Organizations and occupations by now have been thoroughly studied, and the general theoretical principles involved are relatively clear. It is striking how little the technological version of functionalist argument shows up in this area. For the major findings of organizational studies—from the industrial management tradition of Mayo, Barnard, and Simon; from the Chicago tradition of Hughes, Wilensky, Goffman, and Dalton; from the power-oriented analyses of Weber and Michels, Selznick, Gouldner, Etzioni, and Crozier—have been that organizations are sites of incessant maneuvers for autonomy and control. The main actors on the stage are informal groups, alike at the level of workers and managers, staff and line, professional and layman. Technology enters into the picture above all as it affects the resources for power struggles. In Weber and Michels, we see it as a resource for centralized control, especially as it allows for concentration of the channels of communication upon which organizational members depend. More recent studies (e.g., Simon et al., 1954) have shown many instances of it operating with the reverse effect, providing areas of local autonomy down in the organizational structure where members control access to areas of uncertainty and hence shape the communications upon which their hierarchic superiors must act. Crozier (1964: 145–174) and Wilensky (1956, 1968) have raised this to a systematic principle, showing that the power of technical experts can be either high or low, depending on whether they can control exclusive access to an area of crucial uncertainty. Accordingly, technicians are motivated to shape their “skills” for maximal political effect on the organization, in some instances keeping them traditionalistic, in others innovative, depending on where the greatest threat to their autonomy is found. Technology is thus a resource whose use is shaped by its place in the more central organizational process: the struggle of informal groups for favorable power positions.
Theory and research on organizations has put technology in an important place. It is crucial in explaining organizational structure and organizational processes. Organizational structure is correlated with the type of task attempted and with the technology available to carry it out (Woodward, 1965; Collins, 1975: 315–329). Organization theory distinguishes four main types:
1. In unit production, a single unique item or a few items are produced at a time, as in skilled crafts or innovative engineering projects.
2. Mass production consists in the making (or procuring) and assembling of many separate parts, as in most manufacturing of machinery.
3. Process production is a continuous flow of materials from one stage of processing to the next, as in chemical plants or many kinds of food production.
4. In pooled production, a number of independent units or individuals all carry out similar operations, supplied and administered by a central staff, as in banking or many clerical agencies.
Each technology is associated with a distinctive type of organizational structure. It is relatively flat and unbureaucratic in unit production. Mass production organizations are the archetypal organizational pyramid, highly bureaucratic, with many specialized staff and line divisions, and with considerable divergence of informal and formal networks. Process production is organized as a relatively smoothly operating bureaucracy, with few lower-level workers and relatively routine administrative divisions above. Pooled production has a large base and a relatively small administrative component and tends to be bureaucratically rigid and static.
Does this mean that we can speak straightforwardly of technological determinism in organization? Not if we mean to imply that other variables do not remain important for our explanations. The question is: Into what social relations does technology fit? Production is not a simple issue of moving and combining materials, but of controlling, coordinating, and motivating the individual workers. Organizations are structures of power. Since power both leads to intrinsic rewards or pains and is the means for appropriating wealth and physical ease, any power structure generates an implicit struggle to gain, hold, or evade control. Different tasks attempted and different technologies to work with lead to differences in the organization’s control problems.
The principal variable is: How standardized and easily inspectable are the outcomes of work? For example, unit production has a relatively undifferentiated organization because the innovative, highly unsettled work patterns result in difficulties in establishing any rigid hierarchic controls; power is more equally spread among the participants, resulting in a flatter, more informal structure. Mass production, however, results in a sharp differentiation of the lower-level workers with repetitive, highly controllable jobs from the higher-level workers involved in the activities of coordinating the various parts of the process, negotiating its schedules, and experiencing the unpredictable difficulties of this operation. The result is a very hierarchic, complex, and conflictual bureaucracy. Tasks and technologies also affect power struggles by establishing particular degrees of coordination problems. Pooled production involves relatively easy coordination, since administrators have only to provide resources and check up on personnel whose activities are essentially independent of each other. Process production also makes coordination easy, since machines do most of it, linking the various areas together automatically and reducing the worker’s role in checking and troubleshooting. High coordination problems, on the other hand, exist in unit and mass production types, although they have different ways of structurally dealing with this because of their differences in the inspectability of the basic tasks and hence in the autonomy of the lower-level workers.
In other words, technology feeds resources and problems into an ongoing social struggle for control and autonomy and for the goods and rewards one may get from organizations. The correlation between technologies and types of organizational structure is explainable only because different kinds of problems and material resources lead to different sorts of distributions of power in these struggles. A complex technology does not necessarily result in increased skill demands of the workers; its effects can go in either direction. The greatest power within organizations does not necessarily go to those with the highest technical skills nor to those working directly with the most advanced machinery. Rather, the technology merely sets part of the stage for the whole environment of organizational power struggles. The crucial variable is the ease or difficulty of control. If highly productive machinery or highly technical skills are easily planned, their outcomes easily predictable and observable, then the power of their practitioners is low, since it can be routinely controlled by others.
Power depends instead on being in control of an area of uncertainty that is crucial for other organization members. This has been illustrated for modest levels of power as well as for the higher levels. In a rigidly bureaucratic production organization, maintenance men have relatively high power because they alone can deal with breakdowns (Crozier, 1964: 145–208). This does not mean that they dominate the organization, but only that they receive much informal deference, since they can bargain with the ability to hold up others’ work. The maintenance men, however, cannot make organizational policy in general, or hire and fire others; that depends on control of other resources whose provision has an element of uncertainty controlled by a different level of the organization. Top managers (and sometimes certain staff advisors) have this power when they operate in the external environment of the organization in which political and financial support is generated (Wilensky, 1956). Some areas of uncertainty, in other words, are more fundamental than others. Maintenance men may control short-run, intermittent uncertainties of machine breakdowns; top administrators and financial politicians control access to negotiations over crucial resources of money and influence that may keep the entire organization operative or destroy it entirely.
Notice how these uncertainties relate to technology. It is not where technology is most reliable that its practitioners have greatest power, nor where it is least reliable, but where it is in a middle range. Experts must be effective, but not in a fashion absolutely predictable by outsiders. Imperfect technology is of the greatest advantage for power. But this technology also must be astride the flow of production in such a way that others are dependent upon it. Above all, it is how the particular task involved fits into the structure of organizational coordination that is crucial. Analogously, occupations derive greater or lesser power to become autonomous, self-regulating, highly prestigious, and remunerative professions, depending on whether they involve a task skill in this esoteric middle range of effectiveness and whether it is a skill crucial to the profession’s clients (as we shall see in Chapter 6).
The most crucial uncertainties, then, may not be closely related to the basic productive technology at all. The most fundamental resource for any business organization is its finances; for any government organization, finances are mediated by its environment of political support or opposition. Those who control the finances control the very presence of all other participants in the organization. However, this power is not absolute. In day-to-day operations, once an organizational pattern has been set up, maintenance men and other analogous controllers of local, intermediary, and internal areas of uncertainty have their moments of power when they cannot be overruled. In a more complex view of an organization, we would recognize many possible modes of power at different moments; the internal communications hierarchy of the middle-level administrators, by passing along, withholding, and structuring information about one part of the organization for the view of other parts, has power analogous to (but more local than) those of the externally oriented administrators at the top.
From one viewpoint, the “political” side of organizations—the uncertainties of information, finances, and external influences mediated by administrators—can be seen as control of a type of technology, the technology of communications. And indeed, historical differences among types of organizations depend strongly upon the availability of different types of technology (Collins, 1975: 315–413). The shift from patrimonial to bureaucratic forms of organization depended upon the introduction of massive communications technologies in the form of written reports and records. Modern communication and transportation technologies—telephones, automobiles, duplicating machines, computers, etc.—have further shifted resources for control in the administrative arena. The tradition of analysis that most successfully explains power within political organizations, for example, within political parties, focuses on control of the means of communications as the key resource. In one sense, then, we can see organizational control as highly affected by the crucial resources of technology. But we must distinguish between technologies of production and technologies of administration and communication. It is the latter that rules organizations in the sense of giving its controllers the positions of highest prestige, material rewards, and of power in the long-run sense of major organization-shaping decisions.
If we push our analysis one step further, we may see that the crucial factors in administrative power are fundamentally social and political. The key technological resources are the means of communication and administration, but these are the outer forms of the structure of domination. That structure itself is a network of persons making bargains and threats and manipulating expectancies and solidarities. It is a social network in which participants play on the ideas and emotions of the others in an effort to acquire their confidence or awe, which in turn will result in the financial and political credit that can be transmitted into internal organizational domination. It is through successful interpersonal politics of this sort that communications technology is made to pay off, and it is through this kind of success that further access to communications technology is generated. The technology itself is thus appropriated by a political process.
The fundamentals of that process are essentially interpersonal, which is to say cultural. The key resource of powerful individuals is their ability to impress and manipulate a network of social contacts. This depends on their easy understanding of the criteria of membership and prestige, and their ability to use them—first, for social acceptance into the network itself and then for prominence within it. These cultural styles may come from previous status group traditions (such as class, ethnic, or educational cultures), or they may be to some extent created anew within the experiences of that organizational network. But it is ultimately this culture of ‘‘organizational politics” that is the crucial resource, one that can bring all the other resources into its orbit.
Technology, then, is an important aspect of organizations, and it is a variable that correlates with major differences in organizational structure and with differences in power position within organizations. It is not so much the productiveness of the technology that bestows power, however; it is the way that technology plays into the political process of organizations that makes it a variable determining organization structure. Ultimately, the political process rests upon social and cultural ground—the ground of human manipulation of communications into various kinds of interpersonal networks.
There is considerable evidence that the “demands” of any occupational position are not fixed, but represent whatever behavior is settled upon in bargaining between the persons who fill the positions and those who attempt to control them. Individuals want jobs primarily for the rewards to themselves in material goods, power, and prestige; the amount of productive skill they must demonstrate to hold their positions depends on how much clients, customers, or employers can successfully demand of them, and this in turn depends on the balance of power between workers and those who pay for their work.
Much evidence has shown that work groups have informal structures as well as formal ones. Typically, the informal group in a work-place sets and enforces norms controlling the rate of work of its members. Such informal controls over output have been most widely studied among workers in manufacturing, but they have also been found among sales and clerical personnel (Roy, 1952: 427–442; Blau, 1955; Lombard, 1955). The rate of work is usually set at a comfortable level, considerably below maximum possible output.
Managers have some formal discretion to set their own work pace, and they put in a longer working day on average than do other categories of workers (Wilensky, 1961: 33–56). Nevertheless, it does not seem likely that the average manager is forced to work at his or her highest possible level of skill and effort either. An informal organization exists among managers as among other workers, and the amount of work that administrators do depends at least as much on their initiative and cooperation as on formal authority (Barnard, 1938; Merton et al., 1952: 397–422; Dalton, 1959). The widespread existence of “bureaucratic pathologies,” such as evasion of responsibility, empire building, and displacement of ends by means (“red tape”), indicates that administrators often use their discretion in ways that minimize rather than maximize efficient performance. Since administrative work is only indirectly related to the output of the organization, it is difficult for employers to judge very precisely what the performance of their administrators is. Indeed, the busy supervisor rarely checks up on many of his managers’ activities, especially the activities of those in staff positions (Strauss and Sayles, 1960: 463; Dill et al., 1962).
The employer’s conception of the skill requirements of most jobs, then, tends to be rather imprecise. In March and Simon’s terms, the strategy is usually to “satisfice” rather than to optimize—that is, to set average levels of performance as satisfactory and to make changes in procedures or personnel only when the performance falls noticeably below the minimum standards (1958: 140–141). At all levels, wherever informal organizations exist, it appears that standards of performance reflect the power of the groups involved.
Ability Tests
Efforts have been made to predict work performance by objective tests, but without great success at this time. Part of the difficulty has been the problem of directly measuring performance, except on specific mechanical tasks (such as those involved in piloting an airplane) or in tests of academic knowledge (Anastasi, 1967: 297–306; Wesman, 1968: 267–274). Many tests have been based on measurements of attitudes and interests of incumbent job holders, such as samples of current managers. These tests are used to select new recruits who reveal personality traits similar to those already employed. Research psychologists have pointed out that such tests generally lack validation. In order to check them, a control group of persons lacking these traits must also be hired so that the performance of the two groups might be compared. But this has not been done. Given the imprecise nature of job assessment, it seems likely that a wide variety of personnel can perform most jobs satisfactorily; those persons selected by the use of tests do well enough because of the self-fulfilling prophecy involved in the use of nonvalidated tests.
Evidence that the ability requirements of many jobs in modern society are not stringent is provided by a study of 109 retarded children with IQs near 60, followed up when they were over 50 years old (Bailer et al., 1967: 235–327). Few were in institutions. Most worked as unskilled or semiskilled laborers, but 18% held white-collar jobs, including positions as office and sales clerk, policeman, shop foreman, auto and real estate salesman, photographer, laboratory worker, and businessman. Occupational success was not related to intelligence differences within this group. Rather, the more successful workers were those who had the middle-class patterns of dress, speech, and personal behavior, and those who had worked throughout their careers in a single big business or who early attained some skill such as barbering. Other studies have found that there is a considerable range of measured intelligence within any occupational group so that although professionals have higher average intelligence than manual workers, there is a considerable degree of overlap among them (Thomas, 1956: 285–310).
It is possible to see why measures of ability, such as IQ, cannot be used as simple predictors of success. “Intelligence” operates primarily within the context of the school system, and its greatest utility is in predicting grades (Duncan, 1966a, 1966b, 1968; Bajema, 1968: 317–319). Moreover, motivation and the cumulative effects of past learning are deeply entwined with inherent differences in capacity in affecting performance on such tests. The IQ scores of children diverge as they grow older and are affected by different environments; these scores also diverge as children build up different styles of behavior (Turner, 1964; Rehberg et al., 1970).1 It is most likely for these reasons that both school grades and IQ scores are highly related to social class and to race insofar as it corresponds to class distinctions.
The whole subject of innate versus environmental effects on IQ, including recent controversy over racial differences, is given undue prominence as a cause of social inequality. If the utility of IQ tests is confined almost entirely to measuring capacity for academic success—which is not surprising in view of the fact that IQ tests were developed for use in schools and have received few other applications—then the major issue concerns how and why education is tied to occupational stratification. If the ability differences, whether inherent or acquired, that determine school success are only artificially linked to occupational success through the certification value of school degrees, then the psychological problems of individual adaptation to the school system become secondary to the structural problems of explaining the links among social institutions. In practical terms, selection through the schools rewards a particular sort of conforming behavior.
Career Processes
Detailed evidence on the processes of organizational careers shows the overwhelming importance of both informal ties and the struggle to control positions. Studies of promotion patterns show a prevailing pattern of personal sponsorship, whether within industrial organizations, medical careers, or trade union bureaucracies (Glaser, 1968: 191–257). It is because of the importance of personal acceptability to the sponsoring official and group that ethnic background, old school ties, and club memberships are important in careers. Ethnic distinctions among positions have been pervasive in American industry (Nosow, 1956). A study of a New England factory found that promotion to managerial positions went heavily to WASPs or Germans, Republicans, Masons, and local yacht club members (Dalton, 1951: 407–415). A similar study of industries and businesses in one metropolitan area of the South found that promotion depended on membership in civic organizations, family status, and personal friendships (Coates and Pelligrin, 1957: 200–215). Social origins affect one’s career by making one acceptable in the informal culture of the organization, even when they do not directly give one a position by inheritance or family influence.
Even where there are formal tests and evaluations, as in military promotions, often combined with formal rules of “up or out” by certain ages or time in grade, informal connections are crucial in acquiring the kinds of assignments that bring an individual to the attention of superiors and provide the background considered appropriate for promotion (Janowitz, 1968: 211–215). In industrial organizations as well, informal connections are crucial in assignment to those positions that allow horizontal movement and in avoiding too specialized dead-end jobs (Martin and Strauss, 1968: 203–210). The key to success, in other words, has little to do with skilled performance per se, but rather with maneuvering to reach the sequence of positions that lead upward. Positions as assistant to plant manager are a first key; in big corporations, the final line to the top moves through the biggest divisions or plants.
Studies of organizational succession show the same pattern. As in the theory of vacancy chains (White, 1970), such evidence shows that individual success depends much less on the quality of performance than on being in the right place at the right time; insofar as individual cleverness or motivation pays off, it is by differentiating between persons who are aware of such contingencies and those who are not. The very fact that filling a given vacancy sets off a chain reaction of promotions helps to reinforce informal organization around cliques; a vertical group of leaders, protégés, and subprotégés attempts to capture an entire line of vacancies, and part of the “political” strategy of organizational leaders is the deliberate manipulation of vacancy prospects to ensure subordinates’ loyalty (Glaser, 1968: 307–376).
Demotion fits the pattern again. The striking thing about evidence on demotions is that so much effort is spent on disguising demotions, by giving meaningless titles, transfers, and failing to promote; overt demotion into a lower position is very rare indeed for incumbents of white-collar or managerial positions (Glaser, 1968: 259–306). A ruthless application of criteria of efficient performance is rarely seen; instead, one gets the impression that white-collar careers are heavily ritualized and devoted to disconnecting as far as possible any direct connection between performance and reward. The causes cited for demotions do not mention merit judgments; rather, individuals are caught in impersonal conditions such as sharply fluctuating work loads, organizational crises caused by mergers or lack of profits, position obsolescence, or previous promotion to levels where rivalry for further movement becomes too great for the individual’s resources (usually his or her informal connections relative to others’) (More, 1968: 287–294).
One supposes that there must be some “merit” factor in organizational careers, but within any large organizational structure, it seems to be very difficult to measure the individual’s contribution apart from the rest of the organization and the external contingencies of the time. Certainly organizations spend little effort on serious, comparative measures of individual efficiency (i.e., attempting to hold other factors constant). The overriding fact is that an organizational career is made in a political environment, and success goes to those individuals who recognize that fact and act on it most assiduously. The one who makes it to the top is the organizational politician, concerned above all with informal ties, maneuvering toward the crucial gatekeepers, avoiding the organizational contingencies that trap the less wary.
The basic problem of all organizational authorities is controlling members and gaining their active cooperation. Organization theorists have listed a number of means of control and have suggested that different controls may be substituted for one another (see especially Etzioni, 1961; Collins, 1975: 298–315). If an organization can exercise normative control (internalization of organizational goals and values by its participants), it can improve performance and avoid the disruptive consequences of relying on coercive authority or purely financial rewards. The expensiveness of labor turnover adds a direct economic incentive for selecting workers for their loyalty to the organization.
Organizational authorities may develop loyalty or normative control in several ways: by recruiting already committed persons; by indoctrinating new recruits within the organization; by exposing its members to the influence of informal work groups and isolating them from contacts with outside; by offering them possibilities for promotion; by offering pensions and fringe benefits tied to length of service; or by selecting individuals who have acquired the proper values and attitudes through education. The higher the organization’s demand for loyalty, the more of these means of control it may be expected to use.
All jobs require some initiative and cooperation, but some require more than others. Managers have more autonomy and responsibility than production and clerical workers; hence it is most important to exercise normative control over managers. Moreover, the managerial tasks in some organizations require more loyalty than in others. In some organizations there is considerable emphasis on interpersonal coordination, especially through personal contacts; such coordination can be accomplished most easily if the people involved share common values and backgrounds. A manager’s success often depends on his or her “personality” (i.e., social skills and values). According to the personnel managers in 90 large American corporations surveyed in the 1950s, motivation, interpersonal skill, and moral character are the most important qualities for business managers (Gordon and Howell, 1959: 79–80). Extremely high intellect is not considered necessary nor even desirable, and specialized or technical skills are considered to be of minor importance, at least above the level of lower supervisory jobs and specialized staff jobs such as accounting. The importance of common social standards for good managerial coordination is illustrated by the administrative difficulties of the early period in the United Nations, when there were great differences in cultural styles among staff members (Kehoe, 1949: 375–380). Managers often advance in an organization by becoming part of a team or clique, usually working closely as assistants to a sponsoring executive. Failure, too, depends heavily on social skills; a survey of 76 large corporations in 1952 reported that 90% of the managers dismissed from their jobs were reported to have been lacking in desirable personality traits, not in technical skills (Martin and Strauss, 1968: 203–210).
Evidence that education has been used as a means of cultural selection may be found in several sources. Hollingshead’s study of midwestern school children, school dropouts, and community attitudes toward them suggests that employers used education as a means of selecting employees with middle-class attributes (1949: 360–388). School dropouts were largely from lower-class and working-class families, especially because of their conflict with middle-class values in the school. Although their expectation was to improve their economic position by finding immediately well-paying jobs, employment opportunities for them were limited to very menial positions, since employers regarded dropouts as untrustworthy and lacking in the qualities of desirable employees. Although high school graduates may also begin work at lower levels, their possession of education qualifies them in the eyes of employers (and in their own eyes) both for higher pay and promotion opportunities not open to dropouts. Similarly, a 1945–1946 survey of 240 employers in New Haven, Connecticut, and Charlotte, North Carolina, indicated that employers regarded education as a screening device for employees with desirable (i.e., middle-class) character and demeanor (Noland and Bakke, 1949: 20–63).
A more systematic test is provided by a 1967 survey of 309 California organizations.2 The relative normative control emphasis of an organization was indicated in three ways:
1. It is indicated by the organization’s relative emphasis on the absence of police record for job applicants; this proved to be significantly correlated (γ = .25) with an index of the organization’s educational requirements.
2. It is indicated by the relative emphasis on a record of job loyalty among applications; reluctance to hire “job-hoppers” also proved correlated (τ − b = .18) with educational requirements.
3. It is also indicated by a classification of organizational goals, deriving from Etzioni’s comparative study of organizational control forms,3 adapted to distinguish between “public trust” organizations, which emphasize a public image of service ideals, safeguards, and/or confidentiality, and “market” organizations, which are primarily concerned with selling a product or service and making a profit, not with establishing a reputation of honor or trustworthiness. The public trust category includes financial, professional service (medical, legal, educational, engineering, accounting), government, public transport, communications, and utilities organizations. In the market category are included mining, manufacturing, business transportation, wholesale and retail trade, and other services.
Public trust organizations have considerably higher requirements than market organizations. Organizational control type is significantly associated with overall education (γ = .40, p <.001), with white-collar education (γ = .33, p <.01), and with manager’s education (γ = .31, p <.001). There is no significant relationship between organizational control type and blue-collar educational requirements, however.
The three indicators of normative control are highly interrelated. Requirements for police records and reluctance to hire job-hoppers are significantly higher in public trust organizations than in market organizations. These interrelations tend to mutually validate the three indices as actually representing emphasis on normative control over employees. There appears to be a coherent group of organizations with relatively high educational requirements at white-collar and managerial levels, characterized by organizational goals emphasizing a public service image, carefully screening their employees for prior normative violations and for evidence of employee loyalty.
The 1967 survey of California organizations provides yet another piece of evidence to indicate the normative-importance of educational requirements. Employers were asked what actions were taken to recruit for “hard-to-fill” vacancies in the organization. For every occupational group mentioned, one solution to recruitment difficulties was overwhelmingly popular: to increase the number of recruitment channels used. Only one organization in the sample indicated as its primary response that it lowered educational requirements. Apparently, organizations rarely lower their educational requirements except perhaps under conditions of severe labor shortage. Indeed, they almost never relax any of the specifications of the kind of employees they want, even such technically extraneous qualities as sex or appearance, nor do they often dilute the content of jobs or institute training programs in times of labor shortage if they do not have them already. We may term this tendency for educational requirements, once instituted, to change only in an upward direction, the “ratchet effect.” It appears that there are strong normative and status reasons for organizations to keep up educational requirements, even when technical and economic pressures for doing so are absent or even contrary.
Business Administration Degrees
A more specialized use of education for normative control is business administration training. Surveys during the 1950s of employers in nationally prominent organizations indicated that they regarded college degrees as important in hiring potential managers, not because they were thought to ensure technical skills, but rather because they indicate “motivation” and “social experience” (Gordon and Howell, 1959: 121). Business school training is similarly regarded less as evidence of necessary training (as employers have been widely skeptical of the utility of this curriculum for most positions) than as an indication that the college student is committed to business attitudes. Employers may want business values taught in special schools because liberal arts students are oriented toward the professions and their service ideals are indifferent or even hostile to the profit-making interests of business. For example, liberal arts students are more dissatisfied with business jobs than business or engineering students (1959: 124; cf. Jacob, 1957). Furthermore, employers are more likely to refuse to hire liberal arts graduates if they come from a college that has a business school than if their college lacks a business school (Gordon and Howell, 1959: 84–87; see also, Pierson, 1959: 90–99). In the latter case, the students could be said not to have had a choice, but when both business and liberal arts courses are offered and the student chooses liberal arts, employers appear to take this as a rejection of business values.
The 1967 California survey also indicates that organizations concerned with normative control over their managers emphasize the selection of managers with business administration degrees. Overall, “public trust” organizations are more likely to require such degrees than “market” organizations. In a more detailed analysis, it appears that organizations desiring business administration students are concentrated in three industrial groupings: (a) an extremely marketing-conscious manufacturing group (food, fabricated metals, printing); (b) a public-relations-conscious transportation and utilities group (air and water transportation, gas and electric services); and (c) the administrators of large-scale service organizations (hospitals and schools). It is noteworthy that the latter are the service organizations in which administrators have the lowest status in relation to their professional staffs.4
The concern for business administration degrees, then, may be related to the status concerns of the managers themselves in organizations in which they are crucial but consider themselves underappreciated by the “production” personnel. The business administration degree also seems to be considered a loyalty test. As the personnel manager in the national headquarters of a food-processing company put it, his company wanted “business administration or other evidence that liberal arts graduates are interested in business.”
The cultural membership model may also be tested by examining the cases in which it predicts that education will be relatively important in occupational attainment. Education should be most important where two conditions hold simultaneously: (a) where the type of education most closely reflects membership in a particular status group; and (b) where that status group controls employment in particular organizations. Thus education will be most important where the fit is greatest between the culture of the status groups emerging from schools and the status group doing the hiring; it will be least important where there is the greatest disparity between the culture of the school and that of the employers.
This fit between school group culture and employer’s culture may be conceptualized as a continuum. The importance of elite education is highest where it is involved in selection of new members of organizational elites and should fade off where jobs are less elite (either lower-level jobs in these organizations or jobs in other organizations not controlled by the cultural elite). Similarly, schools that produce the most elite graduates will be most closely linked to elite occupations; schools whose graduates are less well socialized into elite culture are selected for jobs correspondingly less close to elite organizational levels.
In the United States, the schools that produce culturally elite groups, either by virtue of explicit training or by selection of students from elite backgrounds, or both, are the private prep schools at the secondary level and the elite colleges (the Ivy League and to a lesser extent the major state universities); at the professional training level, the schools that produce elite groups are those professional schools attached to the elite colleges and universities. At the secondary level, schools that produce respectively socialized, nonelite persons are the public high schools (especially those in middle-class residential areas); from the point of view of the culture of Anglo-Protestant employers, Catholic schools and all-black schools are less acceptable. At the level of higher education, Catholic and black colleges and professional schools are less elite, and commercial training schools are the least elite form of education.
In the United States, the organizations most clearly dominated by the Anglo-Protestant upper classes are the large, nationally organized business corporations and the largest firms (Domhoff, 1967: 38–62). Those organizations more likely to be dominated by members of minority ethnic cultures are the smaller, local businesses in manufacturing, construction, and retail trade; in legal practice, minority ethnic cultures are more often found in solo than in firm employment. In government employment, local governments appear to be more heavily dominated by non-Anglo ethnic groups, whereas particular branches of the national government (notably the State Department and the Treasury) are dominated by WASP elites (Domhoff, 1967: 84–114, 132–137).
Education and Law Practice
Among lawyers, the predicted differences are clear: Graduates of law schools attached to elite colleges and universities are more likely to be employed in firms, whereas graduates of Catholic or commercial law schools are more likely to be found in solo practice (Ladinsky, 1967: 222–232). The elite Wall Street law firms have been most educationally selective in this regard, choosing not only from Ivy League law schools, but also from a group whose backgrounds include attendance at elite prep schools and colleges (Smigel, 1964: 39, 73–74, 117). These same firms have been selective in other ways as well. Membership in elite Wall Street law firms and promotion to partnership within them is greatly facilitated by having a Social Register background, and women, Catholics, Jews, and blacks have been almost automatically excluded. There are also indications that graduates of ethnically dominated professional schools are most likely to practice within the ethnic community: This is clearly the case among black professionals.
Education and the Business Elite
Studies based on both biographical and survey data give a consistent picture of the American business elite. The business elite has always been highly educated compared to the rest of the populace. For example, Newcomer found that 39.4% of the top executives in 1900 had some college education, as did 51.4% of the top executives in 1925, and 75.6% in 1950 (1955: 68–69). These figures may be compared with the total percentage of the male population of ages 18 through 21 enrolled in college in the census years (1870, 1890, 1910) nearest to the time that the median age of each executive group was 20: 4.6% in 1870; 4.7% in 1890; and 6.2% in 1910. Even in the early nineteenth century, the business elite was preponderantly drawn from the best educated section of the populace, as well as from the highest social classes. Those members of the elite from lower social origins also had less education than the businessmen of upper- and upper-middle-class origins.
Approximately 55–65% of the business leaders seem to have received significant assistance from relatives in their careers, the proportion remaining constant through the last century and a half. In the 1925 and 1950 cohorts of one study of executives, 14% had inherited their companies and another 7–14% had acquired them by financial investment; interestingly, the proportion inheriting their positions rises to 27% (1925) and 23% (1950) among those with college degrees (Newcomer, 1955: 80; cf. Lipset and Bendix, 1959: 138). To put a similar finding from another study in a different way, 72% of the more recent business leaders who inherited their positions attended college, as compared with 34% of those who achieved their positions by entrepreneurship (Bendix, 1956: 230). The business elite, then, is drawn from relatively high social origins, and such origins are important in business careers. The business elite has always been highly educated in relation to the American populace, but education seems to be a correlate of their social origins rather than the determinant of their success (Taussig and Joslyn, 1932: 200; Newcomer, 1955: 76; Mills, 1963: 128). In general, the evidence that graduates of black colleges and Catholic colleges (at least before very recent years) have attained lower occupational positions in business than graduates of white, Protestant schools also bolsters this interpretation (Jencks and Riesman, 1968: 357–366; Sharp, 1970: 64–67).5
Comparative Organizational Evidence
The 1967 California survey provides the most direct test of the effects of organizational status on educational requirements. Organizational status was indicated in two ways: (a) by a “national prominence” index, based on whether the organization’s markets and its headquarters were local or national; and (b) by the size of the organization in terms of numbers of employees. Neither index is totally satisfactory as a pure indicator of organizational status, but size and national prominence are significantly associated and both relate to educational requirements in specific contexts in very similar fashion, suggesting that both indicate the same underlying variable. Size is significantly associated with overall educational requirements (γ = .29), as well as with white-collar educational requirements (γ = .28), blue-collar requirements (γ = .25), and manager’s educational requirements (γ = .26). National prominence is significantly associated with the overall educational index (γ = .23), and approaches the .05 level of statistical significance in its association with white-collar education (γ = .24), blue-collar education (γ = .17), and manager’s education (γ = .20). The 1946–1947 study of New Haven and Charlotte, N.C. (see Table 1.2) also reported higher educational requirements in the bigger, presumably more nationally organized firms (Noland and Bakke, 1949: 78).
It is possible to interpret this evidence according to the technical–functional theory of education, arguing that the elite schools provide the best technical training and that the major national organizations require the greatest degree of technical talent. What is necessary is to test simultaneously for technical and status conflict conditions. This is provided by the California employers’ study, which examines the effects of normative control emphasis and organizational prominence, while holding constant the organization’s technological modernity as measured by the kinds of technological and organizational changes in the previous 6 years. Indeed, technological change was found to be significantly related to educational requirements at managerial (γ = .13) and blue-collar (γ = .26) levels (although not for overall or white-collar requirements), thus giving some support to the technical–functional theory of education.
It is possible that the relationships thus far reported are spurious. For example, large organizations may be those with higher rates of technical change, and hence their higher educational requirements may be the result of technological needs rather than organizational prominence per se. In order to control for such possibly spurious associations, each of the major variables—organizational size, technological change, and normative control emphasis—were introduced as controls into the analysis of educational requirements. The results shown in Tables 2.1 to 2.4 indicate that each variable has an independent effect on educational requirements, but each seems to operate most strongly in the organizational context where the other variables are weak. Technological change produces significantly higher educational requirements only in smaller, localistic organizations and in organizational sectors not emphasizing normative control. Organizational size and prominence produce significantly higher educational requirements in organizations with low rates of technological change and in sectors de-emphasizing normative control. Normative control emphasis produces significantly higher educational requirements in organizations with low technological change and in smaller and less prominent organizations. It appears that a ceiling effect is operable: High educational requirements can be produced either by strong normative control emphasis, organizational prominence, or a high rate of technological change. Within contexts where one of these variables produces high requirements, the introduction of other variables does not show significant differences.
Table 2.1 Strength of Relationship between Organizational Type and Educational Requirements, with Controls for Technological Change and Size
| |
γ |
p< |
N |
| Organizational type and education index |
.40 |
.001 |
307 |
| Low technological change |
.54 |
.001 |
144 |
| High technological change |
.21 |
.01 |
148 |
| Less than 250 employees |
.50 |
.001 |
130 |
| 250–999 employees |
.24 |
.20 |
122 |
| 1000 or more employees |
.26 |
.30 |
55 |
| Organizational type and white-collar education index |
.33 |
.01 |
306 |
| Low technological change |
.49 |
.001 |
144 |
| High technological change |
.14 |
.50 |
147 |
| Less than 250 employees |
.53 |
.01 |
130 |
| 250–999 employees |
.16 |
.20 |
121 |
| 1000 or more employees |
.06 |
.70 |
55 |
| Organizational type and blue-collar education index |
.04 |
.98 |
282 |
| Low technological change |
.15 |
.80 |
128 |
| High technological change |
−.07 |
.95 |
139 |
| Less than 250 employees |
−.46 |
.70 |
120 |
| 250–999 employees |
−.26 |
.70 |
121 |
| 1000 or more employees |
.32 |
.50 |
56 |
| Organizational type and managers’ education requirements |
.31 |
.001 |
282 |
| Low technological change |
.51 |
.01 |
131 |
| High technological change |
.07 |
.05 |
136 |
| Less than 250 employees |
.53 |
.02 |
125 |
| 250–999 employees |
.08 |
.01 |
114 |
| 1000 or more employees |
.21 |
.70 |
48 |
Source: San Francisco Bay Area Employer Survey, 1967.
Table 2.2 Strength of Relationship between Size and Educational Requirements, with Controls for Technological Change and Organizational Type
| |
γ |
p< |
N |
| Size and education index |
.29 |
.001 |
307 |
| Market organizations |
.34 |
.01 |
226 |
| Public trust organizations |
−.06 |
.20 |
81 |
| Low technological change |
.44 |
.01 |
144 |
| High technological change |
.10 |
.20 |
148 |
| Size and white-collar education index |
.28 |
.001 |
306 |
| Market organizations |
.37 |
.001 |
225 |
| Public trust organizations |
−.12 |
.70 |
81 |
| Low technological change |
.40 |
.01 |
144 |
| High technological change |
.14 |
.50 |
149 |
| Size and blue-collar education index |
.25 |
.05 |
282 |
| Market organizations |
.50 |
.10 |
217 |
| Public trust organizations |
.25 |
.20 |
65 |
| Low technological change |
.49 |
.10 |
127 |
| High technological change |
.05 |
.95 |
139 |
| Size and managers’ education |
.26 |
.01 |
287 |
| Market organizations |
.33 |
.01 |
209 |
| Public trust organizations |
−.09 |
.10 |
78 |
| Low technological change |
.36 |
.10 |
137 |
| High technological change |
.16 |
.10 |
136 |
Source: San Francisco Bay Area Employer Survey, 1967.
Table 2.3 Strength of Relationships between National Prominence and Educational Requirements, with Controls for Organizational Type, Size, and Technological Change
| |
γ |
p< |
N |
| Prominence and education index |
.23 |
.05 |
287 |
| Market organizations |
.39 |
.01 |
226 |
| Public trust organizations |
.14 |
.10 |
81 |
| Less than 250 employees |
.21 |
.20 |
130 |
| 250–999 employees |
.27 |
.20 |
122 |
| 1000 or more employees |
.19 |
.30 |
55 |
| Low technological change |
.33 |
.02 |
144 |
| High technological change |
.11 |
.70 |
148 |
| Prominence and white-collar education index |
.24 |
.10 |
306 |
| Market organizations |
.39 |
.01 |
225 |
| Public trust organizations |
.01 |
.80 |
81 |
| Less than 250 employees |
.23 |
.70 |
130 |
| 250–999 employees |
.22 |
.70 |
121 |
| 1000 or more employees |
.28 |
.50 |
55 |
| Low technological change |
.30 |
.10 |
144 |
| High technological change |
.17 |
.30 |
147 |
| Prominence and blue-collar education index |
.17 |
.10 |
282 |
| Market organizations |
.22 |
.10 |
217 |
| Public trust organizations |
−.06 |
.30 |
65 |
| Less than 250 employees |
.15 |
.20 |
120 |
| 250–999 employees |
.24 |
.05 |
111 |
| 1000 or more employees |
.02 |
.10 |
41 |
| Low technological change |
.28 |
.20 |
127 |
| High technological change |
.13 |
.20 |
139 |
| Prominence and managers’ education |
.20 |
.10 |
282 |
| Market organizations |
.35 |
.05 |
209 |
| Public trust organizations |
−.09 |
.70 |
78 |
| Less than 250 employees |
.24 |
.30 |
125 |
| 250–999 employees |
.27 |
.10 |
114 |
| 1000 or more employees |
−.02 |
.30 |
47 |
| Low technological change |
.31 |
.20 |
137 |
| High technological change |
.09 |
.50 |
136 |
Source: San Francisco Bay Area Employer Survey, 1967.
Table 2.4 Strength of Relationships between Technological Change Index and Educational Requirements, with Controls for Organizational Type and Size
| |
γ |
p< |
N |
| Technological change index and education index |
.12 |
.20 |
292 |
| Market organizations |
.23 |
.02 |
214 |
| Public trust organizations |
−.14 |
.30 |
78 |
| Less than 250 employees |
.21 |
.10 |
123 |
| 250–999 employees |
−.01 |
.20 |
117 |
| 1000 or more employees |
−.19 |
.50 |
52 |
| Technological change index and white-collar education index |
.12 |
.20 |
291 |
| Market organizations |
.22 |
.01 |
213 |
| Public trust organizations |
−.12 |
.50 |
78 |
| Less than 250 employees |
.20 |
.05 |
123 |
| 250–999 employees |
.04 |
.50 |
116 |
| 1000 or more employees |
−.16 |
.50 |
52 |
| Technological change index and blue-collar education index |
.26 |
.02 |
267 |
| Market organizations |
.30 |
.10 |
205 |
| Public trust organizations |
.14 |
.50 |
62 |
| Less than 250 employees |
.47 |
.10 |
113 |
| 250–999 employees |
.14 |
.20 |
106 |
| 1000 or more employees |
.03 |
.80 |
48 |
| Technological change index and managers’ educational requirements |
.13 |
.001 |
272 |
| Market organizations |
.24 |
.001 |
198 |
| Public trust organizations |
−.14 |
.70 |
75 |
| Less than 250 employees |
.14 |
.01 |
118 |
| 250–999 employees |
.11 |
.05 |
105 |
| 1000 or more employees |
−.06 |
.50 |
46 |
| Technological change index and professionals’ educational requirements |
.14 |
.98 |
225 |
| Market organizations |
.16 |
.70 |
157 |
| Public trust organizations |
.13 |
.98 |
70 |
Source: San Francisco Bay Area Employer Survey, 1967.
The relative effects of normative control, organizational prominence, and technological change may be assessed by comparing the measures of association in the context in which each is strongest (i.e., within the realm where the other organizational variables are weak). At all levels except blue-collar requirements (which are especially influenced by organizational size and technology), the strongest determinant of educational requirements is organizational control type, followed by organizational prominence (measured either by organizational size or national prominence), with technological change producing the weakest effects.6 The relative weakness of technological change in explaining upper-level educational requirements is in keeping with the findings of a study of top business executives in national business, which indicated that the most highly educated managers were found in the least economically vigorous firms rather than in the most rapidly developing ones (Warner and Abegglen, 1955: 141–143, 148).
Finally, let us consider the nature of career channels within organizations. For although the prevailing rhetoric of meritocracy emphasizes the openness of career possibilities, the reality is closer to a castelike separation among major occupational blocs.
The California employers’ survey is especially revealing on this score. Table 2.5 indicates the most generally used recruitment channels for the various occupational levels. Manual laborers are recruited primarily through unions, clerical workers primarily through employment agencies. The only occupational group recruited in any numbers through schools and colleges is professionals. The highest concentration in one channel, however, is for managers: 73% of employers in this sample indicated that they ordinarily promote their managers from within.
Table 2.5 Percentage of Organizations Relying Primarily on Various Recruitment Channels, by Occupational Group
Source: San Francisco Bay Area Employer Survey, 1967.
There is a certain ideological tone about this claim. It is part of a general philosophy expressed by one personnel manager in a cliché delivered with ringing conviction: “There are no dead-end jobs, there are only dead-end people!” Numerous respondents commented in this vein. “There are no artificial barriers to advancement in this company. A man can go as far as his talents will take him.” Like most ideologies, this kind of statement has some connection with reality; but like most ideologies, it also idealizes and overgeneralizes that reality.
If an employer were seriously interested in promoting from within, one might expect that he would have low educational requirements for managerial positions. However, there is no statistically significant difference in educational requirements for managers between organizations that recruit managers primarily from within and those that do not. Even organizations recruiting primarily from within have educational requirements for managers that are a good deal higher than for any other occupational groups except professionals (see Table 1.2). Where, then, do they recruit their managers from? Higher managerial positions are no doubt filled from lower managerial positions, but where do the lower-level managers come from? Some personnel managers stated that they hired managers from within as a general policy, but they then added that the present managers were probably not replaceable from below. Like these organizations, numerous companies meet the problem by actually hiring from outside. Several personnel managers stated that college degrees were required of such outside recruits but not of managers promoted from below.
It can be inferred that managers rarely come from the manual level. When employers speak of every job as being “promotable,” they mean that every entry job has some other job above it to which a worker may aspire. They do not usually mean to imply (except in the vaguest, most ideological sense) that there is a clear channel of promotion from the bottom to the top. There is almost always a promotion ladder from unskilled and service positions to skilled and sometimes foreman positions, and employers in interviews often volunteer comments about how much such upgrading occurred. Nevertheless, no prideful instances were ever cited about movement from manual into nonmanual positions, and some casual comments indicated that employers thought of manual and nonmanual (and clerical) categories as hermetically sealed off from each other, although the opportunities were wide open within each.
Somewhat more systemative evidence confirms this impression. When naming positions customarily filled by upgrading, employers mention managerial positions in relative frequency of rank (see Table 2.6). High-level managerial positions were most often mentioned, but the lowest-level managerial positions were mentioned relatively infrequently. Similarly, upper-level clerical jobs were the most mentioned as filled from within; lower-level clerical jobs the least; and the same pattern may be found within the skilled labor group.
Table 2.6 Percentage of Jobs Customarily Filled by Promotion from Within
| Percentage mentioning particular classifications |
| Managers and officials |
54.7 |
| Managers, officials, administrators |
21.6 |
| Department managers or department heads |
10.4 |
| Assistant department managers or assistant heads |
2.9 |
| Store managers |
2.3 |
| Assistant store managers |
1.6 |
| Supervisors or supervisory jobs |
9.4 |
| First-line supervisors |
4.9 |
| Other managerial occupations |
1.6 |
| Professional and technical occupations |
11.1 |
| Instructors or teachers |
.3 |
| Registered nurses |
.7 |
| Engineers, chemists |
2.0 |
| Accountants |
1.0 |
| Draftsmen |
.3 |
| Engineering technicians |
.3 |
| Laboratory technicians and assistants |
.7 |
| Other professional and technical occupations |
5.8 |
| Sales occupations |
7.2 |
| Salesmen |
6.9 |
| Junior salesmen |
.3 |
| Clerical occupations |
16.9 |
| Upper-level clerical or top clerical jobs |
12.0 |
| Clerical jobs or office jobs |
2.6 |
| Secretaries |
1.3 |
| Stock clerks and shipping clerks |
.3 |
| Other clerical occupations |
.7 |
| Skilled occupations |
19.9 |
| Foremen, leadmen |
10.5 |
| Dispatchers |
1.3 |
| Journeymen |
2.9 |
| Other skilled occupations |
5.2 |
| Semiskilled occupations |
8.1 |
| Truck drivers |
.3 |
| Apprentices in skilled trades |
.3 |
| Machine operators |
2.0 |
| Assemblers |
.3 |
| Pressmen, cutters, adjusters |
.7 |
| Inspectors |
.3 |
| Other semiskilled occupations |
4.2 |
| Unskilled occupations |
1.0 |
| Lube, and wash men (garage) |
.3 |
| Other unskilled occupations |
.7 |
| Service occupations |
5.9 |
| Head custodians |
1.9 |
| Policemen and firemen (above entry level) |
.7 |
| Waiters and waitresses |
.3 |
| Cooks |
1.0 |
| Nurses’ aides and orderlies |
.3 |
| Gardeners and groundsmen |
.7 |
| Busboys, etc. |
.3 |
| Other service occupations |
.7 |
| All jobs customarily filled by upgrading |
19.9 |
| Almost all jobs customarily filled by upgrading |
21.8 |
| No jobs customarily filled by upgrading |
2.6 |
| Specific jobs or classifications mentioned |
55.7 |
| Total (%) |
100 |
| Number surveyed |
307 |
Source: Gordon and Thai-Larsen (1969, Table 8.11).
If one compares the educational requirements for managers with those for skilled workers, it is clear that few skilled workers can hope to be promoted into the managerial ranks, at least not without quitting and getting an education before returning. One personnel manager epitomized this situation by stating, after emphasizing the virtues of promotion from within: “We never hire people for jobs who can’t be upgraded. All our file clerks are trained to become secretaries.”
It is also rarely the case that clerical workers are promoted to managerial positions. The caste system in this case is based on sex: Clerical workers are mostly women, and women are rarely made managers outside of supervisory positions in clerical sectors. There is a clear demarcation between the educational requirements for managers and clerical employees: High school is the modal requirement for the latter, college graduation for the former.7
The primary recruitment channels within organizations for managers are sales and professional positions. Several personnel managers stated that there were no entry jobs for managers and volunteered the information that managers were likely to come in as professionals. It is apparent that employers have a much broader view of “professional” jobs than professional associations do; 25% of employers indicated less than a college degree as the educational requirement for professionals and 10% indicated less than a high school degree (Table 2.2). Similarly, 10% of the employers indicated that they recruit their professionals from within (Table 2.5). It is likely that many employers consider any staff or technical job as “professional.”8 Many studies have shown that there is considerable movement back and forth between staff and line positions and, indeed, between professional and managerial positions.
When employers recruit managers from within, then they mean primarily that they recruit from other high-level positions in the organization and from professional positions. Moreover, it is for “professional” positions, whatever the actual content of the job, that the “recruitment from within” ideology is generally not held to apply. Organizations that recruit their managers from within (very likely from professional ranks) are just as likely as other organizations to have high educational requirements for professionals. Some 74% of the former require college degrees or higher of their managers, as compared to 78% of the latter, a difference that is not statistically significant.
Educational requirements act as a formidable barrier to promotion across the manual–nonmanual caste barrier in organizations. (The clerical–managerial barrier seems less based on educational requirements, partly because it is so noncontroversial; rather, it is based quite blatantly on the ascriptive quality of sex.) The “promotion from within” ideology is used to cover up this barrier, and the ideology is even carried out in practice, but within limits. Through such organizational subterfuges as promotion within the professional group, educational requirements are kept as a major dividing line between managers and manual laborers. One personnel manager of a large, prominent bank commented rhetorically:
The route up from janitor to Senior Vice President is now closed. To get anywhere, an education is now needed so we simply cannot hire out of high school any longer. This is serious for young men who will not continue on with their educations.
Looking back over this mass of evidence, what does it show? Clearly there is a substantially negative verdict on the meritocratic theory of modern society and on the technological–functional theory of education. We do find that within the manufacturing and related material–distribution sectors of the economy, there is some tendency for technological advance to be associated with higher educational requirements, especially for blue-collar jobs. But the bulk of the evidence is in the other direction. Education is not associated with employee productivity on the individual level, and job skills are learned mainly through opportunities to practice them, as retraining procedures for organizational innovations abundantly demonstrate. Social mobility surveys and observational studies of work both show that pressures for technical efficiency are submerged within the social struggles over positions, struggles in which membership in a cultural group is the crucial weapon.
It is not surprising, then, that educational credentials are most heavily emphasized within organizations stressing normative control—that is, cultural socialization—and in the large national bureaucracies; nor is it surprising that these are stronger determinants than technological change. This comparative organizational evidence has both a general theoretical significance and a historical one. For the variable—large national bureaucracy versus smaller local establishment—is essentially the distinction between the organizations controlled by the Anglo-Protestant upper class and those controlled by Catholic and other minorities. Crosscutting this is the second variable—“public trust” versus “market” organizations—which refers basically (if not altogether precisely) to the contrast between the tertiary and secondary sectors of the economy. The evidence here meshes with the historical analysis that we will find in subsequent chapters, which stresses the dual themes of the ethnic organization of the American economy and the effects of super-productive twentieth-century technology in creating a large “sinecure sector.”
