Alternative Sources of Secondary Labor
The analysis thus far has presupposed an extremely simple labor market. There are only two forms of labor: adult men, who are relatively expensive, and women, who are relatively cheap. If the labor force is to be bifurcated, or treated differentially, this split will occur along sex lines. Thus, in Chapter 2 it was assumed that, if an office could not afford male labor, it would turn by default to adult women, while in Chapters 3 and 4 it was assumed that, if there were disadvantageous positions within a firm, women would be recruited for such positions.
In reality, employers who are looking for cheap labor have a wide variety of options open to them. They may hire minority workers, illegal immigrants, juveniles, or members of the white male casual labor force, all of whom work for substandard wages. Each of these groups suffers from a limited demand for their labor, which intensifies the competition among group members for employment. The overcrowding hypothesis works for other forms of secondary labor besides women (Fawcett 1918; Edgeworth 1922). Adult white males can compete for any job in the national labor market; other workers can compete for only a select few. For these groups, the supply of labor is greater than the demand, and the market wage is lowered. Blacks, hispanics, and members of other disadvantaged ethnic groups suffer from overt racism and cultural discrimination. Illegal immigrants experience both these handicaps and the additional burden of not being protected by minimum wage legislation.
Juveniles represent another important source of secondary labor. Teenagers are hurt by their low human capital attainments. Not only do they lack formal training and job experience, but they often lack work discipline and show an extremely high natural propensity to turnover (Price 1977). The lack of these more social forms of human capital can be just as harmful as the lack of years of education.
An important and often neglected component of the secondary labor force is the population of marginal white men. Many majority members find their employment opportunities severely constrained by either low human capital stocks or by permanent blemishes in their career records. The white casual labor force consists of many individuals who are uneducated or unskilled or who are persistent non-performers. This latter group includes both the incompetent and the criminally inclined. Besides these objectively poor workers, there are other individuals who despite reasonable levels of ethics and proficiency are barred from many jobs because of their past records. The employment difficulties faced by ex-convicts, ex-mental patients, or recovering alcoholics are well known. Since all of these groups face the prospect of low wages brought on by reduced labor demand, the fundamental question for the student of occupational sex-typing is to explain why any given low-status job goes to women rather than to one of these other types of workers.
This chapter cannot provide an exhaustive model of the allocation of jobs among alternative sources of secondary labor. However, some components of such a model are suggested here, and their utility is demonstrated by their relevance to the feminization of clerical work. Among the factors that help determine which groups are hired for low-status positions are the relative supply of different kinds of workers, the human capital required for adequate job performance, and the length of employee tenure that the employer prefers.
Mechanisms for Affecting Job Allocation
Since employers have to choose their workers from the labor market in which they are located, they cannot always find workers of a given socio-demographic category. Regional labor markets have different ethnic mixes. These differences involve not only what kind of disadvantaged ethnic group is present but what percentage of the working population is made up of minorities of any kind. Some inner cities are primarily black, while others are divided evenly among blacks and hispanics. A suburban employer may have limited recourse to either of these groups and thus must work with a predominantly anglo labor force. The suburban employer, however, has access to a large supply of white teenagers and adult women who may not be available in certain urban neighborhoods. Some of the change in the hiring patterns of women and minorities may have come from the transformation of urban neighborhoods from white middle class to non-white lower class, with an attendant change in the population of workers available to local firms. There are non-geographical determinants of supply as well. Certain workers may be legally barred from certain jobs. Mandatory schooling has severely constrained the supply of child labor. College students become more available in the summer; there is a comparable seasonality to the industrial availability of those laborers who also work in agriculture. All of these factors will reduce the choices open to an employer seeking cheap labor.
Despite the fact that many secondary occupations are relatively menial, some jobs have legitimate requirements that new entrants possess particular capacities or skills. Positions that require educational skills such as spelling or arithmetic, or social skills such as work discipline, are unlikely to be open to the white male casual labor force. Jobs that require fluency in the English language will not be appropriate for many illegal immigrants. Labor-intensive jobs with extremely high human capital requirements may be filled with specially trained women or college students. On the other hand, there are jobs that are sufficiently undemanding that even the white male marginal labor force can fill them adequately.
Ideal Tenure Length
Occupations differ in which length of employee career produces the greatest economies in staffing. Some jobs are fundamentally seasonal in nature. In these cases, virtually anybody can be hired. However, it may be convenient to use employees whose voluntary job tenures will be short. The amount of on-the-job training can also affect the desirability of long tenures, as was discussed in Chapter 4. The different secondary labor forces are each associated with distinctive turnover patterns. A college student taking a summer job is likely to work for only two months and quit. High school students are likely to last only one or two years. High school or college graduation is likely to produce some job change; furthermore, many adolescents vacillate between one interest and another, producing relatively intense but short-term commitments. Marginal males might have even shorter job tenures, while women and minority heads of families might be expected to stay somewhat longer. Career lengths cannot be predicted perfectly, since turnover depends somewhat on labor market conditions (Price 1977). Nevertheless, turnover rates are sufficiently predictable for them to have an impact in hiring decisions.
Alternative Sources of Secondary Labor in the Office
Clerical work provides an excellent example of how feminization can occur as a dynamic of the changing preferences of employers for different secondary labor forces. Clerical work feminized because of the increasing difficulties associated with the use of teenaged boys in office positions. In Chapter 3, we saw that clerical work did not significantly de-skill from 1870 to 1940. There were a substantial number of marginal office jobs throughout the entire period. Women came to occupy those positions that had previously been held by workers called “office men,” “office boys,” “copyists,” and “messengers.” As other groups became increasingly impractical for use as low-level office workers, women became by default the employees of preference in labor-intense, high-turnover positions.
Casual males, disadvantaged ethnic minority members, and juveniles had all been used for secondary office labor. Casual males served as male copyists and office men. The Great Western Railway gave such individuals the job title “wage clerks,” although in the depots they could also be known as “checkers” or “callers-off.” Wage clerks differed from salaried clerks in that all the terms of their employment were substantially inferior. Their pay rates were lower; their hours were worse; they lacked any guarantee of job security. Because of these relatively unattractive terms of employment, the better candidates applied for the salaried positions, leaving the wage positions to be filled with more marginal candidates.
Clerical work is not well performed by individuals with low levels of education or intelligence. A letter that is misspelled cannot be sent out. Arithmetic errors can add hours to the work involved in balancing accounts; if these errors are undiscovered, they can be very expensive. The savings that can be obtained by hiring casual men can be obliterated by the expenses of ruined work. In the Victorian era, when clerical labor costs were not a large component of total costs, such inefficiencies could be tolerated by the patriarchical employer with a preference for men. However, with the increasing administrative intensity of business, clerical labor intensity increased, and the resulting efficiency movement helped doom the male wage clerk.
The Taylorites on the Great Western Railway were dedicated opponents of temporary male clerks. In the early 1920’s the Clerical Work Committee, a collection of in-house scientific office managers, systematically reviewed the staffing of many of the offices of the GWR. In general, they found most staffing levels satisfactory and thus made relatively few systematic recommendations. However, there was one policy change they strongly advocated: the substitution of women for male wage clerks. In most of the offices they examined, they found that temporary males lacked the qualifications for efficient clerical performance and that the work they did was of very low quality. This was of especial concern because the duties that were being mishandled were not especially demanding in the first place. In each office where they studied the performance of male wage clerks, the most common recommendation was that a woman be employed (Goods Stations 1922).
At the time these studies were done, the Great Western Railway was using very few wage clerks. Although the Taylorite studies did not cover every office in the GWR, wage clerks comprised less than 5 percent of the staffs of those offices the Clerical Work Committee examined. This suggests that the Taylorite recommendations actually reflected current management practice. By 1933, the use of unskilled male office help had become negligible. Wage clerks comprised less than 2 percent of the clerical force of those departments that had no factories. Wage clerks made up a third of the clerical force of the factory departments, but even here their use was strictly confined to the shop floor, having disappeared from the main offices (Staff Charts 1914-37; GWR Staff Census 1933).1
Ethnic minorities seem to have played a minor role in nineteenth-century Victorian staffing. B. G. Orchard’s 1870 monograph on the clerks of Liverpool features a discussion of English clerks’ being undercut by cheap German immigrants, suggesting that the use of immigrants was not wholly unknown. Little is known about the role of ethnicity in either the United States or Britain. However, there are few other mentions of such phenomena either in the historical literature or in the internal documents of the GWR or the GPO. The best guess would seem to be that the impact of ethnicity was fairly small and that minority members tended to become full-fledged clerks rather than constituting an office underclass per se.
In England, one of the most important sources of secondary clerical labor was teenaged boys. In the late Victorian era, boys made up between 20 and 40 percent of the clerical staff of most offices, and in particular cases the share could be much higher. The most accurate global estimates of the use of juvenile clerks come from the British census. In 1871, in England and Wales, approximately 30 percent of the clerks were under twenty years of age. Being under twenty does not necessarily imply juvenile status. Many offices paid their clerks an adult wage at the age of eighteen or nineteen. Thus, in the fifteen-to-nineteen age group, some clerks would have been receiving full wages while others would have been paid at the junior rate, making the percentage of juveniles in the labor force somewhat less than 30 percent. However, 1870 data from Liverpool suggest that the 30 percent estimate may not be far off the mark for the late nineteenth century.
In 1870, Orchard collected data on the office personnel of over three thousand firms in Liverpool. He divided the staffs of his offices into three categories: adults, lads, and apprentices. The exact definition of these terms is not easy to discern. Obviously adults are full adults, and lads are juvenile boys; however, the composition of the apprentice category is ambiguous. In some cases in the text, Orchard describes the apprentices as youths. However, some industries show an enormously high percentage of apprentices. If all of these were youths, some industries would have been over 85 percent juvenile. Probably the apprentice ranks included some younger adults as well as juveniles and thus had the same ambiguous interpretation as clerks aged fifteen to nineteen.2
In Liverpool in 1870, between 18.4 and 44.0 percent of the clerical force was juvenile, depending on how one interprets the age of the apprentices (see Table 8.1). With the exception of banks, loan offices, and insurance companies, the office staff of every industry was at least 10 percent juvenile. Some industries were especially heavy users of boys. Without even considering the apprentices, the offices of ships’ stores suppliers, stockbrokers, general brokers, and lawyers used boys for at least 30 percent of their clerical staff. In law offices, boys made up nearly two-thirds of the clerical force. Other industries seem to have been heavy users of apprentices. Even if these individuals were entirely young adults, a claim that is probably incorrect, they were still an age-based secondary labor force. Over 25 percent of the staff of corn concerns, cotton concerns, forwarding agencies, general brokerages, shipping companies, and timber concerns were made up of apprentices alone.3
Age Composition of Clerks in Liverpool by Industry, 1870
* Closely related industrial categories in the original have been merged for simplicity of presentation.
Source: Adapted from Orchard (1870).
These data are consistent with the claim that roughly 30 percent of the clerical labor force was juvenile. The use of juvenile clerks was not universal, and industries differed dramatically in their dependence on such labor. Furthermore, it is not implausible that more boys may have been used in Liverpool than in other clerical labor markets. Nevertheless, regardless of the precise percentages involved, boy labor seems to have been an important component of the clerical secondary force and a large component of the office population overall.
The Demise of Boy Labor in the Great Western Railway and the General Post Office
Both the Great Western Railway and the General Post Office made extensive use of juveniles. Boys had been introduced in 1862 as part of the creation of the GWR’s internal labor market. Use of the boys had been sparing at first; they made up less than 5 percent of the clerical force after the first year of the program. However, once the savings of not paying adult salaries became apparent, the traffic department began to hire over one hundred boys a year, quickly raising the percentage of juveniles to a third of the clerical staff. According to the 1870 GWR staff census of the traffic department, 35.2 percent of the clerical force was made up of boy clerks. Boys constituted nearly 25 percent of the ticket collectors and were extensively used as van boys, stable lads, machine watchers, depot attendants, and messengers. Nearly 97 percent of all positions that had been created by organizational growth were filled with boy clerks. The Great Western management calculated that the use of juveniles saved the company over £2,300 a year (GWR Staff Census 1870).
The Post Office was a very uneven employer of boys. Some offices used large numbers of boy writers, sorters, and clerks. Others maintained a virtually all-adult labor force. In 1871, 22 percent of the Savings Bank clerks were lads. Thirty-seven percent of all central London sorters were boys, as well as 50 percent of all District sorters.4 In contrast, the Secretary’s Office, the Money Order Department, and RAGO used relatively few boys; those that were hired were used as messengers.
In 1881, a substantial number of juveniles were added to the postal force. The secretary’s office obtained a budget line for copyists that included boys. A new intelligence division was created to facilitate news delivery whose clerical force was over 80 percent juvenile. The money order department hired forty-eight boys, who comprised over 25 percent of the total office staff. The percentage of boy sorters in central London declined to 27 percent; however, in the districts, the percentage rose to a striking 66 percent. The new telegraph service used very few boys for either signalling or office work, although they did hire over two hundred for telegraph delivery. RAGO continued to rely exclusively on adults, while the percentage of juveniles in the Savings Bank actually declined. As in the Liverpool case, the use of boys was not universal; however, a very large subset of offices made intensive use of juveniles, making boys a major component of overall postal staffing.
By the 1930’s these boy clerks had virtually disappeared. On the Great Western Railway in 1933, only 0.8 percent of the clerical force were junior clerks. This low figure was virtually the same for males and females, suggesting that the GWR had become an all-adult labor force.5 In the Post Office, the same thing occurred. In 1931, only 2.9 percent of the Savings Bank was juvenile, with the figures for RAGO and the Money Order office being 1.9 percent and 0.7 percent respectively. Significantly, boys were no longer employed in any of these positions. The only juvenile job title left in these offices was “girl probationer.” In the sorting offices, boys had been eliminated entirely. All of the work of sorting was now done by adult males. Juvenile clerks having disappeared, the only remaining concentration of juvenile labor was in telegram delivery. With the general decline in demand for telegraph service, however, even these workers were being eliminated.
One possible explanation for the declining use of boy labor is that an increased supply of women eliminated the need to use boys. If women were in all cases more economical than boys, then an exogenous increase in the availability of women would have led to automatic substitution by employers of women for boys. The elimination of the youngest males from employment would produce an aging of the male clerical force, as this population became composed exclusively of adult workers. It is unlikely, however that the decreased use of boys was caused by an exogenous process of feminization. For one thing, boys’ salaries were traditionally lower than women’s. An employer seeking to use the cheapest workers available would have chosen an all-boy labor force. Furthermore, such an explanation would be inconsistent with the empirical patterns of change in the use of juveniles. Both the Great Western Railway and the postal sorting offices dramatically curtailed their use of boy labor, but neither of these work settings underwent significant feminization.6 If juvenile labor declined, it was because of some factor endogenous to the mechanics of hiring youth, rather than because of the availability of substitutes.
There were two problems associated with boy labor that led to its ultimate disuse. The first was that the supply of skilled boys began to decrease significantly, forcing employers to turn to alternative sources of labor. The second problem was that boys began to create serious problems of absorption as they matured into adult males. Both of these situations were created by the high human capital stocks employers required for their secondary clerks. As was argued in the case of casual adult males, clerical workers needed to be both good spellers and arithmetically competent, a requirement that limited potential recruits to those workers who were both well educated and naturally intelligent. Children of the middle class were particularly sought after because of their superior educational attainments.
One indicator of the value office employers put on educational attainments is that both the GWR and the GPO had an elaborate formal testing program with which they evaluated all potential clerical recruits. Candidates for entry-level jobs were tested in spelling, arithmetic, and the taking of dictation, with occasional extra sections in geography, foreign language, typing, and shorthand, depending on the job (GWR Women 1906; Hobhouse 1906; GPO Women 1908; Salaried Recruitment 1912). These exams were more than mere formalities. Between 1868 and 1870, over 60 percent of the candidates who sat for the GWR junior clerk exam failed (Staff Statistics 1870). The Railway News reported that in 1910, 40 percent of all job candidates for railway clerkships were disqualified from taking the exam. Subsequently, 60 percent of those who were left did not pass (Railway News, Oct. 8, 1910). In 1895, between 89 and 97 percent of all applicants to the Post Office failed the telegraphists’ and sorters’ exams (Telegraph Chronicle, Sept. 27, 1895). The outcome of the clerical exams is not known, but they were designed to be much harder. The fact that the failure rates were high does not establish that these firms had high human capital requirements. Any candidate selection process with a large number of applicants and a small number of openings will have a relatively high failure rate. What is significant, however, is that these exams were the main criteria by which job candidates were evaluated. This supports the claim that these organizations valued language and mathematical skills.
They also valued middle-class origin. Observers of the Great Western Railway in 1914 noted that recruitment for clerical positions seemed to have been limited entirely to the middle class (Salaried Recruitment 1912). Dorothy Evans noted that postal clerical positions were reserved for children of the middle class, although sorter and telegraph positions were open to more lower-class entrants (Evans 1934). To some extent, this may have been due to elitist prejudice; however, middle-class candidates may also have been the only reasonable source of those educational skills clerical employers valued. Telegraph authorities in 1873 noted that there were difficulties in finding qualified candidates in Hull, Lancaster, Leicester, Sheffield, Nottingham, Birmingham, and Manchester. They explained this in the following way: “It is hard to find learners where the severer forms of labour predominate. Towns in which a large proportion of the population belongs to the lower section of the middle class will always provide us with a plentiful supply of learners” (Telegraph Report 1873). It is unlikely that this scarcity was caused by working-class people being uninterested in the high pay and permanent job security the telegraph service offered. The shortage of candidates probably reflected the Post Office’s own selectivity.
An important consequence of this preference for high-quality recruits was that a potential clerk would have had a wide variety of alternative career possibilities. The parents of the prospective boy clerk would have been assessing all the possible forms of activity that could reasonably lead to a remunerative occupation for their son. Thus, the parents would have been choosing among clerical employment in different industries, employment in careers that did not require holding a clerkship as a first post, such as the military, and delaying entering the labor force to invest in formal education. For the better candidates, taking a boy clerkship would only have been a reasonable option if it could be expected to lead to some form of lucrative lifetime employment. The very best candidates would expect some promise of managerial employment; the next best would expect a promise of a permanent adult clerkship; only the most marginal would accept a purely temporary position.
This meant office managers had to have a coherent plan concerning the long-term prospects of their boy clerks. A teenaged boy could only be expected to receive a juvenile salary for a limited number of years. At the age of eighteen, he would qualify for an adult salary and no longer be inexpensive. One option would have been to fire all boy clerks at the age of eighteen and replace them with a set of younger recruits. This approach would have saved money, but would have discouraged the better candidates from applying for those positions. If the employer wanted a high-quality labor force, he had to guarantee permanent employment after age eighteen at an adult rate. This meant the employer had to be able to absorb an adult male with a prime salary. Over time, if an employer hired the same number of boys every year and then kept them for the full duration of their worklives, his adult male labor force would swell with each class of graduating boys. Thus, over time, there would be a gradual increase in total expenditure that would come both from the increase in the total size of the organization and from the fact that secondary labor would be making up an even smaller proportion of the total staff. The juveniles would become outnumbered by the adults.
The absorption problem is not by itself prohibitive. There are a number of conditions under which increasing the number of full adult clerks will not be a significant concern to management. If the firm is not clerical-labor-intensive, or if it is somehow buffered from office personnel costs, the increase in salaries over time will be considered acceptable. Furthermore, there are some economic situations in which even cost-conscious employers can tolerate absorption. If the skill distribution of an office is being upgraded, there may be less need to depend on secondary labor. The increasing percentage of adults will parallel the growing need for skilled labor. Likewise, if the organization is growing, the absorption will be economical. An expansion in scale will create a new set of adult positions that the graduates can fill. If there is a need to maintain a constant ratio of youths to adults, the size of the boy cohort can be increased at the same rate that the organization as a whole is growing. When organizational growth stops, the firm will have a serious absorption problem with the final cohort of boys, but such a regime could work under conditions of long-term growth. Finally, if turnover is high among adult males, the loss of highly paid employees will both stem the size of the labor force and reduce the ratio of adults to juveniles, compensating for the problem of graduating junior clerks. However, a cost-conscious employer with stagnant or declining skill demands, low organizational growth, and low turnover will find the absorption of his boy clerks prohibitively expensive. He will either have to stop promising permanent employment to his youths or seek an alternative source of secondary labor.
It can be seen that the conditions under which an apprentice system of boy labor can exist are fragile. The Great Western Railway and the General Post Office generally failed to meet these conditions, particularly in the twentieth century. As Chapter 3 showed, it is extremely unlikely that there was any kind of dramatic upgrading of skill levels in these firms. Neither stable nor declining skill levels could provide a solution to the problem of absorption. Adult male exits were unlikely to create a supply of senior vacancies. As the actuarial records demonstrated, male turnover was extremely low. Exit rates were close to zero for most age categories.
The only source of openings that was realistically available to both firms was organizational growth. Both firms grew dramatically from 1870 to 1930. In the nineteenth century, this growth would have been sufficient to support a modest percentage of boy clerks. Growth rates reached their natural limit by the end of the century, however, after which it became impossible for either firm to support a large apprenticeship system.
Tables 8.2 and 8.3 show the rates of growth for total employment in the Great Western Railway and the General Post Office. Since clerk/staff ratios are roughly constant in both firms, these figures provide a good index to the overall size of the clerical force.7 The Great Western Railway more than quadrupled between 1860 and 1884, nearly doubled again between 1884 and 1911, and then grew by less than a half between 1911 and 1931. The growth in this last period is overstated, since much of the new staff that was added during this period came from the national amalgamation of the GWR with other regional carriers. This growth thus represented absorption of existing personnel rather than the creation of new vacancies. This pattern of rapid growth slowing first to intermediate growth and then to relative stagnation is also noticeable in the Post Office, where a growth rate of 7 percent in the 1880’s declined to 3 percent at the turn of the century and to negative growth in the post-war years.8
Total Employment on the Great Western Railway, 1860–1931
Source: Railway Return (1860, 1884, 1911); GWR Staff Census (1933).
Total Employment in the General Post Office, 1871–1931
Source: Postmaster General (1871, 1891, 1911); Civil Service Return (1931).
Table 8.4 shows the rates of organizational growth that would be necessary to absorb the supply of graduating boys while maintaining the ratio of boys to total staff at any given level. The details of its construction can be found in Appendix F. The table is based on the turnover behavior of the GWR and GPO males and makes a number of simplifying assumptions. It thus only applies to the junior clerk programs of these two firms, and even with this limitation its estimates should be treated as approximate.
Small percentages of boys were easy to maintain, even with zero growth. However, any significant increase in the ratio of boys to staff could only be maintained by extraordinary increases in the levels of total staffing. The GWR and the GPO could have maintained a junior clerk program with 5 percent of their staff without requiring any growth to create vacancies. However, expanding the percentage of youths to 10 percent would have required an organizational growth rate of 3.7 percent. Having a labor force of only 10 percent boys would have meant that fully 90 percent of the workforce would have been full-priced males. However, an annual growth rate of 3.7 percent implies a doubling of the size of the organization every twenty years. Not every firm can rely on its staffs doubling in this period of time. To allow 20 percent of the staff to be juvenile, a 7.8 percent growth rate is required, and to allow 30 percent requires over a 10 percent growth rate. These imply a doubling of size in ten and seven years respectively. Such rates of growth can only be maintained for fairly short periods of time.
Rates of Annual Growth Required to Sustain Given Percentages of Boy Labor with the Terms of Employment and Turnover Rates of the Great Western Railway and the General Post Office
Source: See text and Appendix F.
The only period in which the GWR and the GPO could have tolerated an apprenticeship program involving more than 15 percent of their staffs would have been in the 1870-90 period. A staff that was 10 percent juvenile could have been maintained over the turn of the century. In the post-war period, the percentage of juveniles could not have been much more than 5 percent. These firms thus experienced a clearly declining capacity to absorb the graduates from their junior clerk programs. Even in the heyday of these programs, there were limits on the ratios of juvenile clerks that could be tolerated. The 33 percent figure that characterized the GWR Traffic department in 1870 was clearly a short-term aberration. As organizational growth declined, these firms would have had to search elsewhere for their secondary labor. This meant either abandoning the job-guarantee programs and accepting a lower quality of boy, turning to casual adult males, or using women. Only educated women could have provided the numerical and linguistic skills that employers preferred in their clerical forces.
The problems of absorption, while serious, would not have been sufficient to eliminate boys from clerical positions completely. Some clerical positions, such as sorter, required less human capital than others. Presumably, poorly educated youths could have filled these less demanding positions; such youths would not have required a permanent guarantee of employment and thus would have not posed absorption problems. Furthermore, juvenile clerks came to represent less than 2 percent of the clerical employees of these firms, despite the fact that a force that was 5 percent juvenile could have been carried quite easily.
An equally fundamental problem was that boy clerks of any quality were becoming increasingly scarce. Levels of educational attainment were consistently rising during this period. While, on the one hand, this would have increased the number of boys who were capable of doing clerical work, it reduced the number who were willing to take such positions. This was not particularly due to mandatory school attendance laws. Between 1870 and World War II, the age of compulsory school attendance never rose higher than fifteen, a year younger than the customary recruitment age for junior clerks. The greater problem was that many of the more talented youths were voluntarily delaying their labor force entry in order to further their education. Increasingly, from the boys’ point of view, the way to obtain a high-paying job was not to take a clerical apprenticeship but instead to graduate from college. With the loss of college-bound individuals from the clerical recruitment pools, private employers found the quality of their applicants steadily deteriorating.
For the Great Western Railway, this problem seems to have become particularly acute in the early years of the twentieth century. Discussions of recruitment during this period centered around the poor quality of the available junior clerks. On two different occasions, 1901 and 1912, the GWR management circulated memos among itself seeking solutions to the problem of under-qualified clerks. In 1901, they tightened the standards for moving from junior to senior clerk by providing a more difficult senior clerk exam; in 1912, they extended the period of probation to facilitate the expulsion of unsatisfactory recruits (Salaried Recruitment 1912; GWR Petitions 1914).9 Such plans would have encouraged further study among those junior clerks who were already employed; however, it did little to improve the quality of the candidate pool. As one GWR manager put it, “The service has ceased to attract youths educated at higher class schools” (GWR Petitions 1914). The response taken in 1912 was to raise the salaries for entry-level junior clerks. It was felt that by providing a better level of initial remuneration, the GWR would at least become financially competitive with other employers.
One alternative that was considered at the time was the recruitment of university graduates. If the best candidates were all attending college, it made sense to hire them when they finished their education. Despite the presence of strong advocates of hiring university men, the GWR never followed this course of action until after World War II. In part this was due to a somewhat exaggerated concern for preserving traditional chains of promotion. However, an important factor in this decision was that university men could hardly serve the function of providing secondary labor that motivated the original use of boys. To summarize the objections of two managers who considered the use of college graduates for the Audit department, the university men would be quite expensive, and since the work given to entry-level clerks involved “care and accuracy rather than great intellectual power,” the use of university graduates could only be considered in those settings where trainees would be immediately involved in technical and substantive decision-making (Salaried Recruitment 1912). At one time, talented individuals had had to wait for responsible positions, and while they waited they could be put to work copying receipts and filling inkwells. Now, they had to be promoted directly, and someone else would have to be found for the preliminary duties.
It is significant that complaints about the quantity and quality of junior clerks became an issue only in the twentieth century. Complaints about a shortage of juvenile labor can be found in the 1930’s as well as the 1900’s and teens (Labour Report 1930). However, there is no mention of scarcity in any discussion of juvenile labor in the nineteenth century. The 1870 GWR Staff Statistics report an ample supply of qualified junior examination candidates. The 1912 Salaried Recruitment file contains documents from both the nineteenth and twentieth centuries. Although the nineteenth-century documents in this folder were explicitly selected by 1901 managers for their relevance to the current shortage of qualified juveniles, none of these records makes any mention of a tradition of supply problems. No mention of a shortage appears in any other personnel records either. The fact that scarcity problems only started to emerge when education was on the rise strongly supports the claim that competing opportunities were hampering the viability of boys as a secondary clerical force.10
The reduction in the supply of boy clerks played an important role in facilitating the feminization of office work. The elimination of boys did not automatically produce a female labor force. The Great Western Railway placed adult males in many of the former juvenile positions. The Post Office Engineers replaced boys with women only in certain circumstances. Typing positions continued to be predominantly male, particularly in the Engineer in Chiefs office. Many boy clerks would have been trained on the typewriter during their apprenticeships; now the Engineers hired male graduates of the typing schools. However, some of the duties of the former boy clerks were transferred to clerical assistants, an entry-level position for adult women (Engineering Clerks 1919).
A significant blue-collar illustration of the effect of the supply of boy labor on women’s opportunities can be found in the Post Office Savings Bank Printing Shop. The Savings Bank used an extremely large volume of passbooks, certificates, and pre-printed forms. To produce these materials as needed, the Post Office maintained a print shop that employed over fifty people. In the nineteenth century, the staff had consisted of a number of male adult printers and a large staff of boys. The boys did all of the unskilled work associated with form production, such as cutting and separating forms or removing finished work from the press. The Post Office did not run a printing apprenticeship program, its printers being hired fully trained from the outside world. Therefore, being a boy helper was a dead-end job without any prospects of internal promotion or absorption by other departments. Since the human capital requirements for being a helper were low, the Savings Bank staffed the print shop with temporaries. A boy would be hired with the understanding that he must resign at the age of eighteen.
After 1910, it became official civil service policy not to hire any boys who could not be absorbed. This policy came from a national political concern with unemployment. Boys who took dead-end apprenticeships were often burdened with skills that had no future labor market value and were subsequently very difficult to re-employ. In attempting to reduce the national supply of such unhirables, the British government decided to act as a model employer and only hire those juveniles to whom a permanent job could be promised.
The foreman of the print shop was therefore informed that he no longer would be allowed to use boy labor. Since the budget for printing was extremely restricted, it would have been impossible to run the shop with a full complement of men. The foreman vehemently resisted the suggestion that he use female labor. Fully a year and a half was absorbed in a bureaucratic battle similar to that fought by the Savings Bank Controller in the 1870’s. The foreman’s main tactic was to seek out offices that used blue-collar men that would have been willing to commit themselves to creating job openings for print shop graduates. When this failed, the foreman accepted the inevitable and converted most of his staff to women. Thus the Savings Bank Print Shop became one of the few British printing establishments in 1914 that was nearly 60 percent female. The key was the artificial reduction of the supply of boys from which most private printers would have been exempt (Savings Bank Print Shop 1914).
The Decline of Boy Clerks in Britain as a Whole
Although the decline in the use of boy clerks in the GWR and the GPO was especially dramatic, a similar if more modest decline occurred in Great Britain as a whole. Table 8.5 shows the percentage of commercial clerks in England and Wales who were under the age of twenty from 1871 to 1931. During this period, the percentage of clerks under twenty declined only from 31.4 percent to 26.5 percent. These gross figures are somewhat misleading, however, since they conflate two phenomena: the decline in the use of juvenile male clerks and the increase in discrimination against older female clerks. The female clerical force became significantly younger between 1871 and 1931. This is quite likely to have been the result of an increase in the use of synthetic turnover. The selective recruitment of single teenagers, combined with the non-retention of married women, would have had the effect of driving older women from the labor force and lowering the average age of female clerks.
The data for male clerks, which is free from such influences, show a clear trend away from the hiring of juveniles. Of all male clerks, 31.5 percent were twenty or younger in 1871. This figure fell to 20.3 percent in 1931. Boy clerks had lost fully a third of their relative share of male employment. The most dramatic reduction occurred between 1911 and 1921, but most of the other periods saw some degree of relative decline. While a drop from 30 to 20 percent is noteworthy, it is not nearly as striking as the complete elimination of boy labor that occurred in the GWR and GPO.
These figures exaggerate the number of juveniles working in Britain. In 1931, only 8.2 percent of all male clerks were aged fourteen to seventeen. The overwhelming majority of clerks under twenty-one were aged eighteen to twenty and were, for employment purposes, adults. Data limitations prevent the calculation of comparably corrected figures for earlier periods. Given that the Orchard data suggest that a third of all male clerks in 1871 were juveniles, while in 1931 we know for certain that this figure was 8.2 percent, if anything the decline in boy clerkships was even more dramatic than is suggested by Table 8.5. This drop, however, would have been less than the decline that was experienced by the two main firms in the study. While the figures in Table 8.5 are misleading in an absolute sense, the bias they introduce is conservative. What remains to be explained is why these figures declined, and why this decline was less steep than that observed in the two-case analysis.
The more modest nature of the national changes can be attributed to two factors. First, non-railway commercial firms would have been able to absorb a higher percentage of their boy clerks than would have been possible for the GWR or the GPO. The Great Western Railway and the General Post Office had very low turnover rates. Employees were reluctant to leave voluntarily because both firms paid premium wages. Furthermore, the policy of lifetime employment that was practiced by both organizations virtually eliminated involuntary turnover among males. Because other private employers paid lower wages and were more willing to lay off employees, there would have been more circulation of staff in this sector, which would have created vacancies that could be filled by rising boy clerks.
Second, the GWR and the GPO had completed their growth by the end of the nineteenth century. They were thus incapable of absorbing boy clerks through expansion. Other industries would not have been subject to this kind of restraint. Both banking and insurance saw substantial growth in the twentieth century. The number of manufacturing clerkships would have increased both from economic expansion and from the increasing clerical labor intensity of the administration of production. It would have been much easier to maintain large apprenticeship programs in those industries undergoing steady long-term organizational growth.
The pressure to eliminate boy clerks would have come from the supply side. Qualified candidates for boy clerkships would have become increasingly scarce as the brighter youths turned instead to higher education. This would have been exacerbated by the bifurcation of internal labor markets within the office. As employers learned to differentiate between managerial and clerical career paths and started to recruit potential executives to the managerial track only, school-leavers would have become increasingly reluctant to accept clerical positions. Faced with a growing scarcity of educated boys willing to do menial work, employers would have had little choice but to abandon their apprenticeship programs and staff their low-level positions with someone else. Although boy clerk programs were not as severely threatened in Britain generally as they were in the GWR and the GPO, there was still enough of a supply crisis to make the widespread use of boy clerkships problematic. This is why the percentage of male clerks under twenty declined between 1871 and 1931.
We have seen the importance of other secondary labor forces in determining the opportunities that are open to women, with considerations of relative supply, human capital, and ideal tenure length influencing the relative attractiveness of competing sources of cheap labor. The most detailed discussion has been of the trade-off between women and boys, but the same logic could be applied to casual white males or to members of disadvantaged minority groups.
There is suggestive evidence that these factors apply to other occupations besides clerical work. Thomas Dublin has documented a trade-off between the employment of women and the employment of Irish in the Lynn textile mills of the nineteenth century. Originally Yankee farm girls were hired as cotton spinners. While natural training with textiles may have had something to do with this allocation, an equally important consideration was their relative cheapness. In the company houses near the factories, however, a tight occupational community formed that was capable of generating a strong sense of solidarity. This resulted in a dramatic strike by the women operatives against the Lynn mill owners. Subsequently, the owners began to phase out female employment. A new wave of Irish immigration provided an alternative supply of labor that promised docility as well as low wages (Dublin 1979). In this case, women were rejected for Irish, but there is no reason why this history could not have been reversed, with women undermining a group of militant Irish.
An important example from the quantitative literature comes from the work of David Snyder, Mark Hayward, and Paula Hudis. In an investigation of the relationship between sex-typing and location in core versus peripheral industries, Snyder et al. had a striking finding. Over time, the sex composition of the primary sector was generally stable; however, counterintuitively, the second sector was undergoing sex change in both directions. Some occupations were feminizing quickly, while others were de-feminizing. The robustness of these findings is slightly suspect, since the measurement of sectors was fairly crude; however, if these findings are right, they are suggestive. It is unlikely that, in the secondary sector, women were being replaced by white men. It is more probable that women were being replaced by teenagers or ethnic workers of either sex. Since change was occurring in both directions, Snyder et al. suggested that the secondary sector represents a continual flux of recirculating labor forces. It seems quite likely that some of these transformations can be explained by the changing preferences of employers, as they shuttle from one secondary labor force to another looking for marginal advantages (Snyder et al. 1978).
1. The General Post Office also used temporary male wage clerks; their job title was “male writer.” They were largely eliminated by the 1890’s, which could have been due to the inefficiencies of using casual men. However, official civil service policy was to give all employees permanent tenure and avoid the use of temporary personnel. Thus, the elimination of the wage clerks may have been caused by the special circumstances involved with public employment, rather than by any deficiency of the writers themselves.
2. It is likely that different respondents probably defined the term “apprentices” in different ways. The Great Western Railway had a Liverpool sales office that would have been in Orchard’s sample. The GWR had no employees called apprentice per se. They had lad clerks, who were apprentices in every reasonable sense of the term, and they had starting-level senior clerks. How the manager of the Liverpool sales office would have divided his staff between lads and apprentices on his form is a matter of guesswork.
3. The conclusions that are drawn from these materials must be tempered to some extent by a consideration of the quality of the data. The statistics that Orchard generated are somewhat suspect; there is a high percentage of crudely rounded numbers, and the data show a number of arithmetic inconsistencies. Although these materials are probably not accurate enough for elaborate statistical analysis, however, they are probably acceptable as rough estimates of the labor force composition. The errors that were the product of rounding, arithmetic error, and informal estimation are likely to have been random.
4. Central London here refers to the London EC district, which is the main sorting office for the country; it also processes deliveries for the City of London. The districts include most of the central areas that are not in the City of London, such as Piccadilly Circus, Whitehall, or Knightsbridge, as well as more distant areas such as Clapham and Hammersmith.
5. The figures for the GWR as a whole are identical to those of the traffic department considered separately, making these statistics comparable to those calculated for 1870.
6. In the traffic department in 1870, 32.5 percent of all clerks were boys, while in 1933,0.7 percent of all clerks were boys. Since 9.6 percent of the clerks were female, this suggests that the remaining 22.2 percent of the clerical positions, which in 1871 had been filled with boys, were in 1931 being filled by adult men.
7. Reliable estimates of the size of the Great Western Railway clerical force are not available for all dates.
8. The exponential growth rate is the rate of increase that if applied continuously would provide the change in absolute employment observed in the data. It is in many ways analagous to a compound interest rate (Shyrock and Seigel 1975).
9. There were two efficiency checks for junior clerks. There was a probationary period of three to twelve months, during which time a youth could be dismissed summarily. Subsequently, at the end of two years, there was a senior clerk exam, which determined eligibility for promotion to an adult position. The overwhelming majority of junior clerks passed both requirements.
10. The absence of a reference to scarcity could also be due to an idiosyncratic survival pattern among the GWR documents. While the presence of such a methodological artifact cannot be wholly discounted, it is nevertheless significant that an enormous body of personnel records and correspondence survive from this period. Since the GWR archives are far more reliable and complete than most corporate archives, it seems likely that any major staffing problem that merited systematic attention would show up somewhere in the documentation.