglected. In 1887 the company finally took a meaningful step to build markets through a free-installation program, which lent to tenants already at mains the internal pipes, a stove, and a chandelier for the dining room. At last managers had begun to come to terms with investments designed to reap returns over the long run. Their approach to the new policy was cautious and limited, however. Nothing that the PGC did to attract residential customers was original. Gas companies in provincial cities had long lent appliances to customers; indeed, they often offered a wider range of equipment and imposed fewer restrictions. Moreover, Di-

[61] AP, V 8 O , no. 1292, "Rapports et notes émanant des divers services."

[62] Ibid., no. 1303, "Exposition culinaire"; no. 693, deliberations of November 13, 1897. The PGC maintained display spaces on the boulevard St-Germain, rue Condorcet, rue du 4-Septembre, and rue Lafayette.

[63] Ibid., no. 1303, "Voitures actionnés par moteurs à gaz"; no. 691, deliberations of November 13 and 20, 1895.

[64] The deliberations of the comite d'exécution record these agreements. See Ibid., nos. 691-694.

rector Camus at first insisted on the awkward provision that only new customers could receive free stoves; a faithful customer who had used gas prior to 1887 had to purchase a stove.^[65] In spite of the success of the program in attracting new customers and expected returns of 50 percent on the investment, management was halfhearted in its commitments.

Efforts to seek a genuinely mass clientele required still more daring— almost more than management could muster. Some engineers continued to seek business exclusively among the well-off. Thus, Lefebvre, the strategist for the 1889 exposition, rejected a plan to place in the pavilion a simple restaurant serving biftecks; he wanted an elegant café that would attract the grande bourgeoisie (his term).^[66] Nonetheless, the company eventually resolved to democratize gas use, provided the city would compensate it for sacrificing some of its prerogatives. One indication of management's growing acceptance of democratization was a more favorable attitude toward reduced gas rates. In 1880 the company had dismissed protest over thirty centimes as the work of outside agitators. Twelve years later, the PGC sought to organize consumer groups for the purpose of pressuring the municipal council.^[67] In the Sauton project of 1892 the company agreed to gas at twenty centimes, with full compensation for the reduction, and to the fee-free program, which made gas readily available to renters paying less than five hundred francs a year (about 75 percent of all Parisian tenants). As we have seen, the agreements were not realized at once because the Sauton project fell apart; certain aldermen kept insisting on a fee-free program.^[68] Faced with falling consumption in 1892, 1893, and 1894, the company finally found the courage to accept it. With that consent, gas entered petit bourgeois households.

The new mass-marketing arrangement had excellent economic justifications and a political payoff as well. Mounted mains were much more lucrative than ever before as a result of free installations and free stoves. The financial projections made by the PGC's engineers in 1892 predicted returns in the range of 30 percent on investments in mounted mains, pipes, and stoves—less than half the yield from supplying gas to regular

[65] Ibid., no. 24, "Transformation de l'éclairage dans Paris (14 novembre 1892)"; no. 30, report of November 1, 1887; Rapport , March 28, 1888, p. 6. For the policies of provincial and foreign gas companies, see AP, V 8 O , no. 1002, "Note sur les moyens qu'il conviendrait d'employer à Paris pour augmenter le hombre des abonnés . . . (25 octobre 1890)."

[66] AP, V 8 O , no. 1292, "Rapport et notes émanant des divers services."

[67] Ibid., no. 1257, Jules Aronssohn to director, November 23, 1892. In 1900 the director finally admitted to stockholders that Parisians did genuinely want lower prices. See Rapport, March 28, 1901, p. 56.

[68] AP, V 8 O , no. 615, 'Exonération des frais accessoires."

customers but still handsome.^[69] Moreover, management could anticipate valuable political benefits. Thousands of small customers complaining about expensive gas and eager for immediate relief would provide a splendid instrument for softening opposition to a renewed charter on the municipal council. The director could imagine dictating terms to the city, and the strategy almost succeeded with the Chamon project of 1901-1902 (see chapter 2).

With such advantages, it is worth asking why the PGC left the initiative for the fee-free program to the municipality and then implemented it cautiously. The few problems that managers repeatedly cited were simply not so serious as to raise grave doubts. There was the concern that regular customers might resent the special treatment modest renters were about to receive. The head of the customer accounts department worried that the program would vastly complicate the bookkeeping operations and raise the incidence of delinquent accounts—which it did to a small extent.^[70] Surely these objections masked a fundamental aversion to change and risk. The PGC had every reason to expect comfortable profits from the fee-free initiative, yet the decision makers fretted over the large capital expenditures a mass market would require.^[71] In the face of uncertainty, limited though the imponderables were, the status quo had its attractions. Thus leftist aldermen found themselves urging capitalists to pursue a profitable venture and even offering financial incentives for the sake of democratizing consumerism.

Characteristically, the engineers proceeded slowly, cautiously, and conservatively with mass marketing. Rather than using the program as a vehicle for embracing as many customers as possible, the PGC restricted access. The director insisted on verifying the rent levels of applicants and petitioned the prefect to check the claims against tax records. The com-

[69] Ibid., no. 30, "Conduites montantes"; no. 689, deliberations of August 18, 1892. The returns on mounted mains that the engineers reported for the 1890s were much higher than those cited earlier in the chapter for the 1870s. The free-installation and fee-free programs were responsible for the larger yields. Not only did more tenants per building take gas, but their free stoves and fixtures used more of it. Thus each main generated more income.

[70] Ibid., no. 24, "Renseignements concernant les recettes sur conduites mon-tames"; no. 90, "Tableau du personnel du bureau des recettes." The PGC might have tried to supply the common people with gas while avoiding these problems by offering prepaid meters. (The British companies had been doing so for years.) Yet the PGC never permitted this approach to mass marketing.

[71] The PGC would eventually spend twenty million francs on equipment for the program. See ibid., no. 617, "Charges supportées par la Compagnie du fait des abonnements sans frais." The firm purported to have saved small consumers forty-six million francs between 1894 and 1902. The large sum attests to the burden of accessory charges.

party even went so far as to question the qualifications of an alderman who claimed to be eligible. The PGC would not provide buildings with mounted mains until landlords put up a security deposit—a payment many no doubt refused.^[72] Such hesitancy was not justified by the market testing that the company had done prior to the agreement nor by the final results. Accepting mass marketing had been the correct decision. Fee-free clients consumed 80 percent of the increase in gas sold after 1895 and restored, in the firm's last years, the high profit levels of the golden age.⁷³

The quality of the managers' entrepreneurial decisions lacked a simple, uniform profile, so it is easier to describe what that profile was not than what it was. Clearly, the PGC's salaried executives were not Malthusian in any thoroughgoing sense. Nor were they profit maximizers (though they did not eschew huge profits). The engineers of the gas company manifested three valuable talents: they strove tirelessly to reduce costs and found appropriate techniques to do so; they exploited monopolistic opportunities successfully especially when mass marketing was not involved; and they knew how to construct protected markets. Such talents did not lead to basic innovations in business practice; nor, if generalized, did they promise that the rise of large-scale enterprises in France would make the economy function more efficiently. The PGC's managers wished to remain aloof from market forces and largely succeeded in doing so. Even their drive for efficiency in production owed less to the pressures of the marketplace than to their sense of duty and professionalism as engineers. The strengthening of competitive pressures, such as they were by the 1890s, barely changed the quality of decision making.

The debate about whether French entrepreneurial behavior was shaped by a particular mentalité or was entirely a matter of rational responses to the economic environment is impossible to resolve with finality even for one isolated firm. In any case, the more important point is that the managers' cultural presuppositions weighed heavily in their business judgments. They may have displayed an immediate responsiveness to empirical realities on matters of detail, but the engineers' ideology and outlook guided policies regarding larger matters—at least until reality forced adjustments. Moreover, their professional backgrounds imposed certain predilections and expectations. Neither by training nor by vocation were

[72] Ibid., no. 691, deliberations of January 9 and May 4, 1895; no. 693, deliberations of March 13, 1897; Conseil municipal, deliberations of December 8, 1893 (the remarks of Councilman Brousse, in particular).

[73] AP, V 8 O , no. 617, "Augmentation dans la consommation de l'éclairage particulier."

the managers entrepreneurs. They were engineers, destined by their scholastic achievements to responsible posts. Their intellectual formation, reinforced by class values, taught them what a centime saved meant. Their training did not prepare them to take risks nor to proffer their wares to a hesitant public of consumers. The engineers leaned toward administration; their career expectations required stability and orderly change, if change was inevitable. The good fortune of the PGC's officers was that these requirements were consistent with handsome profits for most of the life of the firm.

Culturally shaped expectations especially influenced corporate policies regarding the labor force. In this area, however, frugality and efforts to stay aloof from market forces created results that proved visibly more problematical. Personnel management had no equivalent to gas at thirty centimes for concealing failures.

Labor Policies

Among the many responsibilities of the PGC's management was the direction of an unusually large labor force, probably the largest in Paris belonging to a single employer. The firm's policies and decisions bear attention on a number of grounds. First, the gas company had an uncommon amount of freedom to create the style of industrial relations it desired. As an early example of large-scale industry, the PGC found that patterns of behavior and expectations were not yet tightly drawn. Moreover, the company had fewer economic constraints than most firms. It operated far above the battlefield of intense competition and therefore had the ability to bestow special advantages on its work force if it chose. Liberality was all the more possible because labor was a relatively small portion of total expenses, on average 10 percent. In mining and railroads, by contrast, the wage bill was more on the order of 30 percent of expenditures.^[74] Such freedom from constraint meant that the labor policy of the PGC reflected to an unusual degree the values and priorities of the managers.^[75]

Industrial relations at the PGC are also of interest because they eventually had to respond to a nationwide shift in political economy. We have

[74] Maurice Lévy-Leboyer, "Capital Investment and Economic Growth in France, 1820-1930," in The Cambridge Economic History of Europe (Cambridge, 1978), 8:261. For labor costs in other industries, see Jean Bouvier et al., Le Mouvement du profit en France au XIX siècle (Paris, 1965), p. 77.

[75] On the study of labor policy from a comparative perspective, see Howard Gospel and Craig Littler, eds., Managerial Strategies and Industrial Relations (London, 1983); Elaine Glovka Spencer, Management and Labor in Imperial Germany (New Brunswick, N.J., 1984).

already seen that the company became entangled—perhaps more than any private firm in France—in a search for alternatives to market orthodoxy on the part of some conservatives and some politicians of the left. The latter sought to extend republican rights to the workplace as a matter of justice and as a strategy for integrating workers into a democratic regime. The mobilization of conservatives against collectivism and "state socialism" took several forms, including a bourgeois reform movement dedicated to reducing industrial tensions through private, paternalistic activities. The labor policies of the PGC during its last fifteen years were implicated in the search for social peace. It was not surprising that all sides regarded the firm as a test case for their respective proposals.^[76]

Management, local politicians, and reformers had to develop their strategies for a large and varied labor force.^[77] The wage-earning personnel at the PGC grew from about twelve hundred when the company was founded to five thousand at the end of the golden age and more than eight thousand when the charter expired. The factory laborers were distributed among seven to nine plants (depending on the moment) scattered about Paris and the suburbs.^[78] In addition to the plant personnel there were lamplighters, valve greasers, plumbers, carpenters, navvies, general laborers, and still other workers. In face of such diversity it makes sense to simplify by recognizing two basic groups. The common laborers, who could learn their jobs in a matter of days or hours, comprised just less than 80 percent of the work force. More problematical for management was the other group, the stokers (chauffeurs ), who loaded retorts with coal for distillation. The stokers' work was not skilled in the classic sense of requiring the level of dexterity and judgment acquired through long training; yet their work did demand rare attributes. Stoking called for formidable strength and endurance; most men were not up to the task of toiling in the intense heat of the furnace rooms. Scarcity gave stokers market power. If the company was not prepared for their threats, these workmen could have the firm at their mercy, at least in the short run. Managers

[76] See chapter 2. One argument industrialists used against free trade was that protection would allow them to pacify workers with higher pay and shorter hours. See Lebovics, Alliance of Iron and Wheat, chap. 3. As we shall see, the case of the PGC raises serious doubts about the employers' sincerity.

[77] Only seventeen French firms employed more than five thousand people in 1906. See Georges Dupeux, French Society , 1789-1970, trans. Peter Wait (London, 1976), p. 207. The PGC, with 8,000 workers, was easily in this category.

[78] The plants the PGC operated during the 1860s were La Villette (nineteenth arrondissement), Ternes (seventeenth), Passy (sixteenth), Vaugirard (fifteenth), Ivry (thirteenth), Saint-Mandé (twelfth), and Belleville (twentieth). In 1905 the PGC's plants were in La Villette, Passy, Vaugirard, Ivry, Saint-Mandé, Boulogne, Alfortville, Clichy, and Landy (Saint-Denis).

preferred to think of all their laborers as "simple workers," but in practice they could not apply policies uniformly nor could they expect uniform responses to their policies.^[79]

The gas executives had only one vocabulary with which to conceptualize industrial relations, that of paternalism.^[80] The notion entailed more than a strategy for imposing their interests on the personnel; it was a cultural system and a set of symbolic relationships in which managers themselves were enmeshed. The engineers had matured learning that their authority over workers was based not only on superior mental acumen but also on their responsibility for their underlings. As students in the technical institutes, they were exposed to industrial enterprises that put the principle into practice. Paul Gigot, for example, toured such plants as a twenty-one-year-old student at the Ecole des Mines. This future engineer, who would become head of the PGC's factory division in 1893, found at a zinc plant in Belgium a savings plan, a medical fund, and company housing. Gigot noted in his student journal that these initiatives were "not at all useless, when one considers the workers' complaints about their painful work, which they cannot do for many years."^[81] Even as a young man, Gigot grasped the inevitable connection between paternalism, moralization of laborers, and corporate interest. There was probably not an engineer in the company who would have dissented from the proposition that employers owed workers more than their wages.

The culture of paternalism brought its own unquestioned assumptions—that poverty was inevitable for workers, that most of them were morally weak, and that intervention in their lives was effective only if performed face to face. These and other presuppositions provided a comfortable check on charitable gestures that were too ambitious. Plant superintendent Jones argued in 1865 that "any action that tended to suggest to workers that they are indispensable only spoils them."^[82] The assistant factory chief seconded this opinion and argued that to grant workers' demands provided merely "a facile solution" that gave only "passing satis-

[79] Chapter 7 treats manual laborers in greater detail.

[80] Two important works on industrial paternalism are David Roberts, Paternalism in Early Victorian England (New Brunswick, N.J., 1979) and Donald Reid, "Industrial Paternalism: Discourse and Practice in Nineteenth-Century French Mining and Metallurgy," Comparative Studies in Society and History 27 (1985): 579-607. See also Eugene McCreary, "Social Welfare and Business: The Krupp Welfare Program, 1860-1914," Business History Review 42 (1968): 24-49; Joseph Melling, "Industrial Strife and Business Welfare: The Case of the South Metro Gas Company," Business History 21 (1979): 163-179.

[81] Archives de l'Ecole des Mines, journaux de voyage, J. 1858 (194).

[82] AP, V 8 O , no. 148, report of Superintendent Jones, April 26, 1865.

factions and soon creates new pretentions."^[83] Nearly thirty years later, the new assistant chief, Albert Euchène, reaffirmed that idea, noting that workers "understand no pleasures without a drink in their hands" and that any improvement in pay benefited chiefly the pub owners. Euchène elaborated on the theme by arguing that only proper upbringing and education gave people the capacity for sobriety and providence. He found workers seriously deficient in both. Thus, the conclusion that workers had few, if any, legitimate grievances, or at least that listening to their complaints was a mistake, had a powerful hold on managers. Euchène once devoted considerable energy to "proving" with documents from the company savings plan that poorly paid workers were inclined to save more than better-paid ones.^[84] That this engineer, with years of rigorous analytical training behind him, could present so flawed a study with conviction attests to his capacity for self-delusion about industrial relations. Yet he was hardly unique.

Benevolence was of course the essence of paternalism from the workers' point of view, but not from the managers'. Devoted to duty and to hard work for its own sake, the engineers conceived of their responsibility toward workers mainly in terms of ordering, guiding, and supervising them. In several corporate publications managers proclaimed their adherence to a "firm, prudent, and paternal administration."^[85] A cautious moral superintendence was the essence of managerial paternalism. Though circumstances eventually forced the executives to enter into a dialogue with the labor force, not for a moment did they regard it as legitimate. Indeed, they believed that such a dialogue was fraught with dangers.

This framework for paternalistic management provided justification for the decision makers to do what they wished and ignore what they wished. There was, as well, another limit to their paternalism. The object of the managers' ultimate loyalty was not the moral ordering of society but rather the stockholders' interests. "Firm, prudent, and paternal administration" meant profits, not disinterested service to workers. Most benevolent gestures were costly, and managers could not bring themselves to spend the stockholders' money when the necessity was not self-evident. Hence their paternalistic acts arose as a response to threats and were rather transparent attempts at social control. The impetus for a broader program of reform lasted only as long as the managers felt a loss of control.

The case of the PGC frustrates scholarly efforts under way to rehabili-

[83] Ibid., report of Servier, June 20, 1865.

[84] Ibid., no. 1520, "Rapport de Monsieur Euchène: Mission en Angleterre du mois de mai 1893," 2: 14-16.

[85] See, for example, ibid., no. 1290.

tate the reputation of the nineteenth-century patronat.^[86] The gas managers were calculating and manipulative but not spontaneously generous— despite the enormous profits and freedom from market pressures. To convince Louis Napoleon to confer the gas concession on the new firm, its founders tried to appeal to his socialistic pretensions by promising model housing for workers. The company did in fact draw up plans when it built the enormous new plant in La Villette, then a desolate comer of Paris. When the engineers found that they could assemble a sufficiently large labor force without the housing, however, they abandoned the project entirely.^[87] Plant superintendents were well aware that stokers, who worked in scorching heat and clouds of coal dust, required an adequate shower room at each plant. The head of the Saint-Mandé plant characterized his facility as "immoral, wretched." Yet top management was slow to find the funds for improvements.^[88] Two of the corporation's most solidly established paternalistic institutions, the medical plan and aid to aged workers, were financed through a 1 percent withholding from the payroll. However, the company jealously guarded its control over the allocations and used the power to discipline the work force.^[89] Nor were managers true to their oft-repeated word that encouraging workers to save for the future was one of their unavoidable duties. When wage earners requested a retirement plan to which they would contribute—and thus have a right to a pension—the director rejected the demand as unworkable. Engendering habits of thrift was managers' goal only when it reinforced their control over the personnel.^[90] Indeed, after years of hoping that stokers would take their

[86] Peter Stearns, Paths to Authority: The Middle Class and the Industrial Labor Force in France, 1820-1848 (Urbana, Ill., 1978), and, much less explicitly, Michael Miller, The Bon Marché: Bourgeois Culture and the Department Store, 1896-1920 (Princeton, 1981).

[87] AP, V 8 O , no. 765, "Rapport sur les maisons ouvrières (2 octobre 1856)"; no. 671, deliberations of November 28, 1866; Journal de l'éclairage au gaz , 1853, no. 2 (June 15): 234-235.

[88] AP, V 8 O , no. 768, "Rapports des régisseurs des usines," reports of May 24, 1868, and January 20, 1867. The Vaugirard plant did not have hot water until 1869. See no. 673, deliberations of November 3, 1869.

[89] The PGC had promised to supplement the caisse de prévoyance with a sum equal to the 1 percent, but it ran the fund so frugally that it did not have to dip into the corporate coffers until 1890. See ibid., no. 723, deliberations of March 11, 1858; no. 149, "Caisse de prévoyance." The company created the post of "inspector of the ill" to check on those claiming sick pay. See no. 678, deliberations of September 27, 1876.

[90] Ibid., no. 155, "Retraite ouvrière"; no. 148, "Procès-verbaux de la cinquième audience donnée le 10 mai 1892." Director Godot refused to allow the workers to collect for their mutual-aid society on the shop floor, but the company contributed annually to the mutual-aid society of the Anciens é1èves de l'Ecole Polytechnique.

wage increases in the form of contributions to the company's pension fund (which the PGC accorded as discretionary rewards), Arson pronounced the workers' proposal a "dangerous practice."^[91] To be sure, the engineer-managers were themselves hardworking, self-sacrificing, and modest in their aspirations; they asked nothing more of workers than they asked of themselves. But however true, that assertion does not alter the point that their paternalism was authoritarian and calculating rather than benevolent.

<><><><><><><><><><><><>

These authoritarian paternalists had to formulate specific policies under circumstances that did not always favor their stern principles. Adverse conditions prevailed especially during the first fifteen years of operation. The industrial progress of the Second Empire, accompanied by rebuilding of the capital, created peaks of prosperity that at certain moments gave labor some leverage in dealing with management. Average Parisian wage levels rose more than 20 percent during the 1860s.^[92] Moreover, the PGC significantly augmented the amount of work it extracted from its laborers, particularly stokers, when it rationalized the distillation process. Such conditions generated recurrent labor unrest at the gas plants during the 1860s.

That labor was in short supply, that general wage and price levels in Paris were rising, and that the company was getting ever more work from its personnel did not bring the PGC's engineers to accept tactfully the necessity of occasional pay increases for their underlings. Instead, the mangers railed against the "absurd demands" and "outrageous pretensions" of the workers—even as they calculated that a raise would not halt the decline of unit labor costs. The pay increases they finally accorded came in response to agitation, rarely in anticipation of it. Factory managers resorted to measures of "salutary intimidation" (to use the term of one superintendent) to forestall strikes and always dismissed the leaders of the movement. Even workers who eschewed strikes and politely led delegations to discuss workers' grievances could expect to be dismissed as a matter of course. Like gas consumers, the PGC's workers had to struggle for whatever improvements they were to receive.^[93]

[91] Ibid., no. 711, "Rapport no. 73: Caisse de retraite."

[92] See Jacques Rougerie, "Remarques sur l'histoire des salaires à Paris au XIX siècle," Le Mouvement social, no. 63 (1968): 71-108, for indexes of industrial activity and wage levels.

[93] AP, V 8 O , no. 148, report of June 28, 1865; no. 748, report of Jones to director, April 9, 1861; no. 1268, report of the superintendent of Ternes to Arson, April 26, 1867. To place such behavior in context, consult Michelle Perrot, Workers on Strike:France 1871-1890, trans. Chris Turner (New Haven, 1987), pp. 249-280. The PGC was by no means aberrant.

Managers had difficulty adjusting to an era of rising pay levels because they regarded wages as a subject of morality and authority as well as economics. In truth, they did not want market forces to set wages any more than they wanted them to set the prices of the PGC's products. Rather than according legitimacy to the market as final arbiter of pay, managers attempted to create an internal labor market, in which laws of supply and demand yielded to their moral appreciation of what workers should earn. In exchange for relatively steady employment (in a heavily seasonal industry) and some minor fringe benefits (like the medical plan and small retirement grants), managers expected workers to leave wages to their judgment. That such an arrangement was neither attractive to workers nor realistic did not diminish the engineers' indignation when laborers rejected it. Of course, the common hands were not readily able to brush aside managers' control over them. The implication of the internal labor market for them was low pay relative to other firms offering comparable jobs.

Far from allowing labor to share in the copious corporate profits, management pursued a deliberate low-wage policy A factory superintendent noted in 1865 that the daily 2.75 francs paid to coke haulers could attract "only the old and infirm" by the current standards in Paris. The head of the coke department explained in 1876 that wages for carters (2.75-3.75 francs a day) were far less than mere navvies could earn at most plants. Hence the company had to hire "only the rejects—weak or bad workers in all regards." The lighting engineer brought the director's attention to the plight of the greasemen when their rate of quitting tripled between 1876 and 1881. Since they had not had a raise since 1855 and did thirteen hours of responsible work a day, the engineer had to conclude that their pay might need some upgrading. The director, however, rejected the advice.^[94] These examples of lower management seeming to plea for higher pay should not mislead. They agreed with their superiors on the desirability of low wages, and requests for increases came only when the inability to attract competent workers threatened to disrupt operations. Even in the early 1890s, when the company's labor practices were under public scrutiny as never before, little had changed. The assistant factory chief assembled comparative wage data in the hope of showing the municipal

[94] AP, V 8 O , no. 148, report of Devillier, May 1, 1865; report of Perrin to director, December 24, 1881; no. 90, report of Hallopeau to director, January 14, 1867.

council that the PGC's laborers had no grounds for complaint. His findings demonstrated instead that the firm's pay was on a par only with disreputable industries of the banlieue, glue factories, animal-rendering plants, and the like. At first the engineer tried to explain away the obvious implications through strained reasoning, but eventually he buried the report.^[95]

Not only did the internal labor market mean a low-wage policy, pursued to the extent circumstances allowed; it also meant unequal pay for men doing the same work. The company, for example, employed two groups of carters who delivered alkalies to industrial customers. One received a daily bonus of a franc, and the other did not. Their superior could not justify the differential in terms of work routines. Such pay differences for similar work did not embarrass management nor disturb its sense of equity. Indeed, the director refused to establish companywide standards and even encouraged factory superintendents to pay according to different scales because he wanted to engender divisions among the laborers. The executive committee would not consider a wage demand from stokers on the grounds that the workers had passed the petition from plant to plant.^[96] Unequal pay was an important component of the managers' mission for moralizing workers. Whenever possible, the engineers transformed simple jobs into "careers" by introducing hierarchical pay steps achieved through seniority. Thus, coke haulers would earn 2.50, 2.75, or 3.00 francs daily depending on their years with the PGC. This "segmented labor market," as Richard Edwards calls it, was by no means the invention of the twentieth-century corporation.^[97] The PGC was thoroughly wedded to it as an instrument of control and moralization. Managers justified the practice as an "encouragement to emulation." They undoubtedly viewed it as a natural extension of the career-based pay that engineers themselves had. In all probability workers were impressed mainly by its arbitrary aspects. When they finally developed a collective voice in the 1890s, they complained about being "tyrannized by twenty-five centimes."^[98] Yet the general laborers could not usually resist the will of the engineers, who had scarce jobs to offer. The stokers, by contrast, had the ability to resist and did so. The director tried to impose "grades" of pay on them in 1858 but

[95] Ibid., no. 148, "Renseignements sur les salaires journaliers des ouvriers pris chez différents fabricants."

[96] Ibid., no. 157, report of Linote to Audouin, October 20, 1898; no. 665, deliberations of June 6, 1857; no. 148, "Etat, par usine et par profession, des salaires"; report of Gigot, December 20, 1891.

[97] Richard Edwards, Contested Terrain: The Transformation of the Workplace in the Twentieth Century (New York, 1979). Edwards portrays the segmented labor market as a post-World War II strategy of labor control.

[98] AL V 8 O , no. 148, "Briqueterie et goudrons, 8 avril 1892."

had to withdraw the order within a few months.^[99] The paternal authority that managers affected led them to create an internal labor market that was characterized not by a carefully considered beneficence but rather by low and inequitable wages. The managers could not imagine—nor did they care—how arbitrary their governance might have seemed at times to the labor force.

Armed with their comfortable assumptions about the futility of conceding raises, managers did not do so willingly. When the superintendent of the Vaugirard plant faced an acute shortage of stokers in 1865, he struggled to find an alternate solution. He decided to offer paid overtime work but worried that even that answer might create "unfortunate precedents." Several years later, the manager of the Passy plant railed against stokers who wanted a pay increase when he raised their work load (by having them fill larger retorts). For him, theirs remained "an incredible demand."^[100] The managers' grave reservations notwithstanding, they were sometimes compelled to yield. Then they tried to meet the essentials of workers' demands, but on their own terms and with moralizing provisions attached. Such reformulations of the concessions were important to managers, for that was how they preserved the sense of paternal authority. Surrendering to strikers in 1865, the company granted not a raise but a "bonus" for regular attendance. Stokers would receive fifteen francs a month if they had no unexcused absences. Arson solemnly instructed plant superintendents that the bonuses had to be earned. Superintendent Philippot wished to go further; he favored paying bonuses only at the end of the year "to encourage a spirit of economy and moralization." He also suggested putting the extra pay in savings accounts and making stokers build a nest egg.^[101] None of these ideas obtained. The harsh realities of a tight labor market militated against moralizing projects, even slender ones.

In 1876 the PGC faced another inflationary surge too powerful to ignore. The company capitulated to workers' demands and once again put a patriarchical gloss on its concession. This time the firm accompanied the raise with the creation of a voluntary savings plan (caisse d'économie ), into which workers, it was hoped, would place their pay increases.^[102] The

[99] Ibid., no. 666, deliberations of August 3, 1858, and March 9, 1859.

[100] Ibid., no. 749, report by superintendent of Passy to Arson, April 6, 1869; no. 148, report by Gigot, April 30, 1865.

[101] Ibid., no. 1081, ordre de service no. 253; no. 148, report of June 25, 1865; no. 763, "Distribution des primes."

[102] Ibid., no. 1081, ordre de service (July 18, 1885). Director Camus noted in this directive that he "attaches a great importance to the caisse " and ordered superintendents to "exercise their personal influence to encourage [provoquer ] deposits." Camus insisted that the savings plan should receive only the savings put aside from the raise; he did not want savings from any other source. See no. 678, deliberations of October 18, 1876.

head of the factory division ordered superintendents to assemble the workers and harangue them on the benefits of putting aside their raises. At the Boulogne plant, workers listened respectfully but protested that higher costs of living would consume their gains. The plant manager, resigned to the workers' intransigence, probably did not even sense their evident resentment at the suggestion that the raise was superfluous.^[103] Despite the paltry sums actually put in the savings accounts and the small number of savers involved (usually office clerks, in any case), the director reported on the plan at every stockholders meeting as if it were a fundamental component of the business. The engineers may have forced a moralizing interpretation on wage inflation, but it is likely that they spoke intelligibly only to themselves.

The PGC's decision makers had to struggle with the economic—and the moral—implications of incentive-based pay. Twice during the 1860s the managers met labor agitation with a compromise, retaining a basic rate but adding incentive pay if stokers distilled more than a specified amount of coal. They resisted a full piece-rate plan, however, and granted it only under duress.^[104] Managers found incentive-based pay a problematical issue because it involved a principle they did not willingly accept: the firm might derive benefits from giving workers higher pay. Factory superintendents were united in their opposition to piece rates. They argued that they were able to extract more work from the stokers without it. Furthermore, they feared the wage bill that might result from incentive pay—even if the company was profiting from enhanced productivity. By contrast, their superior, Arson, advocated piece rates on abstract grounds, as a matter of justice and of moralization. Yet even he retreated from his support of full incentive pay when he realized that stokers' wages could reach 9.50 francs a day if productivity continued to rise at its current rate. The convoluted mixture of economic and moral calculation inherent in managers' wage policies was reflected in Arson's reasoning: "This figure is far too large to contemplate. Whatever the interest the company has in realizing such a level of labor output, it cannot reasonably consent to pay

[103] Ibid., no. 714, report of Hadamar to Arson, April 13, 1876.

[104] Ibid., no. 715, report of December 14, 1871; no. 716, "Salaire des chauffeurs (16 mars 1880)"; no. 673, deliberations of April 21, 1869. It is useful to recall at this point that stokers' mode of pay did differ from one plant to another up to 1880.

[those wages]."^[105] Arson's brand of paternalism did not allow him to rejoice in the coincidence of high wages and corporate profits. He and the rest of management dung to moral standards that denied the validity of high wages under any circumstances.^[106]

Management lost its struggle to keep wages within morally permissible limits and even to keep unit labor costs down during the last great burst of Parisian prosperity of the century, between 1876 and 1882. The general inflationary spiral, the exceptional strains posed by the Universal Exposition of 1878, and the labor scarcity created in the PGC itself by the opening of a new plant in Clichy and plans for a new one in Saint-Denis defeated the engineers' efforts to restrain wages. Probably on the insistence of stokers, engineers finally accepted piece rates in 1880 (thereby preserving some semblance of moralization).^[107] Still, pay began to rise faster than productivity for the first time since the creation of the PGC. By this time, as well, the corporate profits were a matter of public scandal, and it was rather convenient to point to the "exorbitant" wages earned by stokers to justify the firm's returns. Thus, a high-wage policy—for stokers, at least—arrived just as the PGC needed it politically.^[108]

The dawn of the electrical era found management resigned to paying stokers quite well by the standards of the day and to devoting an ever-increasing portion of corporate expenditure to labor. The path by which they came to such a situation had no doubt frustrated workers as much as managers and convinced laborers that decisions about pay were arbitrary and capricious. In any case, the focus of industrial relations was about to shift from wage matters to the question of authority. On that subject, too, managers had to adjust their inclinations to unwelcome realities.

[105] Ibid., no. 711, rapport no. 43, "Règlement du salaire de la distillation"; no. 716, "Salaire des chauffeurs (16 mars 1880)."

[106] For another example of concerns that workers would earn "too much" despite rising productivity, see ibid., no. 672, deliberations of May 30, June 3, and June 6, 1868.

[107] Ibid., no. 680, deliberations of March 19, 1880. The police reports noted that the workers had demanded piece rates and that management had granted that mode of pay reluctantly. See Prefecture, B/a 176, report of March 19, 1880. The assessments by the police fit the corporate documents, which stress the "change of mind" among the engineers.

[108] On the jump in labor costs, see ibid., no. 1016, "Prix de revient." Having granted "exorbitant" wages to stokers, the company sought to contain costs by turning to foreign workers, who were paid less. By 1892, 40 percent of the stokers at the La Villette plant were foreign. See no. 90, "Etat numérique des ouvriers italiens"; no. 148, "Union des agents des chantiers à coke."

The managers' authoritarian paternalism was incompatible with the efforts, however modest, by politicians of the left to republicanize the workplace, to endow workers with basic rights before their employers. The PGC's decision makers took the reforms as an infringement on their rightful authority. The law of December 27, 1890, provides an example of the conflict. It was one of the few concrete benefits workers received from the Opportunist Republic. The law called for a week's pay when laborers hired for an indeterminate duration were suddenly dismissed. The PGC could not abide the legislation and refused to honor it. Several workers sued the company and won. The director beseeched the corporate attorney to find some loophole in the law, for he viewed it as a threat to the authority of the firm. "If a foreman cannot fire a worker at will without inflicting a cost on the firm," the director pleaded, "what has become of his ability to discipline?" Alas, the barrister was not able to discover convenient ways of circumventing the legislation. At one point he recommended hiring every worker with a daily written contract but conceded that the necessary paperwork would be overwhelming. In the end he could only suggest one more show of "salutary intimidation" to keep workers from exercising their newly acquired right: have foremen inform laborers that they would never be rehired if they dared to sue the company.^[109]

The firm's confrontation with this piece of republican legislation was prophetic. In the same year that the law passed, the company recommenced negotiating with Paris over the fate of the gas concession. The bargaining was to place the PGC in a most uncomfortable position regarding its industrial relations. The discussions began just as the left majority of the municipal council came to a new understanding about the role of government in industrial society.^[110] The city improved its workers' wages in time to demand the same from the PGC. Wages were not to be the central issue, however. The council also encouraged the creation and growth of unions among the personnel. At the very least, councilmen expected the PGC to comply with the law of 1884, which gave workers the right to unionize free of overt persecution. Many Radical aldermen also wished to see fuller cooperation between management and the unions as well as significant concessions on pay. For the foreseeable future it appeared that the PGC faced an intermediary authority between its executives and a heretofore atomized personnel.

The union posed a formidable threat to the company's authoritarian paternalism. At the same time, it presented a genuine opportunity for a

[109] Ibid., no. 90, "Ouvriers renvoyés ou congédiés."

[110] See chapter 2.

breakthrough culminating in cooperative industrial relations and collective bargaining—if the PGC wished to pursue that path. The workers' union formed in late 1891 showed every sign of being moderate, conciliatory, and respectful of managerial authority. The leaders initiated contacts with management by sending a humbly worded request for an audience with the director. They formally disavowed any intention of striking and described their purpose as exposing the "modifications" that workers desired. The letter ended by assuring the director that they would be pleased if the company conceded only a part of the requests.^[111] That this statement accurately reflected the pacific spirit of the union is established by the fact that workers originally joined with white-collar employees in the same organization, and the clerks by no means included strikes in their arsenal of tactics. The union leaders were delighted when the recently appointed director, Godot, met with them and were encouraged by his ostensibly conciliatory approach. The director tested their goodwill by asking for secret data on the membership of the union, and the leaders gave it to him. Godot was astonished by their consummate naiveté, and the president of the Bourse du Travail (an umbrella organization of Parisian unions) had to issue a severe reprimand to the leaders for violating its rules. The union, then, was clearly not in the hands of militants or revolutionary ideologues. The inexperienced delegates even stumbled into a corporatist stance by promising to be advocates of the PGC before the municipal council in return for concessions.^[112] Implicitly the leaders recognized a legitimate sphere of intervention, entailing compensations, while conceding control over the production process to management. Considering that the PGC was committed to spending more money on its personnel for political reasons, harmonious relations with the union were not out of the question.

Managers, however, never had any intention of negotiating with the union as a bona fide representative of its personnel. The union was not at all a part of the corporate program for social peace. In fact, the director's first response to the formation of a union was to consult with the legal experts about the minimum steps he could take while still adhering to the law of 1884.^[113] Throughout 1892 Godot, who had his own program for

[111] AP, V 8 O , no. 151, Chambre syndicale to director, December 3, 1891.

[112] Ibid., no. 148, "Rapport sur la réunion du mercredi 11 novembre 1891"; clipping from Le Temps , December 31, 1891; report on union meeting of July 25, 1891; no. 150, union's executive commission to director, September 27, 1892. See below, chapter 8, for a fuller discussion of the union's orientation.

[113] Ibid., no. 150, "Syndicats: Jurisprudence." The company did try to intimidate workers so they would not join the union, but the steps could not be too heavy-handed. See no. 148, "Procès-verbaux de la 4 audience donnée le 3 mars 1892."

industrial reform, feigned good intentions because he wanted union activists to bring favorable reports to the municipal council at this moment of intense negotiation over the gas monopoly. When discussions with the city collapsed at the end of the year, the director coldly informed union leaders that henceforth they could communicate with him only in writing. The pretense of collective bargaining ceased, and the municipal council became the intermediary through which the union and the company interacted. The director treated petitions from the leaders as if they came from the individuals who signed them and petitions from the union via the council as if they came from the aldermen.^[114]

Faced with the unpleasant reality of industrial unrest, the PGC's managers became conservative reformers. As such, they were part of a wider movement. Pierre Sorlin has noted that "in the last decade of the century, and particularly between the municipal election of 1892 and the legislative election of 1898, the fear of socialism was a dominant characteristic of French political life."^[115] Monarchists, Catholics, and conservative republicans rallied to the defense of private enterprise. The theorists of social defense called for private, paternalistic initiatives to win workers away from dangerous ideas. These leaders portrayed the workplace as a "natural" social unit, like the family. Emile Cheysson, a particularly influential spokesman because he was an engineer and a professor at the Ecole des Mines, stressed the special role his profession might play in forging social peace. Scorning "selfish" management that consulted only short-term, bottom-line interests, Cheysson counseled engineers to find ways to reconcile laborers to the industrial order through a reasonable, rational generosity.^[116] He and other bourgeois reformers were responding to a moment of economic transition in French society and a crisis of the liberal order just as surely as the Radical solidarists, whose answer to the crisis was to give wage earners republican rights. The management of the PGC was well positioned to listen to the recommendations of the bourgeois reformers.

[114] Ibid., no. 733, deliberations of June 8, 1893; no. 162, "Affaire Leroy." The company ultimately avoided defining the precise status of the union as a permanent institution by adopting the stance that the director was "always" prepared to listen to workers who had something to say to him. The pretense was that the workers were (or should be) unaffiliated individuals. See no. 150, "Projet de lettre, le directeur à Lajarrigue (n.d.)."

[115] Pierre Sorlin, Waldeck-Rousseau (Paris, 1966), p. 358. Sorlin's insights have been amplified by Lebovics, Alliance of Iron and Wheat.

[116] Emile Cheysson, Le Rôle social de l'ingénieur (Paris, 1897); Sanford Elwitt, The Third Republic Defended: Bourgeois Reform in France, 1880-1914 (Baton Rouge, La., 1986).

The paternalistic rhetoric of conservative reform recommended an ordering of industrial relations in precisely the terms that the PGC's managers understood and accepted. The view that social peace depended on employers' goodwill and workers' passive acceptance of their superiors' responsible gestures inevitably made sense to them. There can be no doubt that certain executives of the company were well integrated into the reform movement. Godot was a personal friend of Cheysson.^[117] The brother of the new chief of the factory division, Albert Gigot, was a founding member of the Musée social, the Comité de défense et de progrès social, and the Alliance d'hygiène sociale, all centers for disseminating reform projects.^[118] Management could be expected to consider reform proposals in devising its response to the workers' challenge, such as they imagined it to be. Still, managers turned to industrial reform when they had to. No amount of admiration for theory could have made the engineers overcome their distaste for raising labor costs when they were not compelled to do so. Director Godot may have been something of an ideologue; his appointment was a display of good intentions at a crucial moment. The hard-headed Emile Camus, however, was still in charge (as "delegated board member") and could keep Cheyssonesque influences from getting out of hand. Under duress the PGC became a laboratory for social reform.

Industrial paternalists like Cheysson and Gigot had nothing to say about unions; they presumed that the proper policies would forestall their formation. Thus, the reform literature did not directly address the PGC's most immediate problem: how to make concessions to workers without strengthening the union's influence. Partly drawing on established corporate practice and partly yielding to the insistence of Radical aldermen, Godot and Camus devised a dual strategy. First, the company would woo laborers away from the union through responsible generosity, thereby winning the approval of aldermen as well. Second, the IGC would improve workers' pay without specifically yielding to any union demand. This policy, replete with risks, inaugurated nothing less than a battle between the union and the corporation for the hearts and minds of workers.

[117] AP, V 8 O , no. 1294, Emile Cheysson to Director Godot, April 16, 1890. Cheysson addressed Godot as mon cher ami.

[118] Sanford Elwitt, "Social Reform and Social Order in Late Nineteenth-Century France: The Mush Social and Its Friends," French Historical Studies 11 (1980): 443. Gigot was the author of Les Assurances ouvrières et le socialisme d'Etat (Paris, 1895). See also W. Tolman, Que doit le patron à ses ouvriers en plus du salaire ; suivi d'observations de MM. Levasseur, Gigot, Blondel, Delaire (Paris, 1901). Gigot had served as préfet de police in Paris from 1877 to 1879 and was director of the Forges d'Alais in 1892.

The worker's union, at this point, represented mainly the various sorts of common laborers outside the distillation room—carters, coke handlers, lamplighters, greasemen, and navvies.^[119] The long list of union demands contained primarily the specialized concerns of each of these groups. Underlying the detailed lists were four fundamental grievances.^[120] First, the union pressed for the equalization of pay and hours among operating units and among men who did similar work. Second, it hoped to find some way to alleviate the petty tyranny inflicted on its members by foremen and other supervisors. Workers of nearly every occupational category complained about the surliness and capriciousness of their immediate supervisors. Leaders long had difficulty translating the grievances into specific demands because of their wariness about infringing on managerial prerogatives. Third, the unions asked for raises for almost all categories of laborers except the well-paid stokers. Demands were generally modest, in the range of twenty-five to fifty centimes a day. Fourth, the union representatives insisted on a minimum wage of five francs a day for all workers. The city of Paris had just passed such a provision for its laborers. According to data from the union, 1,256 of some 5,900 workers at the PGC still earned less than that figure.^[121] This demand was the emotional centerpiece of collective bargaining as far as the union leadership and the rank and file were concerned. Certain aldermen also regarded the minimum wage as a test of management's seriousness about social peace.

For their part, the engineer-managers gambled that the rank and file were not attached to these positions or at least would be grateful to a firm that gave them other valuable benefits. Parsimonious as they had been before 1890, the executives were now ready to increase substantially their expenditures on labor to defeat the union and create some goodwill at the Hôtel de Ville. Their concessions intentionally differed from the union's demands, but the managers hoped they would address some of the basic discontents. Management was also perpetuating corporate tradition in disregarding the specific grievances articulated by workers, substituting their own preferences and adding moralizing elements to the concessions. Thus, laborers who asked for raises were instead given opportunities to do extra work for pay. Another alternative to a raise, pure and simple, was to increase the portion of workers in the higher "grades" on the pay scale,

[119] Stokers had formed a union in 1890, but it had fallen apart. See below, chapter 8.

[120] AP, V 8 O , no. 148, contains many lists of union demands.

[121] Ibid., no. 626, "Etat numérique des ouvriers des usines gagnant un salaire inférieur à 5 francs." The number would have been far larger before 1890, for the company responded to the strike of that year with a general round of raises.

thereby expediting advancement.^[122] The most ambitious and costly elements of the corporate program were the pension plan for manual laborers and the profit-sharing bonuses (see chapter 2). Prior to the new era of industrial relations, managers had denied that a regular retirement fund was feasible on the grounds that the manual work force was too unstable. The strategy of social defense and the struggle with the union suddenly made the reform viable. The profit-sharing plan had the attraction of moralizing workers while, presumably, keeping their minds off the five-franc minimum wage. By no means did the package of reforms represent a well-conceived plan for social defense. It was the product of a management that was on the defensive, subjected to multiple forces beyond its control and profoundly distraught by the predicament. Left to themselves, the engineers would surely have continued their frugal authoritarian paternalism.^[123]

After promulgating the profit-sharing plan in November 1892, it was time for management to weigh its success in convincing workers that corporate beneficence was a surer path to improvements than union organization. Factory superintendents, ordered to study labor's response to the announced benefit, found that the union was winning the battle for support, at least in the short run. The superintendent of the Passy plant reported that half his laborers resentfully expressed their preference for the five-franc minimum wage, not the profit-sharing plan. At Landy wage earners told their boss that they would rather have five francs each day than wait all year for a 113-franc bonus. Gigot summarized the reports by arguing that the concessions had reinforced the popularity of the union. He also detected an evil cycle whereby one concession produced still more demands for others.^[124] The company appeared to have blundered. The profit-sharing program was nearly three times more expensive than the minimum wage would have been but provided little satisfaction. Furthermore, the program would continue to disappoint workers because profits were not healthy in the 1890s. The bonus never attained the forty-three

[122] Ibid., no. 1081, ordres de service nos. 340-369.

[123] Recall that the company decided on the profit-sharing plan in the context of negotiations over the Sauton project. In particular, the agreement to put 1 percent (later 2 percent) of the profits aside for workers came in the context of discussions over reducing meter rental rates. The municipal council settled for a lower reduction in the rates when the company promised to bestow on workers the profit-sharing plan. See ibid., no. 154, Troost and Director Camus to president of municipal council, June 12, 1892.

[124] Ibid., no. 148, "Répartition du 2 per cent," report of Gigot, June 13, 1893. The police confirmed the preference of the workers for a minimum wage. Préfec-ture, B/a 1424, report of November 24, 1892.

centimes a day that Godot initially promised. The supplement of thirty-eight centimes that workers received in 1893 fell to thirty-three in 1895 and to thirty in 1899.^[125] Workers inevitably came to believe that they were sharing the problems, not the profits, of their firm.

Managers, unaccustomed to dealing with these sorts of issues and obviously uncomfortable in the role, miscalculated on two levels. Their long-standing disdain for what workers had on their minds led them to underestimate the emotional hold of the union's proposals.^[126] More fundamentally flawed was the engineers' attempt to reduce enthusiasm for the union by granting major concessions after a strong organization was in place. Laborers would surely credit the union for the improvements. The corporate program of reform was not necessarily too little, but it was certainly too late.

In fact, there was not complete accord among managers on the strategies for dealing with organized labor. Aside from individual differences, there was an inevitable divergence of perspective between top and lower management. The director, fully aware of the expectatons of the public authorities and intent on prolonging the life of the firm, realized that open confrontation was dangerous. Production managers, however, in direct contact with workers and immediately threatened by the union, took a narrower view. They wished for a more aggressive policy of repression. The elderly Curry, the most senior superintendent, displayed a paranoia about organized labor from the first moment. He affirmed (in 1892) that leaders formed an "occult authority behind my back"; he asserted that they made threats with impunity; he described the union secretary as "doing whatever he wished" and treating Curry "as an equal." Biju-Duval, the manager of the Saint-Mandé plant, assumed that his duty was to "put the union delegates in their place" before they became too heady with power The superintendents were unanimous in proclaiming that the union was the cause, not the result, of the workers' discontents and that industrial relations had been superbly harmonious before the disrupting organization emerged. Gigot, as head of the factory division, pleaded with

[125] AP, V 8 O , no. 149, "Sommes nécessaires pour payer un salaire minimum de 5 francs"; no. 151, "Montant total des allocations de supplement de salaire"; no. 156, "Gratifications: personnel secondaire." The minimum wage would have cost about two hundred thousand francs a year to implement; the profit-sharing plan required in the range of six hundred thousand francs.

[126] Rolande Trempé might explain the workers' preference in terms of how they conceptualized their wages. The minimum wage related to a view of pay as a means of subsistance. The laborers apparently did not think of their earnings as a share of the wealth they were producing. See Trempé, Les Mineurs de Carmaux , 1848-1914, 2 vols. (Paris, 1971), 1: 369-379.

Godot not to make any concessions on the grounds that they would only reflect well on the union. When the director discounted his advice, Gigot took the unusual step of bypassing sacred hierarchical channels, conducting an unsolicited survey of superintendents' views, and passing the results on to the director. Predictably, lower-level managers regretted the concessions and foresaw the victory of the union in the battle for workers' favor.^[127] Exactly what positive steps superintendents would have had the company take was not a matter they considered. They disregarded proclamations of the municipal council in favor of workers, which so threatened hopes for a renewed charter. Perhaps they would have preferred from the start the demise of the PGC—along with their careers in the firm—to its continuance under existing circumstances. Ultimately lower managers did prove willing to sacrifice their careers to the cause of authoritarian paternalism, but in the early years of the union era top management still hoped for a less drastic solution.

<><><><><><><><><><><><>

Events between the end of 1892 and 1899 seemed to vindicate the intransigence of production managers. Negotiations with the city collapsed, and membership in the union continued to grow. In the meantime, the PGC had burdened itself with much greater labor costs without having won the loyalty of its personnel. The company had to live with expensive, paternalistic policies that it would not have adopted spontaneously. It is no wonder the strategy of top management edged closer to the repressive recommendations of the production engineers.

The breakdown of bargaining with the city at the end of 1892 removed a powerful incentive for maintaining civil relations with the union. The director dropped the pretense of giving the delegates and their grievances a favorable hearing. Union leaders were forced to depend on the municipal council to communicate with the company. Throughout the mid-1890s the director continued to receive many petitions, copies of which had first been sent to the council. He would quickly dispose of the requests by explaining why they were out of the question. Labor's largest victory in the mid-1890s, the six hundred thousand francs allocated to improve wages and salaries in December 1896, arose because the city contributed half the sum and shamed the company into granting an equal amount.^[128]

Having failed to win the hearts and minds of workers, top management

[127] AP, V 8 O , no. 148, report of Cury, March 24, 1892; report of Gigot, December 30, 1891; report of Cury to Gigot, June 13, 1893; report of Biju-Duval to Gigot, June 13, 1893; no. 150, report of Cury to Gigot, June 5, 1893; report of Gigot to director, April 5, 1893.

[128] See chapter 2.

now pursued a policy of containing the union's growth in any way that avoided open confrontation. Godot or Camus devised a strategy for ostensibly honoring the reform commitments that the company had already made while in practice undermining them. The strategy worked by taking advantage of the decentralized administration of the factory personnel and unleashing the festering antiunion sentiments of lower management and supervisors. Union leaders were to find that no one would take responsibility for blatant violations of concessions that they thought had become acquired rights.^[129] Particularly frustrating to the syndicalists was the violation of promises concerning layoffs. The director had affirmed that layoffs, massive in this seasonal industry, would occur strictly according to seniority. Yet the delegates could point to case after case of open infractions.^[130] The union also found itself powerless to stop other abuses—discriminatory firing practices, injustices in fining, intimidation, and demeaning gestures from supervisors. When union delegates complained to Godot, he feigned surprise and promised to look into the matter; but the complaint was lost in the hierarchical chain of command. Occasionally, Godot told the union officials that he was unable to make the superintendents obey his orders to stop the abuses; all he could do was urge the plant managers to abide by the agreements.^[131]

The supervisors, for their part, excused the abuses by placing blame on foremen and other immediate supervisors. Following Godot's example, the plant managers affirmed their inability to control their underlings. Foremen too were profoundly antiunion, seeing the union as a challenge to their authority Neither the superintendents nor the director were willing to take effective steps to curb their abuses. Indeed, management was pleased to take advantage of the foremen's hostility. The engineers did not

[129] The PGC had actually used the policy earlier on the issue of foreign workers. Public pressure had forced the director to promise a reduction in the number of foreign laborers in 1885 (not long after the firm had accelerated its hiring of them). The director issued formal orders, but somehow nothing came of them. See AP, V 8 O , no. 1081, ordres de service, September 29, 1885, and June 23, 1892. Workers frequently accused their supervisors of favoring foreign over French wage earners.

[130] Ibid., no. 150, report of Leroy to Gigot, August 11, 1893; letter of Chapelle to director, August 8, 1893; letter of Darène to director, June 14, 1894; no. 148, "Prociès-verbaux de la 4 audience donnée le 3 mars 1893 aux délégués."

[131] Le Journal du gaz: Organe officiel de la chambre syndicale des travailleurs du gaz, no. 83 (May 20, 1896): 1; no. 110 (July 5, 1897): 1; no. 112 (August 5, 1897): 1; no. 139 (September 20, 1898): 1; no. 133 (June 20, 1898): 2. The director was careful never to promise that rehiring would occur by seniority, leaving open the possibility of eliminating difficult workmen.

even pause to consider that the supervisors could be using their authority not in the company's interest but in their personal interest.^[132]

The matter of decentralized authority and its abuses certainly separated the managers of the PGC from theorists of bourgeois reform. Emile Cheysson had pinpointed the foreman as a major source of industrial conflict. He argued eloquently that the foreman's arbitrary power must be curtailed and given to engineers, who, he believed, were much more capable of impartial and judicious decisions.^[133] Cheysson's warnings about the abusive authority of supervisors proved to be prophetic for industrial relations within the PGC (see chapter 8). Yet gas managers did not intend to act on his recommendation. Their willingness to exploit the prejudices of foremen for the purposes of disciplining the labor force and of persecuting union members illustrated the permanent gulf that separated their authoritarian paternalism from Cheysson's conservative but humanitarian paternalism.

<><><><><><><><><><><><>

The sour yet stabilized relations between the company and the union were destined to deteriorate in 1898. The implicit position of lower managers, that the company might not be worth preserving if a strong union were a permanent feature of industrial relations, then gained weight among top managers. The darkening atmosphere was a result of shifts in the union's membership and goals. Up to the last year or so of the century, the union had not attracted the stokers. As long as it represented mainly the common laborers, the company could easily countenance facing it down if need be. Management's strategy was centered on public opinion and politics, not concerns about the union's power to disrupt production and force concessions.^[134] This situation changed dramatically in the fall of 1898, however, when stokers began to join en masse. They brought a decided escalation of militancy; why they did so and how their joining was part of a wider revolt against industrial authority are points I shall discuss in due course. For the moment, we need to observe that the reorientation of union activity marked a dramatic heightening of tensions between labor and management.

The upsurge of militancy among the stokers engendered a renewed

[132] For the problems that foremen and other immediate supervisors engendered, see chapter 8.

[133] Cheysson, Rôle social de l'ingénieur , pp. 52-53.

[134] AP, V 8 O , no. 148, report of Gigot, April 15, 1892. Gigot referred to the stokers as "the essential element" in determining the kind of threat the union would pose.

campaign among lower-level managers to convince the director that the moment for decisive action had arrived. They warned that instances of disobedience, disrespect, and "arrogance" were becoming commonplace in the factories. Euchène, the chief production engineer, felt compelled to call for a return to the tradition of "paternal discipline."^[135] The superintendents could direct Godot's attention to prudent union leaders who for their own reasons were also alarmed. One of them spoke ominously, in February 1899, about "winds of revolution" and a "dangerous effervescence" in some production departments.^[136] While Godot's concern about a situation getting out of hand grew, factory managers vowed to reassert their authority.

The seriousness of the struggle over authority as well as the desperation of the managers to reassert their domination over the labor force was demonstrated by a bizarre symbolic confrontation in November 1898. The occasion was the funeral of a foreman, Bouttier, who was reputed to have been a model of equity in his treatment of workers.^[137] Union leaders wished to honor his memory by attending the last rites, offering a eulogy, and presenting a mortuary wreath. The widow gave her permission for the commemoration. The company, however, was intent on converting the funeral into a representational showdown with its adversary. Superintendent Hadamar (Bouttier's former boss) tried to get the widow to withdraw her approval for the union's presence. Having failed on that point, he intimidated her into arranging the procession so that the company's delegation (he and Euchène) would walk directly behind the bereaved family while the union members would be at the rear. Without his insistence on this protocol, Hadamar noted, "our dignity would surely have suffered." The superintendent made certain as well that the PGC's wreath preceded all others. Bouttier's last rites were, according to Hadamar's report, a brilliant triumph for the corporation over the union. His problem was to convert symbolic ascendency into reality.

Instead, Hadamar's worst fears were confirmed in March 1899, when a wildcat strike broke out at the Clichy plant, which he managed. On the urging of a foreman, Hadamar had fired two workers; the stokers, perceiving the dismissal as unjust, walked out of the plant. The production managers were unanimous in urging firm resistance, but Director Godot did not have the same freedom of action that he had had over Bouttier's

[135] Ibid., no. 159, "Grèves de mars et août 1899."

[136] Journal du gaz, no. 148 (February 5, 1899): 1.

[137] For this revealing incident, see AP, V 8 O , no. 159, report of Hadamar to Euchène, November 16, 1898.

widow. A new and urgent round of negotiations with the city over the gas monopoly was at hand, and the company faced the alternatives of winning a longer charter or preparing for its liquidation. The director was not willing at this crucial moment to alienate the councilmen of the left by perpetuating labor agitation, so he admitted that mistakes had been made and rehired the wronged workers. Still worse, from the superintendents' point of view, he allowed all the strikers to return to their jobs. Lower management was appalled by the concession. After every other strike production managers had at least been able to reassert their authority by dismissing strikers. Hadamar and his colleagues predicted a general collapse of order; indeed, they declared that such a breakdown was already well advanced. Their reports described a war of nerves between management and workers.^[138] Plant managers were preparing for a general strike, which they expected at any time. Apparently, they managed to convince Godot that the situation was dire, for he asked the minister of the interior to prepare for the army to intervene in case of a strike. The director sent blueprints of the factories to the prefect and requested permission to use soldiers as laborers if need be.^[139] Just a few months earlier, a series of strikes against managerial authority had erupted in the Parisian machine-building industry. Now, the managers of the PGC, from top to bottom, saw themselves as key combatants in the wider struggle to reassert their prerogatives.^[140]

The crisis for which engineers had been preparing came in August 1899. In a troubled and poorly organized move, stokers and some common laborers at several gas plants went on strike again. The demand for an eight-hour day at considerably reduced pay was ill conceived and divided workers more than it united them. To make matters worse, the union treasurer suddenly absconded with the strike funds and dealt a mortal blow to laborers' morale. A well-prepared management was easily able to find scabs and crush the walkout. This time the strikers were not rehired, despite intense pressure from the municipal council and from the press. The union, for the moment, lay in ruins.

Did management foment the strike and use it to crush its adversary? Journalists and politicians soon made that accusation. They did so on the grounds of a secret meeting between a manager and union leaders just prior to the walkout. The motive they attributed to the company was an

[138] Ibid., "Grève, mars 1899."

[139] Ibid., no. 160, Director Godot to prefect, July 4, 1899.

[140] On the wider context in which the gas' strike of 1899 took place, see Lenard R. Berlanstein, The Working People of Paris , 1871-1914 (Baltimore, 1984), chap. 5.

attempt to push the city into renewing the gas monopoly.^[141] The firm easily demonstrated that such evidence and reasoning were unconvincing. Yet the charge was not necessarily baseless. Journalists may have been correct, but for reasons they did not suspect. There is no conclusive evidence of the company plotting to create a strike, and one could certainly understand events without postulating such action. Nonetheless, it is entirely plausible that certain managers did consciously shape the circumstances that led to the strike and hoped it would materialize. They convinced Godot that a strike was necessary or inevitable.

Management's manipulation of the issue of the eight-hour day during the tense summer of 1899 invites suspicion of their motivations. Though the union's demand for eight hours with two charges of coal a day (instead of twelve hours with three charges) had been long-standing, leaders were not pushing the matter that summer. Other issues had come to the fore. It was the company that suddenly proposed to initiate a companywide experiment (itself an irregularity since trials had always occurred on a limited scale) with the shorter day. Moreover, the PGC offered workers a pay formula that not only reduced earnings by a third (to compensate for the lighter work load) but was complicated and sure to sow confusion. The superintendents reported that eight hours was an excellent issue with which to confound stokers, raise dissension between union leaders and members, and bring the spirit of revolt to a head.^[142] Gigot, probably the architect of the strategy noted in a report at the end of July that a strike could easily develop from the experiment because workers would not accept so large a reduction in pay. Without opposing the experiment for that reason, he predicted that there would be angry demands for an alteration in the pay scale and insisted that the company must resist with absolute firmness. Indeed, he ended the report on a note of hopeful tension: "Discipline has been destroyed in all the factories; we must at all costs [italics Gigot's] reestablish it. We must not hesitate. The continuation of the present situation is not possible. The occasion is excellent to return to our former authority."^[143] Two days later, Godot, writing from his vacation home in Brittany echoed Gigot's view: "We must at all costs [italics Go-dot's] finish with the indiscipline that reigns in the factories, and I find the moment opportune. . .. We can take the chance."^[144] At the very moment

[141] AP, V 8 O , no. 159, Godot to Monsieur Defrance, October 19, 1899; Journal du gaz, no. 162 (September 5, 1899): 1.

[142] AP, V 8 O , no. 159, report of Gigot to director, August 2, 1899; Journal du gaz, no. 148 (February 5, 1899): 2.

[143] AP, V 8 O , no. 151, "Essai de 8 heures (25 juillet 1899)."

[144] Ibid., no. 160, Godot to Gigot, July 27, 1899.

these memoranda were circulating within the firm, a union leader warned that foremen were doing their best to stir up the laborers.^[145] Thus the evidence establishes at the very least that executives implemented the experimental eight-hour day knowing that it could easily produce a strike and hoping that it would. The engineers were prepared to "take the chance." They apparently believed that such a strike would provide an opportunity to crush the union and return to an earlier mode of industrial relations.

Even if management did not intentionally foment the strike, their use of it to crush the union—knowing as they did the consequences—is most significant. After seven years of cold war with the organization, the engineers decided to abandon outward compliance with public opinion. They must have known that their refusal to rehire strikers put an end to any hope for a renewal of the charter in the prevailing political situation. Immediately after the strike was lost, newspapers like the Petite République called for rescinding the PGC's charter at once.^[146] Councilmen of the left regarded the company's behavior as the final outrage in a long line of impermissible acts. Crushing the strikers negated the impact of all the financial sacrifices the firm had made up to 1899. As individuals, the engineers also sacrificed the security of their careers after 1905, no doubt a weighty concern for them. Apparently though, the temporary recovery of industrial discipline was worth the price. There is no hint that engineers later regretted their hard line. They certainly did not soften it between the end of the strike and the disappearance of the PGC (in 1907) despite constant pressure to rehire the strikers. Seven years after the event, Euchène (now chief of the factory division of a PGC that was about to be liquidated) still described the strike in the most traumatic terms and still refused to rehire the "agitators" he had excluded from the plant in 1899.^[147]

The Politics of Mechanization

Corporate policies regarding the substitution of capital for labor amplify management's perspective on industrial relations.^[148] In particular, an inci-

[145] Journal du gaz , no. 159 (July 20, 1899): 2.

[146] For a sample of public opinion, see AP, V 8 O no. 159, "Gr<Ap232>ves de 1899— articles de journaux"; no. 160, La Petite République , August 16, 1899. On the policy of refusing to rehire strikers, see AP, V 8 0¹ , no. 160, "Grèves: Août 1899." Godot even refused to receive two aldermen who wished to discuss the issue. No. 160, Lucipia to Godot, January 19, 1900.

[147] Ibid., no. 150, "Affaire Plantin."

[148] For a general study of capital investments in gas plants, see Derek Matthews, "The Technological Transformation of the Late Nineteenth-Century Gas Industry," Journal of Economic History 47 (1987): 967-980. The conclusions, based on inferences, do not specifically examine how decisions were made, nor does the author consider the ties between industrial relations and technological change.

dent that occurred in 1881 reveals a great deal about how the PGC's officers made technological decisions and what sorts of results were paramount for them. Semiskilled workers in the retort-making section of the company's brickyards took advantage of the tight labor market by threatening to strike if they did not receive a raise of fifty centimes a day. Engineer Audouin castigated himself for not having installed retort-making machinery when labor agitation had surfaced three years earlier; the equipment was readily available and had been for some time. Audouin had estimated in 1878 a savings of seventeen thousand to eighteen thousand francs a year on labor, more than enough to justify the investment. This time the manager vowed not to miss the opportunity, even if the strike failed to materialize. He noted that economic considerations were only "secondary" in engendering his resolve. What made mechanization attractive was that "the use of this equipment would have the great advantage of assuring the almost absolute independence of the company over its qualified [spécial ] labor."^[149] Audouin's comportment in this incident suggests that the company was a laggard in adopting labor-saving technology. It also shows that economic rationality was not the central consideration. Managers turned to mechanization as a defensive measure to turn back challenges to its ascendancy over the personnel.

Initially, the gas industry was not heavily mechanized. Production depended on the flow of gases, not on the movement of mechanical parts. In 1878 the PGC required only eleven hundred horsepower to operate its machines.^[150] As the range of new tools and devices grew, engineers kept abreast of them on an experimental basis but introduced them into the production process only belatedly and only when they believed that there was little choice. In effect, the managers made labor, not market considerations, the final arbiter of decisions on mechanization. They provide corroborative evidence for Harry Braverman's justly controversial thesis that class conflict determines the level of technology.^[151]

[149] AP, V 8 O , no. 1067, "Briqueterie de la Villette (11 juin 1881)."

[150] Ibid., no. 716, report of Arson, July 15, 1877.

[151] Harry Braverman, Labor and Monopoly Capitalism:The Degradation of Work in the Twentieth Century (New York, 1974). Braverman's provocative thesis has attracted much commentary. Among the most telling are Richard Price, "Theories of Labor Process Formation," Journal of Social History 18 (1984): 91-107, and Craig Littler, The Development of the Labor Process in Capitalist Societies (London, 1982). In spite of the objections that have been raised to the Braverman thesis, this study suggests that it may hold in a general sense, at least for the authoritarian paternalists I have been describing.

When the ministerial commission of 1890, charged with reviewing the PGC's technological situation, characterized the company's policies as cautious but generally sound, the members did not seem to realize how antiquated were the methods for loading coal into retorts. This operation was done entirely by stokers with shovels and was a "skilled" operation. Yet gas companies all over Europe had long ago introduced a simple tool, the scoop (cuillère ), into stoking to curtail labor shortages and the corresponding market power that hand shoveling had given to workers. Though the scoop did not eliminate the need for uncommon strength and endurance, it did make filling the retort a good deal easier. Workers who could load the scoop and empty it into a retort were far more numerous and easier to train than those who could propel coal into the retorts with a shovel. Perhaps the ministerial commission did not note the PGC's backwardness because elsewhere the scoop had become so much a part of accepted operations that the members did not even inquire about its use.^[152]

Remarkably, the PGC put off introducing the scoop into its furnace rooms until 1890—no less than a quarter century after its counterpart in London. Engineers in Paris had long known about the tool and its advantages. Servier, the assistant factory chief in the 1860s, had inspected its use during a trip to London. No later than 1866, plant superintendents had experimented with it but had taken no further steps.^[153] The reason for retaining shovel work, despite the recognized way it limited the labor pool, was largely, but not entirely, a matter of inertia. All things being equal (which they were not), shoveling had some production advantages over the scoop. The latter tool required some minor alterations in the arrangement of the distillation rooms; it also wore down the walls of retorts so that they had to be replaced more frequently. Moreover, a proficient stoker could pack a retort more quickly and tightly with 150 kilograms of coal instead of 135 kilograms.^[154] Thus, shoveling fit into the efficient hot-roasting method somewhat better than the scoop. Ultimately, the reliance on stokers' shovels for so long reflected the PGC's distinctive operational

[152] Compte rendu des travaux de la commission nominée . . . le 31 janvier 1890, pp. 36-148. On the use of the scoop in Britain in the 1870s, see Eric Hobsbawm, "British Gas-Workers, 1873-1914," in Laboring Men: Studies in the History of Labor (London, 1964), pp. 158 -178.

[153] AP, V 8 O , no. 709, report of Servier to director, May 7, 1860; no. 671, deliberations of May 9, 1866.

[154] Ibid., no. 1068, report of Gigot to Arson, April 11, 1869.

situation. Since gas in Paris was expensive and the firm did not hasten to expand its clientele, demand grew steadily but not massively. This slow growth meant that the PGC had moments of labor shortage but not the chronic and desperate dearth of workers that might have made new production methods essential.^[155] Furthermore, the stokers, basically seasonal migrants, had so far remained unorganized and made only wage demands. Such demands might have displeased the engineers, but they did not frighten them, especially because productivity increases often covered the raises. Management wrung its hands over a wage increase, but raises did not spur it to action. Even when unit labor costs turned upward around 1876, the company was not ready for change, though it paid a price for the inertia. In 1860 stokers' wages in London had been about a third higher than those in Paris; by 1890 the situation was reversed.^[156]

Not high wages but rather the sense that management was about to lose the ability to dominate stokers finally brought an end to shoveling. In 1890 the stokers formed a union and launched a strike with a coherent set of demands. None of the stokers' earlier actions had been so well organized. The PGC, caught off guard, had to make concessions and recognize that a new era of industrial relations was at hand. This experience converted management's complacency into frenzied action: production managers achieved in a few weeks what they had not attempted in the previous twenty-five years. Scoops became a part of daily operations, and the change happened smoothly, without resistance from the stokers. Gigot, echoing Audouin nine years earlier, stated that the purpose of the change was "to assure production, not to economize."^[157]

The lag between the general use of a stoking machine, the next major laborsaving innovation, and the PGC's adoption of it was considerably shorter.^[158] The reason for the accelerating pace of change was the powerful

[155] Pressures generated by demand probably explain why the municipal gas works of Brussels, which charged much lower rates and had made efforts to expand the clientele, was so far ahead of the PGC in mechanization. See ibid., no. 1520, "Voyage en Belgique (décembre 1892)." Engineers Euchène and Louvet noted, with some surprise, that despite the low wages paid to workers, Belgian managers were aggressive in replacing manual labor with machinery.

[156] Ibid., no. 709, report of Servier to director, May 7, 1860; no. 1520, "Rapport de Monsieur Euchène: Mission en Angleterre du mois de mai 1893." The comparison is a delicate one, for patterns of stoking were rather different in the two companies, as Euchène's report makes clear. English stokers had just won the eight-hour day in 1889, but they worked more intensively than Parisian stokers.

[157] Ibid., no. 148, report of Gigot, December 16, 1890.

[158] There were technological advances in charging between the scoop and the mechanical charger, among them vertical and inclined retorts. They will not be discussed here because they did not play much of a role at the PGC. Arson examined the innovations and sometimes even arranged for limited experiments with them but did not apply them. See ibid., no. 711, "Rapport no. 39. Examen du projet des fours à cornues verticales"; no. 1062, "Substitution des cornues verticales (janvier 1876)."

threat of labor independence that emerged in the last decade of the century. Without yet having introduced the rudimentary scoop, the PGC's engineers had shown sporadic interest in the far more sophisticated stoking machine. This apparatus had the potential for eliminating skilled laborers entirely. The company, so hesitant in the application of innovations, was in fact at the forefront of research on a machine. By 1877 its engineers had an operational, if unperfected, stoking machine in the laboratory. The firm even took out a patent on one version.^[159] Had the PGC wished to pursue its breakthrough, the apparatus would have come just in time to stave off rapidly rising labor costs.

Though outside experts spoke of labor savings of 50 percent resulting from stoking machines, managers of the PGC expressed no interest in that aspect of the innovation in their reports. They viewed the machine, as Gigot did the scoop, as a means to discipline labor. In the absence of a permanent threat to their authority, the engineers did not hasten to use the machine. In addition, they worried about the reliability and durability of the apparatus under the difficult operating conditions in the distillation rooms: burly stokers might endure the intense heat and coal dust, but mechanical parts might not. These concerns, coupled with the lack of anxiety over labor up to 1890, postponed the application of the stoking machine.^[160]

Just as the strike of 1890 had promoted the use of the scoop, the strike of 1899, compelling managers to confront a breakdown of authority, decided them on mechanization at all costs. The "spirit of revolt" pushed aside their qualms about the stoking machine's reliability and seemed to leave them no choice but to rush it into use. In little more than a year after the strike, the stoking machine moved from the laboratory into the La Villette factory. The cataclysm of 1899 broke not only the managers' technological prudence but also their caution about capital expenditures. The

[159] Ibid., no. 779, "Chargement mécanique des cornues (13 mars 1877)" (fol. 254); no. 678, deliberations of September 5, 1877. On the benefits and disadvantages of the stoking machine, see C. E. Brackenbury, Modern Methods of Saving Labor in Gasworks (London, 1900).

[160] AP, V 8 O , no. 1520, "Voyage en Angleterre." Engineer Euchène was quite aware that labor costs were a good deal higher in his company than in the British ones, where stoking machines were used (1:35-43). Yet he did not show much interest in the machinery for that reason.

company spent nearly half a million francs on the mechanization of stoking in 1902 and 1903.^[161]

The timing of the investment was peculiar and raises pointed questions about the motivations of management. It came when the future of the PGC was hardly assured. Management planned the capital expenditure during the negotiations over the Chamon project with the Nationalist deputies. It executed the plans as the agreement faced an uncertain, and ultimately hostile, reception before Parliament. Thus, the decision to install stoking machines must be seen as a far more precipitous and emotional move than was customary for these cautious engineers. Mechanization was, arguably a form of revenge or at least a symbolic assertion of mastery that engineers used against the stokers. In any case, the application of laborsaving innovations at that precarious moment was hardly a business decision pure and simple.^[162]

The PGC's policies of technological brinksmanship clarify the social and psychological dimensions of labor management. Apparently, the engineers found wage concessions morally painful but not intolerable. What bothered them about the workers' quest for higher pay was not so much the cost to the firm as the challenge to hierarchical authority and to their right to calculate the appropriate compensation for their subordinates. Indeed, in some ways the officers of the PGC proved less sensitive to bottomline considerations regarding labor than their counterparts in other European capitals, as the divergent policies on the scoop demonstrate. Though Parisian engineers grudgingly accepted the need to spend more on workers, they refused to concede any degree of "independence." Perhaps authoritarian paternalists were incapable of making decisions about laborsaving machinery on economic grounds alone. They focused on power relations inherent in technologies and gave more priority to mastery over the shop floor than to extra profits.

<><><><><><><><><><><><>

Reinhard Bendix reminds us of the broader significance of managerial policy toward labor. He notes that it reflects "the terms on which a society undergoing industrialization will incorporate its newly recruited indus-

[161] For a description of the charging machine that the PGC developed and eventually licensed for use abroad, see H. Laurain, "Le Chargement des cornues à gaz par le chargeur Brouwer," Compte rendu du trentièrne congrès de la Société technique de l'industrie du gaz (1903), pp. 137-150. Management also turned its attention to the mechanization of another of the stokers' functions, the extinction of coke. See AP, V 8 O , no. 736, deliberations of December 26, 1902.

[162] The board of directors awarded hefty bonuses to the engineers who moved the stoking machinery so swiftly from the laboratory into production. AE V 8 O , no. 699, deliberations of July 7, 1904.

trial workers into the social and political community of the nation."^[163] That insight has disconcerting implications as far as the PGC and nineteenth-century France are concerned. Management hoped to consign its labor force to little more than complete subjugation, even though it could easily have afforded to be quite generous. The engineers resented the market forces that gave workers higher nominal wages. Though they were willing to spend more on the personnel when there were excellent political reasons for doing so, they drew the line at allowing workers to seek some control over their work. The managers, secure in their superior intellect and their own devotion to duty could not comprehend workers' wishes to escape from their arbitrary authority. Ultimately, moral considerations—and perhaps a spirit of vindictiveness—drove labor policies at the PGC. The situation was all the more problematical in that the last decade of the century inaugurated a brief era of democratic and republican extension of rights. The balance of political forces shifted just enough to give the common people some grounds for protesting their subordination and some hope for a redress of grievances. Managers tried to come to terms with the shift, but entirely on their own grounds. Their comportment during the strike of 1899 and its aftermath illustrates the depth of their self-destructive frustration that authoritarian paternalism could not prevail under the Radical Republic. The engineers sacrificed much, including the security of their careers, to defend that principle.

In the short run that defense was costly for the engineers. By 1910, however, they might well have claimed vindication. Just as they had given up on a strategy of concessions for fear that the situation had gone too far, so too did large sections of the political elite, only a few years after the PGC disappeared. Did not the crushing of the 1899 strike foreshadow the energetic repression of labor unrest by Georges Clemenceau and Aristide Briand?^[164] The managers of the PGC had always been, and remained, on the forefront of the defense of hierarchical authority.

Conclusion

The business policies of the PGC were not of one doth. Aggressive on small matters, management was usually hesitant on core business concerns. Executives were effective at cutting costs, but the endless search for savings on fuel, for example, was more intense than the concern to cut labor costs to the bone. The managers were impressive in exploiting mo-

[163] Reinhard Bendix, Work and Authority in Industry (Berkeley and Los Angeles, 1974), p. 441.

[164] Jacques Julliard, Clemenceau, briseur de grèves (Paris, 1965).

nopolistic markets for by-products but fell down on doing so for gas. In the end, the decision makers were selective about the market factors to which they responded. The PGC reinforces the impression of French business as conservative, cautious, and risk-avoiding—and not just because the firm functioned with a comfortable monopoly. Management made adjustments to market forces when it had to, but it did so timidly. It treated mass marketing, even though it entailed relatively few risks and promised high returns, as a source of problems, not of opportunities.

The special value of studying the PGC resides in the opportunity to observe the direction that an emerging elite of salaried managers imparted to enterprise. The outstanding point is how little the new category of businessmen innovated in terms of style of leadership and guiding assumptions, although the elite was closer in background and inspiration to high civil servants than to the traditional patronat. Years of rigorous mental training did not help engineer-managers clarify their own uncertainty about market forces. On the one hand, they regarded the pursuit of self-interest as the bedrock of human motivation. That is why they made profit-sharing bonuses so important a part of the compensation package and insisted (as we shall see) on promoting subordinates on the basis of merit alone. On the other hand, they rejected the legitimacy of market forces in setting wages and commodity prices. Managers did nothing to modernize a traditional ambivalence that pervaded French society about the morality of the marketplace.^[165]

The attitude of salaried managers toward corporate assets was most certainly proprietary. They easily put behind them the ethos of state service when they joined a private firm. No duty surpassed that of making money for stockholders, and they did not stop short of morally dubious activities in pursuit of the goal. Thus, the rise of a new type of business management did not create a different sort of agenda for French enterprise.

The decision makers at the PGC threw away a chance to innovate in industrial relations even though that opportunity could have been rather painless. Their objectives were those of the patriarchical industrialist, to preserve autocratic paternalism.^[166] The engineers' decisions on laborsav-

[165] It was not simply the case that management rejected market forces as the mechanism for setting the prices that the company charged. Inflation of the prices that suppliers charged the PGC engendered in managers a conspiratorial mentality. Only partially did they conceive of the economy as subject to impersonal forces. See AP, V 8 O , no. 845, report of Le Maire to director, February 24, 1873.

[166] On traditional mill owners, see Jean Lambert-Dansette, Origines et évolution d'une bourgeoisie: Quelques familles du patronat textile de Lille-Armentières (1789-1914) (Lille, 1954); David Landes, "French Business and the Businessman: A Social and Cultural Analysis," in Modern France , ed. E. M. Earle (Princeton, 1951), pp. 334-353.

ing machinery showed them to be authoritarian paternalists first and calculating businessmen second. They made a mockery of Emile Cheysson's hope that engineers could be impartial and humane experts mediating between capital and labor. No more than the dominating factory owner were the salaried managers capable of understanding the workers' situation.

In all these ways salaried managers at the PGC left business practices and hierarchical relations more or less as they found them. They administered the PGC through a distinctive cultural filter that they shared with the traditional patronat. Their presuppositions often proved problematical. Had management responded mechanically to impersonal market forces or simply pursued pragmatic political calculations, the history of the PGC might have been different—and longer. Their presuppositions proved equally unhelpful in dealing with another subordinate group created by economic development, the white-collar employees.