The Legitimation of Exclusion
One of the most important factors contributing to the limitation of women’s opportunities in the General Post Office and the Great Western Railway was the sexism of male managers. Employers’ belief in innate female inferiority and their preference for interacting with male employees led to women’s exclusion from offices that were non-clerical-labor-intense. Sexism may have contributed as well to the confinement of women to low-status jobs. Excluding women from jobs with long learning curves was consistent with synthetic turnover and the economic need to create circulation among the staff. However, such a policy was also consistent with the interest of male managers in patriarchy and, in particular, the preferential allocation of promotions to men.1
However important this concern with sex privilege may have been, it was a dubious basis on which to justify a business decision. All major personnel policies at some point had to be defended before the superior. Department heads had to submit staff budgets to the Secretary’s Office. The Secretary had to report to the board of trustees, if the firm was private, or to the Treasury Department, if it was public. Even decisions that were made on the most personal, idiosyncratic grounds had to be rationalized with arguments that used good business sense or incorporated sound public policy. In private firms, the former rather than the latter were especially important. Given that men are more expensive than women, the exclusion of women always had to legitimated to local referees in terms of a larger organizational policy.
The significance of such ideologies has to be placed in proper perspective. A good explanation was neither necessary nor sufficient to exclude women from the job. People are intellectually very capable of bending the terms of a given set of beliefs to justify the policy they want to pursue. Max Weber was able to take the anti-materialism of early protestantism and twist it into a justification for capitalism (Weber 1958). “Equal pay for equal work,” which sounds like a reasonably feminist position, has been used by male unionists to legitimate paying women less than men (Kessler-Harris 1982).
Nevertheless, a manager could significantly contribute to the maintenance of an all-male labor force by developing an ideological defense of such a policy that could withstand the scrutiny of his immediate superiors. A sexist manager whose covering explanation was weak could expect constant challenge from budgetary watchdogs. The use of males would require extended confrontations with his colleagues and the expenditure of a substantial amount of time and energy. A well-accepted justification could be the basis of permanent organizational policy; once the principle was established, a precedent could be set that would allow for the easy exclusion of women from comparable positions in the future.
Not all such arguments are intellectually interesting. In the setting of the sex-type of any particular job, managers are likely to make a fairly large number of arguments, each of which may contribute in some way to the final decision that gets made. Those ideological statements that are particularistic and case-specific will have little predictive power in a general discussion of sex-typing. Within the Great Western Railway and the General Post Office, however, there were a number of principles that became important general guides to policy. These principles became accepted because they were consistent with the objective economic requirements of each firm; once instituted, however, they attained a life of their own. They introduced minor rigidities that kept women out of job assignments for which they would have been economically appropriate. They also provided a cover with which managers could obscure large-scale de-feminizations and rationalize the preferential allocation of rewards to men. The economic considerations motivating the original policies by themselves would have justified the exclusion of women from some attractive forms of employment. However, these programs were implemented in a fashion that was blatantly discriminatory against women. Furthermore, management justified these policies both with economic considerations and with bald, forthright discussions of the inferiority of women. Thus, for many managers the appeal of these policies was that they provided a legitimate, defensible cover for the extension of male privilege.
Women and Nightwork
One of the most important of these policies was the prohibition of women from working at night. In both the Great Western Railway and the General Post Office, women were generally only permitted to work shifts that started no earlier than seven in the morning and ended no later than nine at night. Related to this ban on nightwork was the rule that only men were allowed to work on Sundays. Thus, all positions that required irregular attendances were given by default to male employees.
The importance of the nightwork ban in limiting female employment cannot be overestimated. Most of the work in sorting mail occurred between three and seven in the morning, when sorters prepared mailbags for morning trains. Depot clerks often had to work at comparable hours preparing the invoicing and accounts for early departures. These early-morning duties were one of the fundamental reasons that women were barred from sorting offices and railway depots. Nightwork was also important in telegraph work. A substantial proportion of telegraph transmission was done at night. News stories would be transmitted late at night for use in the morning papers; non-urgent messages were also sent at this time, when the lower volume of traffic on the wires justified charging a lower rate. Women were traditionally given the daytime duties in signalling offices, while night duties were given exclusively to men.
Sorting offices, depots, and telegraph offices were the most important cases of women’s employment being limited by paternalistic restrictions on hours. Most other offices worked on a nine-to-five schedule and had limited requirements for their staff to stay late. It was true that during annual peaks, or when offices were understaffed, overtime could be required. In these cases, men had to remain at work while women went home. However, most clerical offices never generated enough overtime to require a night staff, thus preventing the rise of a set of positions from which women would be prohibited.
To some extent, the banning of women from nightwork was voluntary on the part of the firms. They were not forced to adopt their hours policy because women were unwilling to accept night positions, nor were they under legal compulsion to make such restrictions. Women themselves would have been entirely willing to accept clerical employment at night or on weekends. This was actually demonstrated during World War I, when the shortage of male labor forced the Post Office to use women for all duties, including those at night. The Post Office had been reluctant to have women sort at all. After it became necessary to bring in women temporary sorters, they still attempted to give men the night duties. However, complaints from the men and the growing labor shortage finally required that women be placed on non-daylight shifts. Originally, women were put on a three P.M. to midnight duty. This proved to be extremely unpopular with the women because they were forced to go home in the middle of the night on dark, unsafe streets. As an alternative, the women were put on a shift from nine P.M. to six A.M. This solved the transit problem, since women could now take the late trams to work every evening and return home during daylight on the early-morning trains. The commuting was made more pleasant by their going in the opposite direction from rush-hour traffic on both halves of their daily route. Furthermore, the new shift was compatible with the women’s family responsibilities. The three-to-midnight shift kept women away from home during the dinner hour. The nine-to-six shift allowed women to be home during the day, prepare dinner for their families, and see that the children were in bed before reporting to work. This new shift was acceptable to the women and was generally adopted during the war whenever female nightwork was required (WWI Nightwork 1918).
Many employers were legally forbidden to use women at night. Britain had an extensive body of protectionist legislation that put limits on the situations in which an employer could hire women or juveniles. There were restrictions on the kinds of industries in which women could be employed and there were limits on the hours that women could work. Nevertheless, none of these laws applied to clerical positions on the Great Western Railway, and they had no relevance whatsoever to the General Post Office. Office employment was exempt from most protectionist legislation. Whatever laws did exist applied to factories, and specifically to those parts of the factory where manufacturing actually took place. An office within a factory would have been exempt both from the hours limitations of the Factory Acts and from the health and sanitary provisions as well. The Railway Clerks Association lobbied in the 1920’s to extend the Factory Acts to railway offices, but never received even modest support (Walkden 1928). The Post Office would have been exempt from protectionist legislation in any case, because they were an agency of the crown. Factory laws were limited in their application to the private sector.2
However, this said, it should be noted that the Post Office often voluntarily followed policies that were mandated by law for other industries. In order to facilitate the enforcement of labor legislation, the civil service usually attempted to run the government by the same laws that it was advocating to industry as sound public policy. Thus, when the government began to encourage the use of industry-wide collective bargaining, it set up formal negotiating machinery for its own employees. When, at the turn of the century, national employment policy consisted of discouraging employers from hiring boys to dead-end positions that could create later unemployment problems, the government voluntarily reduced its own reliance on juvenile labor, even though this entailed substantial costs. The Post Office’s tendency to act as a model employer may have carried over into its nightwork policies. Although there may have been external political pressures for the Post Office to abstain from nightwork, there would have been few formal legal barriers. However, the Post Office’s willingness to cooperate with national policy was probably facilitated by some internal considerations. Not only did the nightwork ban appeal to the local patriarchical concerns of Post Office management, but it had economic advantages as well.
The restriction was made economically viable by the side benefits that came from imposing large amounts of nightwork on men. Late shifts were generally unpopular among the men. Most night shifts were staffed with men who had been assigned to such rotations against their will, since the Post Office could rarely find men who would take such hours voluntarily (Telegraph Substitution 1880; Hobhouse 1906). Duncan Gallie has a perceptive discussion of why in general workers do not like working late shifts. Nightwork seriously disrupts people’s personal lives. Early evening is when most people have their social lives. Nightwork not only restricts workers’ ability to see their friends, but keeps them from attending meetings or cultural events. Late shifts also provided a source of tension for family life. Late hours keep the worker from interacting with his or her family; futhermore, the presence of a day sleeper inhibits the patterns of living of those family members who are home during the day and have to remain quiet. Most important, the irregularity of sleep and meals that come from an off-cycle schedule can be seriously detrimental to a worker’s health. The change in eating habits can have as much of an impact as sleep loss in producing physical problems, exhaustion, and psychological distress (Gallie 1978). In the GWR and GPO, both men and women avoided late hours whenever possible; there are few cases of women’s volunteering to work evening hours. Women were no more incapable of working late than men were, but both sexes preferred daytime employment.
Because of the unpopularity of nightwork, it was against Post Office policy to assign anyone to such a position permanently. The Telegraph Department explicitly kept a set of permanent day positions open as promotions for those workers who had worked disproportionately at night. The special reserve of day jobs was the linchpin of the economic strategy that made the scheduling of nightwork a tool for sex discrimination. If only men could do nightwork, and there was a reserve of day jobs left over for nightworkers, then this day reserve would also be all-male. The same logic applied for the supervisory force. Nightworkers would need superintendents. However, the night supervisory positions were the least attractive of the better positions. Thus, there would have to be a reserve of supervisory day positions left open to provide relief for the night supervisors (Hobhouse 1906).
The night reserve system had important economic advantages as well. The Post Office incentive system was based on annual salary increases coupled with the prospects of promotion. However, the number of superior positions was limited by economic constraints; not every male worker could expect regular promotion throughout his career. Since the supply of both annual pay increments and promotions was financially limited, the telegraph service needed a scarce resource that could be distributed preferentially and that would not cost the Post Office any money. Daytime positions were such a resource, so the Post Office began to use the night/day distinction as an inexpensive way to provide mobility for the men.
By 1880, the outlines of the shift incentive system were in place. There was no permanent night shift per se. Instead, all entering workers were subject to irregular mandatory night duty. The night jobs would be first filled by volunteers; then, after that, the supervisor would assign workers temporarily to the late shift until all posts were filled. All workers were subject to being tapped for late-night duty until they could get promoted or transferred to an all-day position. Some skilled jobs were only performed by day, making promotion to such a position a double benefit. Likewise, some offices were only open by day, making a transfer to such a site automatic protection. If a worker could not get one of these positions, he had to wait until a job in the day reserve opened up and then he would be guaranteed freedom from the night draft. Note that both the lateral transfers and the day reserve represented prizes that a worker could compete for, which cost the Post Office nothing to distribute.
This explains why women were barred from nightwork in the telegraph department. The development of long-term incentives was only an issue for permanent staff. Women were temporary staff who were expected to leave upon marriage. The Post Office wanted to restrict the improvements a woman would receive in her career so that as many “promotions” as possible would be concentrated among the men. There are two regimes that would have been consistent with such a strategy. All women could have been put permanently on nightwork, along with a minority of men. The minority of men would have then competed for a daylight assignment. The Post Office chose the other alternative, putting all women on daylight assignments and making all men work to obtain a daytime job. Because women’s hours were constant, they received no improvements over their career, excluding them from the Post Office’s incentive system.3 Not only was the entire population of men thus put into a competitive situation, but the value of a daytime job became inflated. Removing women from the draft pool meant that the odds of any individual man’s having to do nightwork were increased, just as the supply of daytime jobs was artificially restricted. This increased the pressure felt by men to compete for a daytime slot and thus further bound the men to the postal incentive system. Thus the linking of nightwork to gender was intrinsically related to the policy of synthetic turnover that divided workers by the need to provide them with long-term incentives.
There were other points of convergence between the nightwork ban and the operation of the postal internal labor market. In the Telegraph Service, the confinement of women to low-status jobs, which was an essential component of the synthetic turnover regime, was originally instituted as a by-product of the creation of the male day-shift reserve. The vigorous implementation of the nightwork ban was facilitated by the fact that it provided a second justification for the artificial encouragement of early female retirement and the elimination of women from complex, highly paid positions.
In the early days of feminization of telegraphs, women were given a number of duties that were far more responsible and complex than those they would receive in later periods. This was due to the idiosyncrasies of the market for telegraph operators in the early 1870’s, when women were first introduced to telegraph work. The management of the Post Office committed itself early to the principle of a female signalling force. There were many factors that contributed to heavy early feminization, such as a concern with reducing labor militancy and an overall need to economize on personnel due to the high labor intensity of hand-transmitting messages. Once the telegraph force had been made primarily female, however, it was difficult for the Post Office to confine women strictly to the low-status positions. In the late 1860’s and early 1870’s, competent Morse code operators were extremely scarce. Telegraph signalling is a highly skilled occupation that requires years of training. A signaller needs at least a year to learn the rudiments of Morse code and then several years after that to become proficient. The Post Office could not train its labor force gradually because, at the time of nationalization, the telegraph service was rapidly expanding, creating an enormous number of vacancies. This created a very large number of jobs that had to be filled by whomever was available to fill them. Under these circumstances, the Post Office could not afford to be too particular about which jobs were open or closed to women. Women made up 60 percent of the learners in any given year (GPO Establishment Books 1876). There were so few operators of any sex that could proficiently work the more difficult lines that the Post Office assigned these positions to anyone who could do the job.
One important implication of this was that many women found themselves operating long-distance inter-city lines. In general, the difficulty of operating a telegraph line correlated with its length. Local lines within a city were relatively easy to operate, because the connections on these wires usually produced a clear signal and the transmitters used for local messages were not particularly complex. Inter-city connections were both weaker and more prone to breaking down, requiring subtlety in one’s fingering technique and the ability to do quick repairs. The long-distance transmitters were substantially more complex than a simple Morse device and varied significantly between one manufacturer and another. In the 1870’s, women worked longdistance lines alongside the men and handled these duties with great proficiency. In 1876, in the Central Telegraph Station, the men raced the women sending out the Queen’s opening speech to Parliament. The detailed results of the test are not known, but in the aggregate the women significantly outscored the men. According to the Controller of the Telegraph Service, in normal service, under non-competitive conditions, the performance of male and female signallers was roughly equal (Telegraph Report 1876).
This came to an end in the late 1870’s. By the 1880’s, the telegraph service was strictly segregated into male offices and female offices. Local branch offices were worked by individuals of a single sex. In the Central Telegraph Station, women were removed from the inter-city galleries and forced to work in an all-female Metropolitan gallery that only did local signalling. The primary reason for this transformation of telegraph work was the introduction of the reserved male day shift. Women had been excluded from night duty since the beginning of feminization. By 1875, however, the ban on women’s doing nightwork was creating real difficulties for the staffing of the late shifts. The ratio of women to men at the Central Telegraph Station was at an all-time high, possibly due to an overcommitment by management to female labor. The late duties were falling on an ever-decreasing supply of men. Since the same individuals were consistently being selected for involuntary late shifts, morale became extremely bad. Telegraph management realized that in order to create an acceptably large pool of candidates for nightwork, the sex ratio of the telegraph service would have to become more male. The need for males became intensified by the creation of the male day-shift reserve, which was first developed and implemented during this period.
The Post Office was then faced with the problem of how to adjust the ratio of males to females to comply with the new nightwork-related requirements. Some of the adjustment in the sex ratio could be accomplished through the natural growth of the Post Office. The new vacancies created by expansion could be given to men, producing a painless strategy of de-feminization. The Post Office took the maximum advantage of this kind of tactic by closing its training program to women telegraphists and refusing to make any more female hires. However, the supply of male entrants was insufficient to produce a proportion of men that would be large enough to support both the night shift and the day reserve. Some of the change in the sex ratio would have to come from established female employees leaving the service. So in the mid-1870’s the Telegraph Service began a campaign to encourage women signallers to resign from postal employment (Telegraph Substitution 1880).
The first step was the imposition of a marriage bar. The Telegraph Service had not used a marriage bar in the early 1870’s, since the general shortage of qualified operators motivated management to encourage women to remain on the job (Scudamore 1871). By 1876, however, a bar was in place to encourage a high rate of female exits. Postal officials noted, however, that even with a marriage bar, women were not resigning fast enough to bring in the required number of men. To further stimulate the quit rate, the Post Office froze all promotions of female staff. It was decided that until the sex ratio of the service was brought to an acceptable level, all supervisory vacancies were to be filled with men. In the same spirit, the women were taken out of the Provincial Gallery, where the work was complex and technically stimulating, and put in the Metropolitan Gallery, where they could operate the simpler machines. Women were also taken out of branch offices where nightwork was performed and put in all-female branch offices that only operated by day.
Given the enormous scope and scale of the personnel readjustments that were being made, it is unlikely that the nightwork ban alone can totally account for these changes. The need to reduce the percentage of women in the Telegraph Service was relatively short term; however, the redistribution of women that resulted was a permanent transformation of Post Office practice. Additional support for these policies may have come from particularly sexist male managers who would approve of the limitation of female opportunity purely for patriarchical reason. However, what made these transitions economically viable was the fact that those policies that in the short term would help to produce an all-male reserve day shift would in the long term help contribute to the synthetic turnover regime. By the time that the staff adjustments required by nightwork were complete, the Post Office could reasonably expect to be suffering from the problem of an overly mature labor force on tenure-based salary scales. The solution to both the nightwork and salary-scale problems were the same: encourage women to leave, and allocate women to low-status positions to minimize the human capital loss entailed with these organizationally induced exits. The complementarity of nightwork policy with other important organizational concerns contributed both to the implementation of the staff adjustments of the 1870’s and to the long-term viability of nightwork itself as an accepted organizational principle.4
Segregation by Sex
There was a second process equally important to the limitation of female employment, the sex segregation of offices. Both the Great Western Railway and the General Post Office maintained separate promotion chains and salary schedules for their male and female employees. The General Post Office went one step further and wherever possible kept men and women in different rooms or buildings. Local London telegraph offices were either all male or all female; if an office was worked by both sexes, the women worked by day and the men worked by night. Money order work was arbitrarily divided into two sections, money order and postal order. The processing of the two orders was virtually identical, except money orders had a lower limit on the size of a transaction. Throughout the nineteenth century, money orders were processed by an all-male staff, and postal orders were processed by an all-female staff. The postal order women were further isolated by being placed with other women in the Receiver and Accountant General’s Office. The sections were merged in 1906, although this was combined with staff changes that made the new division over 90 percent female. The Receiver and Accountant General’s Office was further divided into two divisions, the main section and the telegraph accounts section. The first was all-male and the second was all female, although this distinction was relaxed after World War I. The segregation of the sexes even applied to temporary workers. During World War I, when temporary women sorters were introduced to the London sorting office, they were confined to an all-female floor, while men were restricted to the rest of the building. Sex segregation was also practiced by the Great Western Railway, but in a far more limited fashion. The need for isolation of the sexes dominated the early discussions of feminizing, and sex segregation was implemented up through World War I. After World War I, however, the practice was abandoned, and male and female employees were mixed freely.
Sex segregation had a mixed impact on women’s economic opportunities; it reduced the absolute number of jobs for women, but it upgraded the quality of those positions that remained. Sex segregation produced rigidities in the determination of occupational sex-typing by making the sex-type of a job dependent on the aggregate properties of an office rather than on the particular qualities of the job itself. The operation of segregation can be seen most clearly with a simplifying example. Assume that the only determinant of occupational sex-typing was the status of the position. If there were no sex segregation of offices, all high-status jobs would go to men and all low-status jobs would go to women. Under a regime of sex segregation, however, the sex composition of the office will be determined by the average properties of all jobs within the office. Thus, departments with a high proportion of supervisory positions will be male, and those with a low proportion of such positions will be female. However, jobs with atypical status will be misallocated in both kinds of departments; high-status jobs in the low-status offices will go to women, and low-status jobs in the high-status offices will go to men. Most offices have more unskilled positions than skilled ones. Therefore, a trade of skilled for unskilled positions leaves the side receiving the unskilled positions with a larger number of jobs. Segregation ends up defeminizing a large number of female jobs and feminizing a small number of formerly male jobs. This reduces the total amount of female employment, but improves the quality of those jobs that remain.5
Sex segregation was by itself substantially less important than many of the other factors affecting sex-typing. It involved the misallocation of a minority of jobs in departments whose majority was correctly sex-typed in terms of its economic attributes. Furthermore, serious misallocations could be corrected by the physical relocation of personnel. In the male Money Order Department, much of the filing and sorting was done by women. Since technically they were not supposed to be part of the Money Order Department, which was by definition male, they were hired as part of the female staff of the Savings Bank. Part of the explicit job duties of female sorters in the Savings Bank was to do sorting and filing in other departments, sparing these departments from having to hire their own female staff (GPO Establishment Books 1901).
Sex segregation was a comparatively transitory device for the exclusion of women. Sex segregation had been originally justifiable because Victorian managers feared that the mingling of the sexes would lead to a weakening of work discipline, as normally productive male workers gave in to the temptation of flirting, courting, and making casual conversation. Such fears seem groundless now and were probably groundless then. Workers in all-male environments lose time in stag conversations; introducing women would have changed the content but not the amount of the non-work-related fraternizing. Thus, to the extent that sex segregation represented a sincere concern by management with the deleterious effects of mixing, such concerns were likely to be allayed by actual experience with a coed labor force.
Furthermore, managers who attempted rigid sex segregation found this led to a wide variety of practical difficulties. The isolation of one sex from the other was often allowed to lapse through benign neglect. A particularly short-lived case of sex segregation occurred in the Foreign Office of the British civil service. The Foreign Office hired a small number of female typists, whom they sought to isolate from the male clerical staff, fearing that, if men began to socialize with the new typists, there would be an uncontrollable disruption of normal work activities. Therefore, they placed the women in an isolated room in the attic, far from the normal flow of Foreign Office activities. Male clerks were strictly forbidden from entering the attic, and the women’s presence was never officially announced except for an arrow on the attic staircase that said “To the Typewriters.” Obviously, under such circumstances, the female typists could not be kept a secret for long. Matters came to a head the first time that the women were to be paid. It was strictly inappropriate for men and women to fraternize in the corridors. However, it was equally inappropriate to give a worker’s check to anyone other than the worker himself. Thus, if the women were to be paid, they would have to report during working hours to the Cashier’s Office, which was located on the ground floor. Somehow, the women were going to have to get from the attic to the ground floor and back again without anyone either seeing them or having a chance to talk with them. So on payday, when the women’s turn came to receive their checks, all the corridors of the building were evacuated. The doors of all the offices were shut. Then the women were sent running down the stairs as fast as possible to get their checks and sent running back up to the attic again. As soon as they were safely concealed, the corridors were reopened and business was allowed to continue as before. Obviously, such a disruptive and cumbersome procedure could not have been made part of the weekly routine; management quickly perceived the difficulty of maintaining total segregation and moved the typists downstairs with the rest of the staff (Martindale 1938).
A comparable situation occurred in the Secretary’s office of the Great Western Railway. Most of the Great Western Railway was unsegregated. However, the supervisor of the Registration section of the Secretary’s office preferred a segregated staff. The Registration Office was responsible for servicing the GWR’s obligations to its stockholders; it sent out dividend checks every quarter and kept records of the ownership of individual shares. The office occupied three rooms, two of which were all male and the third of which was all female. To justify this physical division of labor, all women were given the job title “addressograph and duplicating machine operator."6 The men were given a wide variety of job titles reflecting the diversity of tasks within the office. Thus, the women could be relegated to the room with the addressograph machine and kept apart from the male employees.
However, the separation of the sexes quickly broke down. First, for the women to make addressograph plates, they needed to know the addresses of the stockholders; these addresses were kept in the main files located in the mens’ offices. The regular passage of women to and from the files provided ample opportunities for conversation. Second, it was impractical to confine women merely to addressograph work. Every three months, dividend checks had to be distributed, producing regular peaks in the workload of the office. During these peak periods, it was necessary to assign women to check preparation in order to relieve some of the pressure on the men. Since both men and women were working on the same tasks, it was impossible to coordinate the work without a substantial amount of interaction between the sexes. Thus, in practice, there was neither a rigorous isolation of the sexes nor an ironclad division of labor between men and women. It is easy to see how as a practical matter sex segregation would make a poor basis for the legitimation of all-male offices (GWR Women 1906).
The Great Western Railway followed the practice of most modern corporations in freely intermingling the sexes. The Registration Office maintained the fiction of hiring only men for two rooms and only women for the third, but in this regard it was an idiosyncratic exception. However, the Post Office maintained fairly rigid sex segregation up through World War II. The persistence of sex segregation in the Post Office is not easy to explain, but several factors may have contributed to it.
The first consideration is that the Post Office grew by the steady creation of new departments and divisions, as opposed to simple growth in numbers or growth from the absorption of pre-existing units. The creation of new units meant that there were always a large number of vacancies that could be arbitrarily filled with members of either sex; in the case of simple growth or growth through absorption, although vacancies are created, the units in which these new positions occur are already partially filled with pre-existing job holders. The presence of this earlier cohort of workers produces a pre-existing set of occupational sex-types that may not be consistent with an economic sex-segregated regime. It may be theoretically possible, given a particular distribution of jobs in a department, to create an arbitrary division of some offices as male and others as female that would produce an optimal overall sex ratio for the entire department. If a given department should be 40 percent male, it might be possible to arrange a set of single-sex offices so that 40 percent of the total staff is male, with no individual offices suffering from a grossly incorrect sex-type. However, the number of possible administrative solutions to such a problem will be small. Pre-existing staff in the wrong locations can add to the cost of setting up a segregated regime, as personnel have to be reallocated and retrained in new positions. The continual creation of new organizational units in the Post Office, such as the Telegraph Accounts Branch, the Postal Order section, the Telephone sections, and an immense number of local telegraph offices, provided the staffing flexibility to create a viable segregated regime.
Second, because the Post Office had been feminized during the nineteenth century, the labor shortages of World War I had a negligible impact on local patterns of occupational sex-typing. World War I had a far more disruptive impact on offices that in 1914 were predominantly male. Military enlistments created large numbers of vacancies that were randomly distributed throughout the male clerical ranks. Female replacements were required in virtually every office of most firms; since women were being brought in as short-term temporary help, they were assigned to the simpler jobs within each office. Thus, in a firm like the GWR, where World War I produced the first large-scale feminization of clerical work, women would have been distributed among low-status positions within every office and department within the organization. The widespread dispersion of women among departments would have eliminated any immediate prospects of creating segregation. The crisis atmosphere of maintaining smooth operations during the war would have reduced the prospects of developing a long-term segregation plan before 1919; after 1919, the post-war adjustments to the new economic environment would have pushed segregation far down the list of managerial priorities. In the years between 1919 and 1921, the Great Western Railway management was dealing with a series of crippling strikes by their manual laborers and an impending nationalization threat from the government. Many other industries were dealing with comparable changes in labor relations and government regulation. By 1923, when the business climate stabilized, the post-war gender regime would have been established for four years, making segregation an expensive and time-consuming reorganization.
In contrast, the Post Office introduced women through a series of planned innovations in which individual offices and departments were feminized one at a time. The staffing decisions were not made in a period of general organizational crisis, but instead under conditions in which the staffing of each individual department could be made the single preoccupying concern of management. Furthermore, there was no exogenous force scattering women throughout the organization, but instead the limited introduction of female clerks into those positions that were designated as being suitable for women by management. By feminizing in a gradual fashion, postal management was able to plan its sex-typing. By not having war, or past policy, predetermine various job allocations, the Post Office had the flexibility to develop an economic segregated regime. Thus, by being an early feminizer experiencing growth through diversification, management could custom-tailor their sex decisions to fit their own patriarchical needs.
However, it is one thing to explain how it was that the Post Office had the ability to create segregation; it is another to explain why they bothered to maintain such a system. In the early years, segregation seems to have been motivated by a genuine concern with work discipline and propriety. Nevertheless, given the impracticalities of maintaining segregated systems, it is hard to see why the Post Office persisted in its policies. It would have been administratively simpler and more cost-efficient to let supervisors in every office hire both male and female employees and put each sex in those positions that were appropriate for each sex.
The persistence of segregation may have been due to the unwillingness of postal management to handle the administrative difficulties that would have been associated with mixing pre-existing single-sex office staffs. Dorothy Evans, a civil service administrator of the 1930’s, compared the segregation policies of older and newer divisions of the British civil service. She found the older departments to be much more rigidly segregated, with the oldest department, the Post Office, being the most divided of all. To some extent, this might be explainable by older departments’ being run by older and more conservative administrators. However, Evans noted that there was a second problem; those departments that had been founded early had tended to feminize early and developed several cohorts of mature, experienced women. Some of these women had been promoted to supervisory positions and within the women’s sections exercised considerable authority and responsibility. The integration of these women into a male-dominated middle management force would not have been easy. The fairest solution would have been to integrate the male and female promotion chains completely and let the most senior supervisors take the combined responsibilities. In this case, however, many women would have outranked their male competition, which would have been intolerable to both middle and top postal management. A second alternative would have been to let men supplant the women supervisors, de facto demoting these women back to common workers. This would have been an extremely awkward situation, because these men would have had no experience in the work of the women’s divisions. The men would have had to have been trained by the very women whose jobs and status they were usurping. This would not have been an insurmountable organizational problem, since firms often depose supervisors when a change of leadership is needed. Given that the women were probably perfectly adequate supervisors, however, one can see how the Post Office might have been loath to engage in the large-scale disruption that would have been required to transfer the superior female positions back to men (Evans 1934).
For whatever reasons segregation survived, it nevertheless produced a small but real impact on women’s economic opportunities. It created a microscopic number of advanced positions for women that would not have existed otherwise. In 1931,7.7 percent of women in the Post Office held non-entry-level jobs. This figure was only 4.7 percent in the unsegregated Great Western Railway. These figures have to be treated with caution. There may have been technical differences between the two firms in the quality of work that could account for these status differentials. The differences could also be due to the different metrics used by the organizations to measure entry level. There is a suggestion, however, although not by any means a strong confirmation, that segregation benefited women by opening up some positions of autonomy.
The study of particular patriarchical constraints on female employment can provide some interesting insights into the process of ocupational sex-typing. Nevertheless, the observer should be wary of assigning any particular mechanism a central role in the overall explanation of the use of women. Nightwork and segregation played a part, but their impact was small relative to such purely economic factors as clerical labor intensity or the dynamics of synthetic turnover. The impact of the nightwork ban was quantitatively tested by including a dummy variable representing the presence of nightwork in the postal equations that were estimated in Chapter 2. There was the expected zero-order relation between nightwork and sex type; women were more likely to appear in those offices that only operated by day. However, when clerical labor intensity, the percentage of nonpolitical jobs, and the percentage of non-entry-level jobs were controlled, the coefficient for nightwork invariably became insignificant. Such a statistical finding is not surprising given the actual distribution of women in the Post Office. Although women were excluded from the sorting offices, which were often open twenty-four hours a day, the telegraph offices contained the heaviest concentrations of women in the Post Office. Telegraph signalling was clerical-labor-intense, and the telegraph offices had no carry-over employees from the pre-1871 era. Both economic factors caused the Telegraph Service to hire massive numbers of women in the 1870’s. The adjustments to allow for the demands of nightwork were trivial in comparison.7
Predicting occupational sex type from particular cultural or patriarchical ideologies of management is not easy. This is because it is fairly hard to specify a priori precisely what non-economic utilities a management will attempt to maximize in developing its gender policies. Male managers like an all-male labor force, if they have one already. They are usually in favor of confining economic opportunities to men, a value that is partially economic but partially cultural. However, individual managers can often differ dramatically in what other gender-related goals they would like to attain. Postal managers were very concerned about restricting conversation between the sexes; this has hardly been a universal concern of capitalists.
An example of the diversity of non-economic sex utilities can be found in the case of sexual harassment. Some employers have taken sexual advantage of their position, either by discriminating on the basis of physical attractiveness or by actually making personal demands on their female workers. Other employers would find such behavior morally indefensible, even as they discriminate against women in other ways. Managers make a free choice between abusive privilege and conformity with accepted principles of proper behavior. Both companies and individuals vary dramatically in which of these two selections they are likely to make.
Cultural considerations can have general predictive power when they are imposed by an external force on the labor market as a whole. The legal prohibition against women’s working at night in blue-collar occupations is an example of a cultural restriction that has been given universal applicability by the power of the state. In the absence of such environmental restrictions, however, cultural limitations on women’s employment come from the imaginations of managers themselves. There may be universal trends in the preferences of these managers; the general exclusion of women was quite common at one time. However, more particular preferences are likely to random out, losing their ability to affect more than a fraction of the labor force.
Even the most important cultural considerations are likely to be subordinated to economic considerations. In the GWR and the GPO, nightwork and sex segregation were only maintained when they could be conveniently integrated into the economic functioning of the firm. Even the preference for a male labor force was overruled when such a policy was counterbalanced by a pressing need to economize on labor. Legitimating ideologies are important in understanding the history of particular workplaces, but they operate within a very limited domain. The major parameters of sex-typing are determined by technology, labor markets, and cost structure; the minor details are then affected by local managerial preferences.
Notes
1. Note that the data in the Chapter 4 chapter cannot be wholly explained by simple sexism, with no reference to the underlying economics. It would be difficult using patriarchy alone to explain why discrimination against older women was less pronounced in blue-collar jobs than in white-collar jobs. If anything, the machismo environment of the factory should have intensified rather than reduced such discrimination.
2. I am very grateful to W. B. Creighton for a personal written communication that provided the legal analysis in this section. Creighton is the author of Working Women and the Law (1979), the definitive published work on the evolution of protectionist legislation in England.
3. This is one form of “exclusion” of which most women approved.
4. It is ironic that a policy designed to promote male employment, provide supervisory openings for men, and shorten female careers resulted in the immiseration of the lives of entry-level male clerks. There was a parallel situation in World War I, when the Sorters Union was attempting to forestall the use of female temporary sorters. The union leaders offered to have their men work double shifts if this would guarantee that the Post Office would not introduce women (WWI Sorters 1915). Both the sorters case and the telegraph case seem to illustrate Michael Reich’s contention that competition between majority and minority workers can lead to a worsening of conditions for both parties (Reich 1981).
5. Unfortunately, it is difficult to test these propositions quantitatively. There is no meaningful data on within-firm variance in office sex segregation, ruling out large-sample analyses such as those in Chapter 2. Differences between the GPO and the GWR are more likely to be due to clerical labor intensity or to differences in the status structures of the firms.
6. This job title only applied within the office. For salary purposes, the women were either grade 1 or grade 2 women clerks.
7. Unfortunately, a parallel test can not be done for the GWR, since nightwork is perfectly collinear with depot status. The most that can be said is that some caution should be used in interpreting the meaning of the depot coefficient.