Division of Research Graduate School of Business Administration The University of Michigan April 1984 WHITE COLLAR HUMAN RESOURCE MANAGEMENT IN LARGE JAPANESE MANUFACTURING FIRMS Working Paper No. 374 Vladimir Pucik The University of Michigan FOR DISCUSSION PURPOSES ONLY None of this material is to be quoted or reproduced without the expressed permission of the Division of Research.

INTRODUCTION This paper focuses on white collar Human Resource Management practices in large Japanese manufacturing firms, filling the gap in the literature on Japanese organizations. According to numerous accounts of Japanese management, firms in Japan manage their human resources from the perspective that "people are the most important assets". However, very little has been written up to now about specific Human Resource Management practices, especially as applied towards white collar employees. In particular, past research on personnel policies in Japanese firms was focused mainly on blue collar workers (e.g. Cole, 1971; Dore, 1973; Marsh & Mannari, 1976), with only limited attention given to white collar and managerial employees. Also, available analyses of white-collar personnel management deal mainly with the service industry (Rohlen, 1974), and with the public sector (Skinner, 1983; Noguchi, 1983). In contrast, no systematic treatment of the white collar HRM functions in large industrial organizations has been attempted so far. It may be that the large industrial firms in Japan are not representative of a typical Japanese organization. Indeed, a majority of Japanese white collar employees work in small and medium size firms (Clark, 1979). In that respect, many aspects of the HRM systems described below are observed in large organizations only. Yet these large Japanese firms are precisely those which today pose the greatest competitive challenge to many American industries. A better knowledge of their internal HRM 2

practices is essential to a better understanding of their external competitive strategies. JAPANESE ORGANIZATIONS AS INTERNAL LABOR MARKETS At present, in most large Japanese firms (defined as those with over 1000 employees), hiring employees with previous work experience is unusual. Essentially, there are only two ports of entry into a large Japanese firm matched with two categories of recruits. The first port of entry is designed for new high school graduates, leading primarily to blue collar or "low-ranking" technical jobs; the second is for graduates freshly out of college, slated primarily for managerial-level jobs in both technical and administrative fields. Nearly all promotions to management jobs are from within the organization. This system is characterized in economic literature as an Internal Labor Market (Doeringer and Piore, 1971, Cole, 1979). The employees are hired, under the implicit assumption that they will remain with the company until they reach the compulsory retirement age, usually between the ages of 55 and 60. During most of their careers, employees are granted annual seniority pay increases, in addition to salary increases based on merit. Layoffs are exceptional and the maintenance of so-called "life-time employment" (a somewhat incorrect but widely used term) are some of the principal -- and publicly acknowledged -- concerns and practices of the corporate management. The origins of this system of employment are relatively recent. It was not until the 1920's that, in response primarily to 3

economic factors, the first Internal Labor Market features emerged as a distinct pattern in the employment practices of some large Japanese corporations. And it was only after World War II that the consciously developed ILM structure began to dominate Japanese personnel practices, again fostered by the economic environment (Clark, 1979). This may seem a point of limited interest to all but economic historians. However, this also implies that white collar HRM practices in Japan -- as observed today in large industrial firms -- are less a derivative of their cultural environment, and more a result of a deliberate organizational choice. Internal Labor Markets and Organizational Hierarchy Japanese white collar personnel practices are closely tied to the complex structure of most large corporations in Japan, even though the actual reporting lines may, at the same time, be very simple. A number of organizations utilize a so-called dual career system, one for status ranks (shikaku), the other for job titles (yakushoku). A convenient analogy to such a structure would be a typical military organization. Promotion to a given rank opens eligibility for a range of job titles, but the match between ranks and titles is often imperfect. However, there are firms in Japan that use only one hierarchical ladder, and also those that use three complementary ladders, adding position class (shokumokyu) or salary class (tokyu). Often the terms used to label individual grades within the rungs of job hierarchies are not exclusive. For example, the term kacho 4

(Manager) is used frequently to describe a specific status rank as well as a position class or a job title, yet not all employees with the kacho status are assigned to kacho positions. To complicate the matter even further, each company develops its own vocabulary with its own specific order of ranks and positions. As a result, the Internal Labor Market is reinforced by a unique set of customary organizational practices. To present a better picture of how this system actually works, the organizational hierarchy and the basic HRM functions of large Japanese manufacturing firms will be discussed in the context of data collected in two corporations, both are among the largest manufacturing firms in Japan with respect to sales volume, asset base and the number of employees. The first company, ELCO (electric machinery firm), was established in the early 1920's by a spin-off of two facilities manufacturing electric equipment from a major shipbuilding concern. In 1960, ELCO employed approximately 30,000 employees, 1,400 of whom were college graduates. Sales reached the level of about $300 million ($1/360yen). By 1970, the number of employees increased twofold to nearly 55,000, and the number of employees with college degrees increased threefold to 4,500. Sales volume jumped to $1.2 billion ($1/360yen), an increase of 400 percent. However, during the 1970's ELCO's growth began to slow down. By 1980, sales volume reached $4 billion -- partly inflated by adjustments in exchange rates ($1/230yen). The number of employees actually decreased to little over 50,000, but the number of college graduates continued to rise to about 8,500. 5

The second firm -- STEELCO -- is in the steel and engineering/ shipbuilding business with origins going back to the World War I period. The first STEELCO mill was designed to produce steel for a nearby major shipyard from pig iron transported from another location. In the late 1950's the Steel Division accounted for 70% of $200 million of total corporate sales, with shipbuilding providing the rest. By 1970, the share of steel increased to about 80% (out of over $1.5 billion), with the rest divided equally between shipbuilding and heavy engineering. By 1980, shipbuilding and engineering were merged, but their combined share of company sales slipped further to 15% out of the total of over $ 5 billion. During the same period, the number of employees increased from approximately 15,000 to more than 35,000 (which was, however, 20% below the employment peak of 1974). The distribution of employees across divisions was approximately proportionate to sales. ELCO is one of the firms with a triple hierarchical structure. For white-collar employees who are the focus of this study, there are seven skikaku ranks: the entry level nikkyu shuin (Professional II), ikkyu shuin (Professional I), shuji (Assistant Manager), shukan (Manager), sanji (General Manager), sanjo (Junior Executive) and riji (Senior Executive). The youngest Junior Executive in the company at the time of this study was 47 years old, the youngest General Manager was 42, the youngest Manager was 36, and the youngest Assistant Manager was 32. Job titles (yakushoku) are given only to employees in the ranks of shuji and up. Forty-nine different job titles were used at the 6

time of the study. The position (shokumokyu) hierarchy for regular white collar employees has 10 main position classes (24 other classes are reserved for special assignments or other employee categories): Unclassified Positions 1-6, Specialist, Supervisor (kakaricho, Manager (kacho) and Senior Manager (bucho). Table 1 presents the distribution of ranks and positions for ELCO's white-collar college graduates in the three middle skikaku ranks. TABLE 1 - ELCO: STATUS RANKS AND POSITION CLASSES RANKS POSITIONS GEN.MANAGER MANAGER ASS.MANEGER SENIOR MANAGER 336 142 MANAGER 15 1047 83 SPECIALIST - 196 SUPERVISOR - -941 UNCLASSIFIED - - 88 In STEELCO, the hierarchical structure is substantially simpler. There are seven shikaku ranks for white-collar employees: General Manager (bucho), equal to the rank of sanji in ELCO, Senior Manager (jicho), Manager (kacho), Supervisor (kakaricho), Junior Supervisor (shukan), Professional I (shuji I) and Professional II (shuji II). (The variations in meaning of the original terms between ELCO and STEELCO are striking.) The top-level executives are unclassified. The age pattern was similar to that observed in ELCO. The youngest General Manager was 48 years old, the youngest Senior Manager was 43, the youngest Manager was 36, and the youngest Supervisor was 32. The rank distribution by Business Groups is 7

presented in Table 2. The position classes (in STEELCO called shokunokyu), are, for all practical purposes, identical with the status ranks. The use of job titles is restricted, unlike at ELCO. Six titles are reserved for those with line responsibilities (plant manager, department manager, group manager, area manager, team supervisor, foreman), two others are for special assignments, but the majority of white-collar employees of managerial ranks are classified as shunin buin (Senior Staff). TABLE 2 - STEELCO: STATUS RANKS BY BUSINESS GROUPS STEEL ENGINEERING TOTAL GENERAL MANAGER 92/86* 49/44 141/130 SENIOR MANAGER 243/209 174/139 417/348 MANAGER 794/580 522/353 1316/933 SUPERVISOR 978/438 668/309 1646/747 JUNIOR SUPERVISOR 910/530 659/379 1569/909 PROFESSIONAL I/II 1319/588 1237/389 2556/977 *all white-collar/college graduates First of all, the data indicate that promotion opportunities for white collar employees in the Steel Business Group are better than the opportunities for those employed in Engineering. At the lower hierarchical levels, the number of employees is nearly equally distributed between the two sectors, but the distribution ratio is approximately 3:2 for middle managers, and 2:1 for junior executives. In other words, opportunities for advancement are twice as good in Steel than in Engineering. This is caused by two factors: a relative decline of shipbuilding/engineering activities in general, and the lower white-collar labor intensity of steel-making. In that sense, we can observe a dual Internal 8

Labor Market within a single organization. Even though the hiearchical structure is the same in both business groups, promotion opportunities vary. Table 2 illustrates one more important factor concerning career achievement in Japanese organizations; only very few employees who do not have a college degree secure promotion to upper level managerial ranks. Those who succeed are generally graduates from special professional schools (often military schools). However, the rank of Manager is the highest rank that a high school graduate can reasonably expect to reach. It should be pointed out that in comparison to the recent high school graduates, those who joined large Japanese firms in the immediate post-war years had much better chances to reach the upper levels of middle-management. Although their promotion speed was generally slower than for employees who graduated from college, many were promoted to middle management positions, benefitting from the rapid growth of most large Japanese firms during the twenty years following the Korean war. The slowdown in economic growth during the early 1970's changed this picture dramatically. Just as the cohorts of employees who entered employment in the late 1950's or early 1960's reached the seniority necessary for promotion to first-level middle management jobs, positions suddenly became more scarce. As a result, many companies introduced various measures to tighten promotion rules, either directly by lowering the promotion quota, or indirectly by devices such as written management exams. The 9

vast majority of high school graduates do not qualify under such a system and are eliminated from the competition for all but first level managerial positions. For example, in STEELCO, before being considered for promotion to the position of Supervisor, an employee must pass a written exam testing knowledge of functional area and business in general, basic communication skills, and even a comprehension of English. Any employee of the firm can have up to three tries to pass the test, but only after at least eight years of work experience in the case of college graduates or after twelve years for all others. In 1975, 80 percent of college graduates who took the test passed, and 92 percent of them were immediately promoted to the position of Supervisor. Among high-school graduates the respective percentages were 34 and 50. In 1980, with a more rigorous exam, only 74 percent of college graduates passed, and only 45 percent of them were promoted. Among the high school graduates, only 14 percent passed, and only 28 percent of them -- 4 percent of all high school graduates who took the test (and many did not even try) -- were actually promoted. The STEELCO executives defended the exclusion of high school graduates from consideration for upper level managerial positions by arguing that nearly all who are potentially qualified to be managers take advantage of the opportunity to go to college and therefore a college degree can be "objectively" used as a first "discriminating" factor in determining promotion potential. In fact, national educational statistics indicate that in 1960 only 17 percent of male high school graduates went on to college, but 10

by 1978 this share increased to 36 percent (Bowman, 1981). It should be added, however, that corporate policies excluding high school graduates from consideration for higher managerial positions are still strictly "unofficial" and most companies are firmly committed to the promotion of high school graduates to at least the lower level managerial ranks. In fact, the top of the hierarchical ladder for a typical high school graduate, the position of a Supervisor (kakaricho or kojocho), is equal in rank to the first step of the management ladder reached by most employees with college degrees. Occasionally, a former blue-collar worker can be promoted to Manager or even to a Senior Manager position. These cases, however, are rather unusual, even though stories of such careers are favorite items of company folklore. The system remains flexible, in case an exception is in order, but for all practical purposes, most blue-collar workers and white-collar high-school graduates understand that they have little chance to become managers, nor are many of them truly interested in such career progression (Marsh & Mannari, 1976, Kamata, 1982). HUMAN RESOURCE MANAGEMENT FUNCTIONS Given the Japanese emphasis on managing "human assets" it is not unusual that the Personnel Department is one of the most powerful and prestigious departments in a typical Japanese corporation. The HRM functions of selection, development, appraisal and rewards are to a large degree centralized. The manpower requirements are carefully monitored, and the Personnel staff has 11

the final say in almost all promotion and transfer decisions. The Personnel staff is also a full partner in long-range strategic planning. In addition, Personnel serves as a center for organizational learning by supporting an exchange of information on corporate goals and performance between management and employees. This all is accomplished with a staff comparable in size to that found in American firms. The above general characterization of the Human Resource Management system applies equally to ELCO and STEELCO. However, when their HRM organization and functions are analyzed in greater detail, several differences between the two companies emerge. Fundamental functional similarities still remain, as the ILM structural requirements, such as limited entry and gradual promotion, limit the scope of available personnel policy alternatives. At the same time, the emphasis on "management by socialization" puts a highly visible and "unique" institutional imprint on personnel practices in each firm. Personnel Organization In ELCO and STEELCO, the Personnel and Labor Relation functions are integrated into a single department headed by a General Manager who report to the President/CEO through a Senior Managing Director in charge of Administration (in some other firms that I have observed, the two functions are managed separately). The departments are divided into several sections that deal with planning, transfers and promotions, recruiting, and with administration of salaries and benefits. In ELCO, the Training 12

and Development staff, while nominally separated from the Personnel Department, reports to the same General Manager. In STEELCO, T&D is treated as an independent unit with its own General Manager, although the two departments interact daily and staff rotation between the two groups is common. In both ELCO and STEELCO, the personnel policies for overseas employees and for expatriates are supervised by semi-autonomous sections in the head office reporting to the General Manager of Personnel. On the plant level, Personnel is represented through plant sections focused on personnel services, such as salary and benefit administration. The plant personnel staff formally reports to the plant deputy manager in charge of administration who also very often has an extensive HRM experience. The plant manager has the direct operational responsibility, but policy decisions are centralized and controlled from the head office. Large sales and engineering offices are controlled in a similar fashion. At the corporate level, the most prominent section within Personnel is "Organizational Planning" that has direct access to key corporate officers and to Corporate Planning. This section oversees performance appraisals for white collar employees, all transfers and promotions on the managerial level, and is also responsible for long-term human resource planning. In its work, the OP staff is a heavy user of computerized HR data banks though which information on all white collar employees is available online. An assignment to the OP staff is considered very desirable as it is seen as a stepping stone to future executive positions. 13

In ELCO, the OP staff consisted of personnel specialists who were thoroughly familiar with the careers of several thousands of white collar workers. For example, they had no trouble identifying employees from anonymous data sheets on transfers and promotions. In contrast, the OP staff in STEELCO also included managers specialized in other functional areas, such as sales or engineering. These were assigned to Personnel as a part of their long-term developmental training. In this sense, the OP staffing patterns reflected the differences between the two firms. In ELCO, the Personnel function was rather specialized, with employees migrating out but seldom back in. In STEELCO, the cross-functional transfers in and out of Personnel were more frequent. One exception to the latter trend was the Labor Relations group that deals with the company union (the union includes blue collar workers and also white collar employees below the rank of Manager). Employees assigned to Labor Relations might move from one plant to another, or to the head office, but seldom outside the LR function. The objective is to maintain stable communication links with the union officials. In fact, several managers on the LR staff in both firms were former elected union representatives. In the Japanese context, this is not unusual, as about 16% of executives in large Japanese firms are reported to have occupied a union post sometime during their careers (Shirai, 1983). Corporate Recruiting is another section where communication abilities are emphasized, even though the CR staff turnover is 14

faster than in Labor Relations. As explained in greater detail in the next section, successful recruiting efforts for white collar employees are dependent on the company's reputation and on the quality of its ties with university faculty and staff. During the actual recruiting period, most PD staffers are involved in the selection interview process. The job of the recruiting staff is to ensure that the pool of job candidates is of desired quantity and quality. Wage administration, especially at a plant level, is often the first assignment for future HRM specialists. Salary and benefits administration is a pure service area, as policy decisions concerning the reward structure originate, depending on the level of employees involved, from Organizational Planning and from Labor Relations (all white and blue collar workers are on salary). Salaries and non-cash benefits are administered separately. For the latter, personnel functions include management of dormitories, company houses, housing subsidy programs, sport and recreation facilites, and of an extensive medical care system, including company hospitals. However, only a few of the PD staffers involved with benefits programs are regular employees. Most facilities are staffed by reemployed company retirees, or by "special status" employees (e.g. doctors and nurses). Selection The recruitment of college graduates takes place once a year only. The selection process starts during the summer when company 15

recruiting materials are distributed to college seniors (the academic year in Japan begins on April lst.). Under "administrative guidance" from the Ministry of Education and the Ministry of Labor, in order to safeguard the educational integrity of the last college year, formal employee selection interviews should not be conducted before October and actual hiring decisions should not be made before the beginning of November. Officially, all companies adhere to this schedule, but most attempt to get a jump on the competition by persuading the more able students, identified through contacts such as employee referrals or an introduction from a friendly professor, to apply early and informally conclude the process before the official recruiting season actually begins. The selection process format depends on the students' future occupational class. Students who major in the social sciences, law, or the humanities will enter administrative (jimukei) jobs, such as planning, personnel, sales, or purchasing. Students majoring in technical disciplines will enter technical (gijitsukei) jobs. The selection procedures differ for each of the two occupational classes. Aspiring administrators are expected to apply for jobs directly to employers. Following the formal application, the job candidate is asked to appear for a set of interviews with company employees, managers and executives. The basic criteria for hiring, besides an employee's potential or ability, are "balanced" personality and moderate views. The evaluation of job candidates is often supplemented by background checks assigned to 16

private investigators who interview the candidate's neighbors and acquaintances, check local police records and examine the family history. Those who pass the last round of interviews are invited to sit for the company entrance examination at the beginning of November. Officially, this exam should determine who is best qualified for the job, but both in ELCO and STEELCO over 90% of candidates are already preselected on the basis of the earlier interviews. The exam usually asks essay questions on such topics as family background, career/life-objectives, the applicants strengths and weaknesses, etc. A number of companies are actually using the exam as an assessment tool to determine the career interests of new employees. For example, in STEELCO, the entrance examination results are used in decisions on employees' first assignments. The prospective technical employees, college seniors majoring in technical disciplines, are approached through their professors and departments. Although a direct application to a company is not ruled out and would be acted upon, companies generally prefer to deal with an intermediary. One reason is the difficulty of evaluating a student's technical potential on the basis of a short interview only, when the majority of interviewers have little up-to-date technical background. For reasons pointed out above, if the company waited for a written exam it could not preselect and would risk losing the best candidates to the competition. 17

The second reason for the reliance on intermediaries is the competition for top technical talent. Recruiters cultivate good relations with college professors (again, preferably those from prestigious schools) who then recommend individual firms to their students. By relying on intermediaries, the companies expect to get their "fair share" of talent and at the same time prevent a self-defeating bidding war that would not only raise the starting salaries for selected jobs, but, more importantly, would disrupt the carefully balanced compensation structures of Internal Labor Markets. Hiring through intermediaries makes the selection process less costly, as it reduces the number of candidates to be screened. In fact, a similar process has begun to appear for jimukei applicants as well. As the number of students even at the best universities increased dramatically, the seminars (courses taken with a professor's permission) of certain faculty members are considered to be more selective than others and develop an "elite" reputation. While students in these seminars still have to apply for employment through regular channels, who in the senior seminar is going to apply to what firm is often decided in advance under the professor's guidance. The difference in the selection process between administrators and technicians is not accidental. Other personnel policies, in particular those concerning training and development, also differentiate between the two groups of white-collar employees. Both ELCO and STEELCO keep separate personnel statistics concerning the two categories of employees, as do most other 18

large Japanese firms. With few exceptions, their career paths rarely intersect. College grades are not an important selection tool for either occupational class. What matters are the educational credentials of the school from which the student is graduating. Given the rigorous competition to enter the first-tier schools (most of which are public, such as Tokyo or Kyoto Universities), the companies rely on the university entrance examination as an indicator of an employee's "latent ability" (Iwata, 1981). This bias in favor of graduates from the more prestigious schools reinforces the competition to enter the appropriate college. In the early postwar period, most large firms and government agencies targeted their recruiting on the graduates of a few prestigious schools. Gradually, as the demand for university graduates increased, the recruiting drive was expanded to other schools. For example, since the late 1950's both STEELCO and ELCO recruit at more than fifty campuses. Without question, educational credentials play today only a secondary role in the selection process at least on the entry-level. Whether or not the same can be said about career progression over the long-run is, however, another matter. Development One of the reasons why Japanese companies prefer to hire only the new school graduates is that a "virgin work force can be readily assimilated into each company's unique environment as a community" (Hazama, 1978:148). The process of assimilation, which 19

is the cornerstone of the career development program in Japanese firms, begins with the initial training which is geared toward familiarizing new employees with the company; this can last from several weeks to several months. In STEELCO, the training program lasts four months, in ELCO six months. During the course of the program, the recruits learn about the business philosophy of the company and experience work on the factory floor as well as, in the case of ELCO, in the sales offices. They are expected to become "company men", and therefore, their functional specialization is of secondary importance (Clark, 1979). Both the careful screening during the selection process and the subsequent socialization program are designed to develop the homogeneity of the personnel in the firm. After the initial training program is completed, the new employees are assigned to their first jobs, the vast majority in the "field". The choice of assignments is influenced by a number of factors: current company needs, assessment of the new employees' skills, and also, but only to a limited degree, the personal preferences of the new employees. With a few exceptions for hardship cases (illness in the family, etc.), these choices are limited to the area of initial specialization, not to the specific location or region preferred. In addition to the initial socialization, a "resocialization" (Katz, 1980) takes place each time an employee enters a new position, as he has to familiarize himself with a new set of people and tasks. Average tenure in a job varies, depending on 20

the firm, the employee's functional specialization and career stage. However, it is argued that rotation is more frequent in early career stages, when learning new tasks is especially critical (Suzuki, 1981). Employees are transferred for two main reasons. First, they are assigned to new positions to learn additional skills in on-thejob training programs which form the core of the companies' training and development efforts. Off-the-job education is usually limited to subjects such as human relations, leadership, negotiations, etc. Second, transfers are part of a long-range experience building program through which the organization grooms its future managers. This program usually takes the form of periodic, lateral, interdepartmental transfers. While employees rotate semilaterally from job to job, they become increasingly socialized into the organization, immersed in the company's philosophy and culture, as well as bound to a set of shared goals (Pucik & Hatvany, 1983). Job rotation in Japanese firms is usually described as "generalist" in nature. In fact, the proposition that Japanese managers follow a so-called generalist career path is probably one of the most frequent statements concerning the "unique" features of Japanese organizations. For example, Suzuki (1981:34) asserts that, "(An employee) is regularly rotated from one section to another. While climbing up the seniority ladder this cross sectional movement scales up to a cross-functional rotation and continues up to the age of 40."'In the same manner, Ouchi (1981:32-33) commented that, "people conduct careers between 21

specialities but within the organization....All managers rotate through all areas of business." Clark (1979) and Vogel (1979) arrived at similar conclusions. Strictly speaking, according to these statements, the majority of Japanese middle managers move along a differentiated career path from one functional specialty to another. However, given that most managerial careers lead to a plateau somewhere in the middle of the organizational hierarchy rather than to a top executive job, it is also important to consider whether the career path pattern attributed to most managers is also characteristic of the elite group among them. In this respect, Vogel (1981) and Suzuki (1981) stated explicitly that career path differentiation is a characteristic of an elite career, the specialist functions being, with exceptions, relegated to those with lesser potential. However, the perception of Japanese managers as generalists discounts the possibility that career patterns might differ depending on the functional specialization. This is taken into consideration in only two surveys by Tawara (1979, 1980) covering 1,200 managerial-class employees from more than 40 firms. With the sample subdivided according to the manager's functional area, varied career patterns became apparent. Unfortunately, one major shortcoming affects Tawara's studies. The cross sectional sampling method used in the surveys makes it impossible to analyze the relationship between career path patterns and career achievement, as it is not known how the respondents' career progression compares with the companyspecific "elite" timetables. To examine such a relationship, 22

organization-based cohort analysis would be more appropriate. Indeed, a cohort analysis of mobility patterns of white collar employees in ELCO and STEELCO shows a picture substantially at odds with the "generalist" proposition (Pucik, 1984). First, the scope of the interfunctional and inter-divisional (interdepartmental) mobility was found to be closely related to the manager's occupational class. This is seen as a consequence of task uncertainty facing managers in different occupations. The careers of administrators were characterized by a broader functional and positional mobility than the careers of managers in technical positions. Second, the observed relationship between the interfunctional mobility and career achievement of managers was in the direction opposite to the one predicted. Narrow -- rather than wide -- functional specialization is associated with career success. A moderating impact of the manager's occupational class was observed. In both firms, this relationship is stronger for technicians than for administrators in terms of its impact on career achievement, but only in ELCO is the difference statistically significant. Appraisal Both in STEELCO and ELCO, periodic appraisals of white collar employee performance are conducted several times a year. For example, in ELCO, performance evaluations precede salary increase decisions due in April of each year, the summer and winter bonus determination,. and the annual career development review, with timing dependent on the employee rank or position. In addition, 23

for many, but not all employees, performance evaluations linked to decisions on promotions are conducted in early fall. In STEELCO, the type of appraisal is similar, but with less frequency, as the evaluation for salary increase and the career development review are combined. This kind of frequent performance appraisals is characteristic of most large Japanese firms. What is important, however, is that in all cases intensive appraisals occur regularly from the very first year a new employee enters the firm. These evaluations clearly discriminate among employees as each is compared to other members of his peer group and ranked accordingly. In ELCO, the peer group is the cohort of entry and therefore the ratings are fairly straightforward. However, in STEELCO, the peer group is that of employees in the same position class. The rating score reflects not only an employee's relative performance, but also his years of tenure in the particular position. As a result, direct intracohort comparisons across position classes are difficult. As in many American firms, the evaluation criteria depend on the position class. For example, for white collar employees in STEELCO, there are two sets of appraisal criteria, one up to the level of Supervisor, the other for those above. The evaluations for managers have four major components: - a single achievement score, - six scores measuring job-related abilities such as human relations skills, business judgment, coordination and planning, - two outstanding job-related attributes (selected from a list of eight, such as creativity, leadership, and reliability), - two outstanding personality-related attributes (selected from a list of twelve, such as sociability, flexibility, confidence). 24

These items are first evaluated by the direct supervisor, then by at least two other higher-level managers or executives. One noticeable difference between appraisal systems in Japan and those in the U.S. is the lack in the former of feedback concerning the appraisal results. As shown in interviews with the Personnel Managers in 34 of the top 100 Japanese corporations, only two of the firms had a formal policy recommending explicit performance feedback (Tawara, 1979). In the other cases, the form and frequency of feedback is left to the discretion of individual managers, who may or may not provide it. Also, some personnel executives expressed the opinion that an employee's ability to solicit feedback is a mark of understanding how informal communication channels operate and is thus a part of the appraisal process itself. Given the unattractive exit opportunities, however, the majority of employees who are not the "fast trackers" may find it preferable to avoid feedback so they do not have to confront directly the reality of failure. Even if individual supervisors are willing to provide specific feedback to their subordinates, they often do no more than inform an employee about what rating was suggested as ratings are usually adjusted at the corporate level to smooth out differences between departments. The results of these adjustments are very seldom transmitted back to the first-level evaluator. However, ranking results can be inferred by comparing one's paycheck with those of the peer group or with announced average salary increases. Obviously, those with higher than average increases 25

encourage informal comparisons to make their success known. Those with less than average performance would rather avoid it. The competitive nature of the system and the resulting intracohort rankings are not very visible during the first 10-12 years of tenure in the organization, as the rankings are not disclosed. This has led many observers (e.g. Ouchi, 1981) to propose that evaluation in large Japanese firms is long-term, based on years of careful judgments and comparisons. However, when ranking in the cohort becomes visible, it is usually too late to do anything about it. As discussed in the next section, in many Japanese firms the chances for recovery from low ratings are slim. Rewards The effect of seniority on rewards is probably one of the most frequently misunderstood aspects of Japanese personnel policies. As the length of tenure in the firm increases so does the basic salary, which for many years is equal for all from the same cohort. However, at the same time the ratio of the seniority component to total compensation gradually falls. First of all, starting in the second decade of employment, seniority increases begin to taper off and the seniority wage curve becomes virtually flat for employees with twenty or more years of tenure. In addition, merit pay increases are awarded annually on the basis of performance evaluations. For junior personnel the differences in monthly pay are merely symbolic, but sufficient to foster intense competition (Iwata, 1980). Sometimes, as in ELCO, merit and basic salary increases are combined. Each year an 26

employee is awarded a pay raise equal to an average increase for a given cohort plus or minus his merit rating. Thus, the career achievement of individuals within a single cohort can be compared by measuring the differences in salary increases over time. However, numerous other allowances unrelated to performance (e.g. family, housing or commuting allowances) that may increase the take-home pay by up to 10-15% should be taken into account in any such comparison. On top of that, position class allowances are introduced, beginning with the Supervisor or Manager class, depending on the firm (the former in STEELCO, the latter in ELCO). Consequently, after approximately twenty years of gradually widening differences, the gap between those on the "elite-track" and those at the bottom of the cohort is about 30% of the average cohort salary. Given the corresponding difference between two position classes, the salary gap is about half of that usually observed in the U.S. (Mahoney, 1979). The impact of seniority is also often exagerated with respect to promotions. With promotions being awarded on the basis of merit evaluation as seen above, seniority is a necessary, but not a sufficient, condition of success. Most companies, including ELCO and STEELCO, set up minimum tenure requirements for a position or minimum tenure in a grade before the next promotion or both. As a result, career progression lines are rigid and chances to recover after a slippage in ratings and a missed promotion are limited. In this sense, the Japanese competition rules ressemble closely the "tournament" model of career progressions developed 27

by Rosenbaum (1979). Both systems allow "recovery" of a fallen ranking only in exceptional cases; the common pattern is to keep one's standing within the cohort, or to slip down. For example, in a another firm which I examined (Pucik, 1981), according to company regulations, it takes at least 16 years to be promoted from the lowest management position of Supervisor to that of General Manager (the same as in STEELCO and one year longer than in ELCO). The first hurdle, promotion to Manager, comes four years after the initial promotion to Supervisor, which is granted automatically. Among six cohorts analyzed, only 5% of those who missed the earliest promotion to Manager (but by no more than one year) succeeded in reaching the General Manager level early enough to be seriously considered for any executive position. In addition, two years of delay eliminated chances to reach the class of Senior Manager altogether. Most white collar employees, with the exception of a few toplevel executives, are "eased out" to a subsidiary in conjunction with their mandatory retirement from the parent firm (which occurs at the age of 60 in most large companies), unless they choose not to seek further employment. However, an early involuntary transfer to a subsidiary is a significant threat to many managers. Although for a certain period (usually one or two years) the parent company may pay the salary differential, eventually the salary is adjusted downwards to the subsidiary level. Therefore, involuntary transfers, while infrequent, function as a powerful control mechanism to insure acceptable performance even with the virtual guarantee of job security. 28

However, important differences were observed in a cohort analysis of career advancement for administrators and technicians with 20 years of seniority in the firm. The average rank/position status of ELCO's and STEELCO's administrators was found to be significantly higher than that of technicians. In Table 3, the "success" of ELCO's administrators, as opposed to technicians, is obvious. Although they account only for 28 percent of the cohort examined, among those belonging to the elite group of senior managers, two-thirds are administrators. TABLE 3 - ELCO: POSITION CLASSES BY OCCUPATIONAL CLASS POSITION CLASS ADMINISTRATORS TECHNICIANS SENIOR MANAGER 35% 7% MANAGER 61% 83% SUPERVISOR - 1% SPECIALIST 4% 3% UNCLASSIFIED - 6% TOTAL 100% 100% In STEELCO, approximately 43 percent of administrators in the cohort are ranked as Senior Managers (jicho) but only 32 percent among the technical personnel. The rest are ranked as Managers (kacho). This again indicates better promotion chances for administrators, even though at the top executive level both companies are headed by engineers. In other words, fewer technicians succeed as managers, but those who do may go a long way. One more important point concerning careers and rewards should be mentioned. In organizations with Internal Labor Markets where ports of entry are limited to the entry-level positions and 29

promotions are only from within, vacancies in higher positions emerge sequentially as each cohort progresses through the hierarchy. Promotions are thus unlikely to be rapid unless the organization is expanding rapidly. This limited upward mobility encourages lateral job rotation, as carefully planned job transfers add substantial flexibility to the reward system. At the same time, not all jobs at the same hierarchical level are equal in their centrality or importance to the organization's activity (Schein, 1971). By assigning individuals to jobs that are at the same level but vary in their centrality, the organization can de facto discriminate in terms of both promotion and demotion among employees who, within the formal system, share the same status, salary and job perquisites. CONCLUSIONS This paper described the ILM structure and related Human Resource Management practices in large Japanese manufacturing firms. On a number of dimensions, the HRM practices as applied towards the white collar professional and managerial employees were shown to be different from "typical" Japanese management as described in the previous studies on this subject. This is probably due to the fact that the earlier studies were focused on blue collar personnel practices, or on personnel practices in public organizations, rather than on white-collar employees in the manufacturing sector. In particular, it seems that successful Japanese middle-managers are functional specialists, rather than generalists. The 30

findings also indicate that selection practices, mobility patterns, and distribution of rewards differ by occupational class, and that, on a number of HRM dimensions, the differences in the ILM structure influence differences in HRM functions between organizations. While most large Japanese corporations endorse the concept of life-time employment and management principles related to it, the actual personnel practices may vary greatly both within and across firms. The HRM policies in large Japanese firms are carefully designed to integrate a relatively stable and homogeneous labor force into highly competitive organizations committed to excellence. However effective in their local context, it seems that a "wholesale" replication of Japanese methods may be misplaced. Although an in-depth comparisons were not attempted here, differences with accepted American HRM practices were observed on a number of critical dimensions. At least three such differences stand out. First, selection is highly centralized, and screening criteria involve a variety of job-unrelated characteristics. Second, direct feedback in the appraisal process is infrequent, if there is any at all. Third, within-cohort compensation levels are significantly compressed. Change towards the Japanese approach may be highly controversial and often, as in the case of the selection criteria, socially unacceptable. Proceed with care. 31

REFERENCES Bowman, M.J. Educational Choice and Labor Markets in Japan. Chicago: Univ. of Chicago Press, 1981. Clark, R. The Japanese Company. New Haven, Conn.: Yale Univ. Press, 1979. Cole, R.E. Japanese Blue Collar: The Changing Tradition. Berkeley, Cal.: University of California Press, 1971. Cole, R.E. Work, Mobility and Participation: A Comparative Study of American and Japanese Industry. Berkeley, Cal.: Univ. of California Press, 1979. Doeringer, P.B. and M.J. Piore. Internal Labor Markets and Manpower Analysis. Lexington, Mass: D.C.Heath, 1971 Dore, R.P. British Factory, Japanese Factory: The Origins of National Diversity in Industrial Relations. Berkeley, Cal. Univ. of California Press, 1973. Hazama, H. "Characteristics of Japanese-Style Management." Japanese Economic Studies, 6, No. 3-4 (1978), 110-173. Iwata, R. Gakureki-shugi no Hatten Kozo. Tokyo: Nihon Shoronsha (In Japanese), 1981. Kamata, S. Japan in the Passing Lane. New York: Pantheon Books, 1982. Katz, R. "Time and Work: Toward an Integrative Perspective." In Research in Organizational Behavior, Vol. 2. Greenwich, Conn.: JAI Press, 1980, 81-127. Mahoney, T.A. "Organization Hierarchy and Position Worth". The Academy of Management Journal, 22 (1979), 726-737. Marsh, R.M. and H. Mannari. Modernization and the Japanese Factory. Princeton, N.J.: Princeton Univ. Press, 1976. Noguchi, P.H. "Shiranai Station: Not a Destination But a Journey." In D. W. Plath. Ed. Work and Life Course in Japan. Albany, N.Y.: Suny Press, 1983. Ouchi, W.G. Theory Z. Reading, Mass.: Addison-Wesley, 1981. Pucik, V. "Promotions and Intraorganizational Status Differentiation Among Japanese Managers." Proceedings of the Academy of Management, 1981, 59-63. Pucik, V. Careers and Internal Labor Markets: Determinants of Career Achievement among Managers in Large Japanese Firms. Unpublished doctoral dissertation, Columbia University, 1984 32

Pucik, V. and N. Hatvany. "Management Practices in Japan and Their Impact on Business Strategy." In R. Lamb. Ed. Advances in Strategic Management. Greenwich, Conn.: JAI Press. 1983. Rohlen, T.P. For Harmony and Strength: Japanese White-Collar Organization in Anthlroporogicar Perspective. Berkeley, Cal. Univ. of California Press, 1974. Rosenbaum, J.E. "Tournament Mobility: Career Patterns in a Corporation." Administrative Science Quarterly. 24 (1979), 220-241. Schein, E.H. "The Individual, the Organization and the career: a Conceptual Scheme." Journal of Applied Behavioral Science. 7 (1971), 401-426. Shirai, T. "A Supplement: Characteristics of Japanese managements and Their Personnel Policies." In T. Shirai. Ed. Contemporary Industrial Relations In Japan. Madision, Wise.: University of Wisconsin Press, 1983. Skinner, K.A. "Aborted Careers in a Public Corporation." In D. W. Plath. Ed. Work and Life Course in Japan. Albany, N.Y.: Suny Press,1983. Suzuki, N. Management and Industrial Structure in Japan. New York: Pergamon Press, 1981. Tawara, J. Career Patterns of General Managers. Tokyo: Center for Human Development (In Japanese), 1979. Tawara, J. Career" Patterns of White-Collar Employees. Tokyo: Center for Human Development (In Japanese), 1980. Vogel, E.F. Japan As Number One. Cambridge, Mass.: Harvard Univ. Press, 1979. 33