THE COMPETITIVE SEMICONDUCTOR MANUFACTURING HUMAN RESOURCES PROJECT:

Second Interim Report
CSM-32
Clair Brown, Editor

9. The Evolution of Skill Demand and the Nature of the Employment Relationship in a Technology-Intensive Firm
Vincent M. Valvano

9.1 Introduction
9.2 Changes in the Structure of Employment at NewTech
9.3 The Organization of Work in a Technology Firm
9.4 Earnings Determination at NewTech
9.5 Career Length and Turnover Patterns
9.6 Summary

9.1 Introduction

Surprisingly little empirical work exists in economics on the internal organization of firms. Recent developments in economic theory have brought attention to questions concerning the nature of the employment relation, work organization, and incentives inside the firm. But theory has outpaced empirical analysis of these issues, reflecting, in part, the difficulty in gaining access to appropriate firm-based data. Given the small number of existing studies based on personnel data, it is far from resolved whether or not existing theories of the employment relation are capable of explaining the important features of observed firm compensation and promotion systems. One aim of this focus study is to contribute to the empirical literature on the internal organization of the firm, using longitudinal personnel data from a large technology firm which is identified by the pseudonym "NewTech." Data on the compensation and careers of employees at NewTech are used to test predictions of several models of the employment relation.

A second aim of the study is to consider the implications of knowledge-intensive work for internal organization and labor market dynamics. The organization of knowledge workers (in particular, engineers and scientists) remains a relatively neglected area of inquiry in both labor economics and the economics of organization. In the standard model of corporate hierarchy, higher-ranking employees specialize in decision-making while those in lower ranks focus on information-processing or execution. A defining characteristic of knowledge work is the rejoining of decision-making and problem-solving activities with information-processing and execution across a wide range of jobs in the firm. Of interest is whether this integration of decision-making and execution has implications for the firm's hierarchy, and consequently, for the observed career and pay patterns of knowledge workers in the firm. For example, one might predict the existence of a relatively flat hierarchy in such settings, with fewer managers at upper levels and larger spans of control.

An additional important characteristic of knowledge workers derives from the nature of their skills and skill acquisition. Such workers acquire a large amount of specialized training (in the form of technical education credentials) prior to the start of a job, and they continue to rely on sources outside of the firm (journals, conferences, etc.) to maintain and update their specialized knowledge. In the language of human capital theory, the general or industry-specific component of these workers' human capital is large relative to the firm-specific component (acquired through on-the-job experience and training). If general or industry-specific human capital is relatively more important for this class of employees, we might predict differences in observed career and turnover patterns relative to employees whose human capital and training is mainly acquired on the job and is firm-specific.

These and other issues are being investigated with personnel data from NewTech, a large technology-intensive firm based in the United States but with extensive operations overseas. An annual year-end record for each employee in the principal business unit of the firm is available for the years 1976-1994. During the sample period, employees in the business unit accounted for 40-50% of total employment in the firm (see Figure 9-1); at the beginning and end of the sample period the ratio exceeded 50% while during the 1980s it remained closer to 40%.

The chapter proceeds as follows. In Section 9.2, changes in the structure of employment at NewTech during the sample period are presented. Such changes give an initial indication of how skill demands have evolved at NewTech. Section 9.3 discusses the organization of work in technology firms generally, with reference to NewTech. The wage structure of the firm is discussed in Section 9.4. Career and turnover patterns of managers and professionals at NewTech are analyzed in Section 9.5. Section 9.6 concludes.

9.2 Changes in the Structure of Employment at NewTech

The occupational structure at NewTech changed significantly during the sample period, as employment shifted towards higher-skill professional occupations. Table 9-1 presents worldwide employment by major occupation category in 1985 and 1994. These two years are interesting comparison points because employment at NewTech became much less cyclical beginning in 1985 and remained relatively stable during the 1985-1994 period (Figure 9-1). Between these years, professional occupations' share of total employment in the firm grew by 50%, while the shares of managers and technicians both declined. The employment share of operatives, which are the largest occupation category at NewTech, remained stable at 40%. Because total employment in the firm declined by 4% during this period, only the employment level of professionals increased in absolute terms between 1985 and 1994.

Table 9-1. Worldwide Employment Levels and Shares by Occupation at NewTech

⁺Other includes sales, skilled crafts, and service workers.

The occupational distribution of NewTech employment in the United States exhibits a similar shift toward higher-skill professionals between 1985 and 1994 (Table 9-2). The share of professional employees in the U.S. grew by 70% during the period, while manager and technician shares both declined and the operative share of employment remained stable. The actual level of total firm employment based in the U.S. declined by 2% between 1985 and 1994, which was less than the 4% drop in worldwide employment. Hence, as the bottom row in Table 9-2 indicates, the U.S. share of total employment at NewTech increased slightly over the period.

Table 9-2. U.S. Employment Levels and Shares by Occupation at NewTech

⁺Other includes sales, skilled crafts, and service workers.

NewTech employment in the United States is relatively more skill intensive than worldwide employment. Professionals and technicians account for larger shares of employment in the U.S. than overall at NewTech, while the share of operatives based in the U.S. is much lower than worldwide. During the 1985-1994 period, the share of NewTech professionals based in the U.S. increased while the U.S. share of technician and operative jobs remained stable. In 1985, 46% of NewTech professionals were based in the U.S., increasing to 53% by 1994. The proportion of technicians and operatives based in the U.S. remained at 55% and 17%, respectively, during the period.

U.S. employment levels and shares are also reported for 1976 in Table 9-2. Trends in the occupational distribution differ in the 1976-1985 period when compared with the later period. The shares and absolute levels of managers, professionals, and technicians increased substantially. But the share of operatives in U.S. employment declined by 37% between 1976 and 1985. Because the gains in the growing occupations were not large enough to offset this drop in operative employment, total U.S. employment at NewTech decreased by 18% between 1976 and 1985.

Finally, the proportion of managers in the NewTech workforce appears to be significantly lower than the shares reported for firms in other industries. At NewTech, the share of managers in worldwide employment was about 10% between 1985 and 1994 while the U.S. share of managers declined from 13% to 10%. These are strikingly smaller shares than those reported elsewhere. For example, in the services firm analyzed by Baker et al. (1994a), management constituted a stable 20% of the workforce over the period 1969-1988. At post-divestiture AT&T, the managerial share of employment increased from 29.4% in 1984 to 45.7% in 1990 (Batt, 1995). This comparison supports the idea that technology firms rely less than other firms on managerial employees whose principal job is to supervise other employees and relay decisions along a vertical chain of command.

This initial description of the structure of employment at NewTech raises a number of questions. Overall employment trends during the sample period present one puzzle. In the period prior to 1985, employment in the principal business unit displays a cyclical character, but in the years since 1985 employment has remained quite stable (Figure 9-1). Correlates of changes in the employment shares of particular groups will be analyzed in more detail. For example, the secular decline in operative employment might be attributed to cyclical changes in product demand, or to contracting out of work previously done inside the firm, or to skill-biased technical change. The significant increase in professional employment during the sample period may be concentrated in particular occupations and may be related to changes in technology or in the product mix of the firm. Changes in work organization or technology may also underlie the change in the ratio of professionals to technicians, from approximately 1:1.6 in 1985 to 1:1 by 1994. Analysis of these issues is currently underway.

9.3 The Organization of Work in a Technology Firm

Several studies have suggested that hierarchical structure and career patterns in technology-intensive firms are fundamentally different from those in other firms. The internal structure of such firms is held to have departed from the classic m-form model characterized by "tall hierarchies, vertical communication orientations, a functional bias in organizing activities, relative centralization, and slowly changing technology" (Kanter, 1984, p. 109-10). Instead, decision making has been decentralized through the use of multi-dimensional reporting relationships (matrix organization) and reliance on project teams and task forces. Many career moves of both professionals and managers are "ad hoc responses to current needs" rather than based on prescribed ladders or mobility chains (Kanter, 1984, p. 121). Bailyn (1991) suggests that professional careers may be characterized by four distinct career moves: 1) a switch to a managerial ladder, 2) progression through a series of professional jobs on a technical ladder, 3) project-based moves, or 4) a technical transfer in which the professional moves with a process technology from the development lab to the manufacturing facility. This quasi-formal career structure seemingly offers as many horizontal as vertical moves. If such career patterns are important, they may have implications for the form of incentive provision, the nature of skill development and/or the length of employment at the firm.

The relative importance of various career paths for technical professionals, and their affects on pay trajectories can be measured directly. NewTech organizes work within each business unit on the basis of cost centers, departments, divisions, and groups (in ascending order of scope). To the extent that cost centers in the firm define projects, they may be particularly important in technology-intensive firms for defining careers and earnings. For example, Ferguson and Morris (1994) propose a "Silicon Valley" model of firm organization in which modular satellite groups work in parallel with a core design and development group on various system components sharing a common architecture. These groups form and dismantle depending on the developmental needs of the architecture and its success or failure in the market. This model implies that careers are more likely to be based on a series of moves among projects rather than on a vertically-oriented job hierarchy, and that moves between projects may also entail moves between firms.

Although little work on these issues has been done in economics, an ongoing concern of the behavioral social science literature has been that the promotion systems of large firms are maladapted to managing technical human assets. Two adverse consequences are identified. First, it is allegedly difficult for professionals to maintain advances in terms of pay and perquisites without shifting onto a managerial career ladder. The firm is thus at risk of misallocating or losing technical talent. "Dual career ladders" or "technical ladders" are proposed as a response to this condition but there is no consensus that they have effectively equalized differences between ladders when established (Bailyn, 1991; Mohrman and Von Glinow, 1990). Second, the internal incentive systems of firms are alleged to be ill-equipped to deal with the more rapid obsolescence of the skill base of professionals in a technology-intensive environment. For example, Bailyn (1991) suggests that the average half-life of technical knowledge is 5 years.

Several implications follow from these potential costs of internal organization. For example, given rapid skill obsolescence, the allocation of senior professionals into management may represent an efficient allocation of labor if technical professionals are as capable of managing as other employees (an assertion that is often debated). Roberts and Biddle (1994) argue that the transition of technical professionals into management is efficient for the firm for a different reason high-performing professionals turn out to be high-performing managers as well. Using personnel data from a medium-sized defense contractor, they find that "good" technical employees (as defined by annual performance evaluations) tend to become "good" managers. Whether such results can be generalized to other technology-intensive industries remains to be demonstrated. Selection bias may also be a factor in such results. Bureaucratic formalism is said to prevail in aerospace firms to a much greater extent than in the newer high-tech industries. Formal criteria such as seniority, education, and supervisor evaluations are primary determinants of promotion; job descriptions, lines of authority, and advancement criteria are explicit and standardized in the more mature aerospace industry (Robinson and McIlwee, 1991). The resulting "corporate culture" is likely to generate different incentives and behavior than the "quasi-formal structure" said to characterize the newer technology industries.

If the incidence of skill obsolescence is higher in technology-intensive firms, such firms may have additional incentives to provide training or training subsidies to their employees. NewTech provides a range of training opportunities for its employees. The effect of observed training on the subsequent performance and earnings of NewTech employees can be estimated. Such estimates can provide evidence of the degree to which skill obsolescence is an issue in technology-intensive firms.

It is also unclear a priori why a technical job ladder can not overcome the incentive and allocation problems of a promotion system based on a one-track managerial ladder. One potential explanation is that administered pay systems give considerable weight to the number of employees supervised when assigning salary levels to jobs (Bailyn, 1985). Paying technical professionals who lack supervisorial responsibilities comparable upper-management-level salaries may be contested within the firm on merit grounds. A direct comparison will be made of subsequent earnings for professionals who do and do not switch to a managerial job, controlling for education, experience, and performance. The size of any "earnings premium" for those who switch to a managerial track can thereby be estimated. The distribution of managers and professionals across grades at NewTech (Figure 9-2) indicates that a "technical ladder" extends about two-thirds of the way up the grade hierarchy. The top six grades are populated exclusively by managers who account for less than 1% of the total population of (U.S.) professionals and managers.

An initial analysis of the movement of employees between professional and managerial jobs is presented in Table 9-3. Such movement is sizeable and the direction of the flow is evidently two-way. The distribution of job spells per employee is reported for the pooled sample of managerial and professional employees during 1983-1994. Panel A presents the distribution of spells for employees who entered the firm in a managerial job title, where a spell is defined as a continuous period of time in one job title. Over 17% of these employees experienced at least one spell in a professional job during their tenure at the firm. Panel B indicates that about 9% of employees who entered the firm in a professional job experienced one or more managerial job spells. In absolute magnitude, the flow of personnel from professional to managerial job titles is larger, given the larger population of professionals in the firm. Nonetheless, the significant number of managers who experience professional job spells implies that many managers have technical backgrounds and that technical professionals serve for defined lengths of times as project or department managers, much in the manner of an academic who may serve a term as the department chair.

An additional measure of the importance of the professional-to-managerial transition in the firm can be derived from examining the sources of entry into the population of managerial job titles. I have analyzed entry into the 50 most important managerial job titles, representing 86% of managerial employee-years, over the period 1983-1994. In this data set, entry into a managerial job can occur through a new hire, a transfer from another business unit in the firm, a move from a non-exempt job title, or a move from a professional job title. Professionals are clearly the most important source of entry into managerial titles, accounting for 53% of all entry into managerial jobs during the period. New hires comprised 33% of entry, while the combined category of business unit transfers and moves from non-exempt jobs accounted for 14% of entry into managerial titles.

Go to Next Section of this Chapter
Go to Table of Contents for this Chapter
Go to Table of Contents for the CSM-HR Interim Report