University of Southern California Dissertations and Theses

Career advancement in managerial hierarchies in the United States firms: A multi-theoretic model and empirical tests of the determinants of managerial promotion

(USC Thesis Other)

Career advancement in managerial hierarchies in the United States firms: A multi-theoretic model and empirical tests of the determinants of managerial promotion

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UM1
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may
be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in
reduced form at the back o f the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6 " x 9" black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly
to order.
A Bell & Howell Information Company
300 North Z ee b Road. Ann Arbor. M l 48106-1346 USA
313/761-4700 800/521-0600
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CAREER ADVANCEMENT IN MANAGERIAL HIERARCHIES IN THE U.S. FIRMS:
A MULTI-THEORETIC MODEL AND EMPIRICAL TESTS OF THE
DETERMINANTS OF MANAGERIAL PROMOTION
by
Yong-Min Kim
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Business Administration)
August 1995
Copyright 1995 Yong-Min Kim
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
UMI Num ber: 9616977
Copyright 1995 by
Kim, Yong-Min
All rights reserved.
UMI Microform 9616977
Copyright 1996, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90007
This dissertation, written hy
Yong-Min Kim
under the direction of hh?. Dissertation
Committee, and approved by all its members,
has been presented to and accepted by The
Graduate School, in partial fulfillment of re
quirements for the degree of
DOCTOR OF PHILOSOPHY
Dean of Graduate Studies
Date . . f S L L ’ J E l
DISSERTATION COMMITTEE
i/i Chairperson* I
.......................
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
TABLE OF CONTENTS
CHAPTER 1: INTRODUCTION AND OVERVIEW.............................................................1
Objectives of the Dissertation Research....................................................................4
Dissertation Overview..................................................................................................5
CHAPTER 2: THEORY AND HYPOTHESES........................................................................7
Career Advancement in Managerial Hierarchies.................................................... 7
Structural Models— Managerial Internal Labor Markets (MILMs)
and Promotion............................................................................................................9
Individual Models— Human Capital and Social Capital Factors
and Promotion..........................................................................................................21
Tournament Mobility Model and Promotion.........................................................35
Relational Demography Model and Promotion................................................... 43
CHAPTER 3: DATA AND METHODS................................................................................ 50
Sample........................................................................................................................... 50
Measures: Dependent Variable............................................................................... 51
Measures: Control Variables..................................................................................... 53
Measures: Independent Variables........................................................................... 55
Data Analyses..............................................................................................................67
CHAPTER 4: RESULTS...........................................................................................................69
General Description of MILMs................................................................................ 69
Results of Event History Analyses.......................................................................... 75
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
CHAPTER 5: DISCUSSION AND CONCLUSION.......................................................... 93
REFERENCES.......................................................................................................................104
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
FIG URES A N D TABLES
Figure 1: An Integrative Model of Executive Promotion................................................ 10
Table 1: Frequencies and Percentages of Entries, Promotions, and Exits
by Hierarchical Level.............................................................................................71
Table 2: Frequencies and Percentages by Types of Promotion.................................... 74
Table 3: Means, Standard Deviations, and Correlations for All Variables
in tire Study.......................................................................................................75-77
Table 4-a: Logistic Regressions of Executive Promotion............................................ 80
Table 4-b: Logistic Regressions of Executive Promotion........................................83-84
Table 4-c: Logistic Regressions of Executive Promotion.............................................. 86
Table 4-d: Logistic Regressions of Executive Promotion........................................88-90
APPENDICES
Appendix 1: List of tire Sample Firms......................................................................... 102
Appendix 2: Elite Educational Institutions................................................................. 103
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
ABSTRACT
Career advancement of top managers has not been well studied in the literature of
organization science for decades. This dissertation develops an integrative theoretical
model of career advancement in the managerial hierarchies of major U.S. profit-seeking
organizations. The theoretical model draws on structural models of internal labor markets,
individual models of career mobility primarily led by human capital and social capital
theory, the tournament career mobility model, and theories of relational demography.
A series of event history analysis were used to model the occurrence of promotion of
1576 top managers collected from 92 manufacturing firms drawn from 3 industries: 32 firms
from the computer industry, 34 firms from the chemical industry, and 26 firms from the
natural gas industry. The data were drawn from a population of the largest firms in each
industry for up to 13 fiscal years between 1978 and 1990 as long as the data were available.
Promotion was identified exclusively on the basis of status enhancement in terms of rank or
level in the managerial hierarchy to avoid any confusion or ambiguity.
The findings can be summarized as follows. First, managerial internal labor markets
(MILMs) can be characterized as internal promotions to fill vacancies at higher levels, job
ladders based on functions, and stable hierarchical promotion systems. Second, human
capital and social capital factors were associated with promotion in the expected directions.
General and firm-specific human capital variables, age, managerial task knowledge, and
with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
managerial productivity were correlated with the likelihood of promotion. Two key social
capital variables, external networking ability and internal network centrality were positively
associated with the likelihood of promotion. Third, all three aspects of tournament, overall
career velocity, losing (or winning) the present tournament round, and losing (or winning)
the latest previous round, were associated with the likelihood of promotion as predicted.
Fourth, except in the case of educational affiliation similarity, all the other dyadic relational
demographic characteristics between the CEOs and the managers affected the likelihood of
promotion. Functional similarity, age similarity, and outside status similarity were
positively associated with the likelihood of promotion as predicted.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
1
CHAPTER 1: INTRODUCTION AND OVERVIEW
Top managers of America's large business corporations, primarily CEOs,
presidents, and various levels of vice presidents, are viewed as occupying an
elite group at tire summit of its class structure (Galbraith, 1971; Useem &
Karabel, 1986; Kurtz, Boone, & Fleenor, 1989). As Galbraith (1971) succinctly
summarized it, "The decisive power in modern society is exercised not by
capital but by organization, not by the capitalist but by the industrial
bureaucrat." Top managers are also viewed as key strategic decision makers
who can affect the survival and success of their entire organizations as
evidenced by numerous researchers (Andrews, 1971; Barnard, 1938; Child, 1972;
Donaldson & Lorsch, 1983; Hambrick & Mason, 1984; Hickson, 1987). More
recently, an increasing body of literature documents the intervening role of top
managers between environmental contingencies and organizational outcomes
(Finkelstein & Hambrick, 1990; Hrebiniak & Joyce, 1985; Milgrom & Roberts,
1992; Priem, 1990; Bantel & Jackson, 1989; Wiersema & Bantel, 1993). More than
any other single set of people, their decisions have decisive impact not only on
tire destiny of organization, but also on the competitiveness of the entire
industry or of tire nation (Porter, 1990).
The question of who is selected and promoted into a relatively small
cadre of senior managers has always attracted ordinary men's attention in this
country (Krutz et al., 1989). Yet, oddly enough, career advancement at the
executive level has not been well-studied (Vancil, 1987; Powell & Butterfield,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
1994; Tharenou, Latimer, & Conroy, 1994), and exceedingly little has been
written about the characteristics of the U.S. business executives' promotion
(Swinyard & Bond, 1980). Let alone the promotion of employees in general
(Markham, Harlan, & Hackett, 1987), both the process and the determinants of
career advancement of executive managers in managerial hierarchies are largely
unresearched.
Three major problems, theoretical and empirical, have contributed to this
lack of attention to executive career advancement. First, past research on
executive promotion has narrowly focused on the very top positions in
organizations, primarily CEOs, in die context of managerial succession under
the assumption tiiat the CEOs have die ultimate power in determining
organizational destiny (Helmich & Brown, 1972; Dalton & Kesner, 1985;
Lieberson & O'Connor, 1972; Smith & White, 1987). This inordinate theoretical
focus on CEOs has prevented researchers from looking at the mobility processes
below the CEO level, which can be equally interesting and important
phenomena from an "upper-echelon perspective" (Cyert & March, 1963;
Hambrick & Mason, 1984; Finkelstein & Hambrick, 1990). Mobility processes
below die CEO level can be meaningful especially because CEOs are usually
selected from diis "upper-echelon."
Second, promotions at die upper-echelon of organizations are sensitive
political events often involving issues of personal egos (Dalton, 1951;
Rosenbaum, 1984; Pfeffer, 1989). In organizations, dominant coalitions
endeavor to institutionalize tiieir power through the manipulation of personnel
at the top (Cyert & March, 1963; Pfeffer, 1981). Due to this sensitivity, top
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
managers are reluctant to reveal information concerning their political
behaviors. Also, unlike those at lower ranks in organizations, those at the top of
the organization possess power to prevent the interested students of
organizations from intruding (Bird, 1988). In addition, there is no mandatory
requirement for firms to make information regarding executive promotion
practices open to non-key personnel outsiders, whether they are inside or
outside the firm (Bird, 1988).
Third, studying executive promotion often requires extensive data
collection, especially longitudinal data. Past research has not adequately tested
career advancement of people in organizations due to its cross-sectional
analyses (Abraham & Medoff, 1985; Cannings, 1988; Stewart & Gudykunst,
1982; Whitely, Dougherty, & Dreher, 1991; Zenger, 1992). Although there has
been a significant volume of longitudinal research on die basis of personnel
records of single organizations (e.g., Baker, Gibbs, & Holmstrom, 1993,1994a;
Forbes, 1987; Rosenbaum, 1979; Sheridan, Slocum, Buda, & Thompson, 1990;
Stewman & Konda, 1983; White, 1970), these studies are exposed to lack of
generalizability due to the idiosyncratic nature of their samples (Rosenbaum,
1984).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
4
Objectives of the Dissertation Research
In this dissertation, I will seek to develop an integrative model of career
advancement in the managerial hierarchies of major U.S. profit-seeking
organizations. The underlying research philosophy of this dissertation is that a
single theory alone cannot explicate complex organizational phenomena such as
the promotions of high ranking managers. Toward that end, various research
paradigms including structural, individual, tournament, and relational
demography models will be explored to explain career advancements by
particularly focusing on the top-management echelon including vice presidents
or higher ranks. In the process, various related disciplinary areas such as labor
economics, organizational and occupational sociology, organizational behavior,
and social psychology will be examined.
Specifically, the following questions will be addressed: (1) What are the
structural characteristics of die managerial internal labor markets (MILM) that affect
the likelihood of promotion for senior managers? (2) Why do certain individuals
have more promotion opportunities than others? What are the individual
characteristics, including human and social capital factors, that affect the likelihood
of promotion for the executives? (3) What principles of career mobility prevail in the
MILMs of the major U.S. firms? Do firms follow random processes in determining
whom to promote, or do firms operate clearly rational career systems through
tournament competitions for top managers? If so, do the fast trackers who are
rapidly promoted at one level tend to carry the advantage to the next level? (4)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
What relational demographic properties affect the likelihood of promotion for top
managers? After all these questions are addressed, the following questions will be
posed and discussed in the end. Do these considerations serve any sort of efficiency
purpose for die organization and its members? What conclusions can be drawn
from the examination of various theoretical traditions and empirical results?
Dissertation Overview
Chapter 2 of the dissertation develops a theoretical model of career
advancement in managerial hierarchies and the research hypotheses. The
theoretical model draws on structural models of internal labor markets, individual
models of career mobility primarily led by human capital and social capital theory,
the tournament career mobility model, and theories of relational demography that
have been developed in social psychology. Each research paradigm and its
philosophy is examined to explain various underlying processes of career dynamics
of top managers. The model presented here is necessarily of an integrative and
interdisplinary nature due to tire complexity of the promotion phenomenon in
question. Based on tire discussion of each research tradition, a set of hypotheses is
then presented to test the theory.
Chapter 3 presents the design of the sample constructed for the dissertation
and the methodology used. First, tire sample and data sources for this study is
explained. Second, the operationalization and the measurement of the dependent
variable of promotion as a discrete event, are explained in detail. Third, the
operationalizations of the independent variables and their data sources are
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
presented. Fourth, data analytic methods including the use of discrete-time
methods for event history analysis and the determination of the promotion hazard
rate are explained.
Chapter 4 presents and discusses the empirical results. General
characteristics of MILMs are briefly discussed, and the hazard rate models of
executive promotion will be presented. Descriptive statistics including means,
standard deviations, and correlations are reported. The results of the individual
tests of hypotheses are then presented with a series of event history analyses.
Chapter 5 presents the discussions, general conclusions and the limitations of
the dissertation. Implications for further research in career advancement in upper
management in major U.S. firms are also discussed.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
7
CHAPTER 2. THEORY AND HYPOTHESES
Career Advancement In Managerial Hierarchies
Career advancement of managers provides many benefits to both
organizations and individuals. Moving managers through a series of progressively
more demanding positions allows an organization to train, evaluate, and motivate
people over an extended period (Martin & Strauss, 1959; DiPrete, 1987; Milgrom &
Roberts, 1992). For individuals, promotions provide higher status, higher pay, and
broader experiences (Baker et al., 1994a; Rosenfeld, 1992; Markham et al., 1987).
Despite these important roles, our understanding of managers' status attainment
and their career mobility is extremely limited (Powell & Butterfield, 1994;
Rosenbaum, 1984). Knowledge of career mobility of managers and its underlying
mechanisms has manifold implications for the students of organizations, firms and
their employees as well.
First, studying die upward career mobility can expand our knowledge of
MILMs (Baker et al., 1993,1994a; DiPrete, 1987). While there has been sizeable
research on blue-collar jobs in organizations, the study of white-collar jobs,
especially of MILMs, has been scarce (Osterman, 1981; Kanter, 1984).
Second, die study of MILMs reveals important information about how social
stratification and inequality in organizations are determined and maintained (Pfeffer
& Cohen, 1984). It corresponds to a growing academic endeavor of connecting the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
study of social stratification to the study of organizations (Baron & Bielby, 1981;
Baron, 1984; Pfeffer & Cohen, 1984).
Third, stratification and prospects for mobility affect organizations through
their impact on managerial motivation (Kanter, 1977; Anderson, Milkovich, & Tsui,
1981; Markham et al., 1987; London & Stump, 1982). The prospect of career
advancement serves to motivate individuals as they compete to move to positions of
higher status and pay (Lazear & Rosen, 1981; Lazear, 1991). Even though
individuals seek many kinds of rewards from work besides status and authority,
organizations allocate these nonstatus rewards to jobs in ways that are related to job
status (Rosenfeld; 1992; Rosenbaum, 1984). In this sense, an efficient promotion
system from organization's perspective can function as a powerful motivational tool
for the managers (Dreyfuss, 1938).
Fourtli, opportunities and prospects for promotion may be fundamental
determinants of a variety of other attitudes and behaviors that are more important
than an individual’ s personal traits in the organization (Kanter, 1977). For example,
promotion affects career aspirations (Argyris, 1957; Chinoy, 1955; Hall, 1976),
leadership style (Levenson, 1961), job satisfaction (Wilensky, 1956; Anderson et al.,
1981), organization commitment (Grusky, 1966; Kanter, 1977; Hall, 1976), and
turnover (Porter & Steers, 1973; Vancil, 1987; Sicherman & Galor, 1990). These
organizational behaviors and attitudes are critical in determining organizational
outcomes (Markham et al., 1987).
As much as an organization is a multifaceted entity affected by various
complex economic, socio-psychological, and political forces, so is career
advancement within managerial hierarchy of the organization. This dissertation
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
9
will investigate various theoretical traditions to explicate the complex dynamics of
managerial career advancement in major U.S. profit-seeking organizations. Four
models of managerial promotion— the structural, individual, tournament, and
relational demography models— will be discussed to assess the determinants of
promotion (see Figure 1). This integrative approach provides a broader
understanding of executive promotion which has to date been studied only from a
narrow disciplinary focus. This approach also makes it possible to determine the
relative importance of different theoretical explications. As such, this dissertation
contributes to both theory testing as well as theory development in the area of
managerial promotion.
The Structural Model— Managerial Internal Labor Markets and Promotion
Structural models analyze how social structures in organizations affect
careers. In this section, the major characteristics of internal labor markets, in
conjunction with the concept of vacancy, will be examined to explain managerial
career advancement. Structural models contend that structural aspects of the labor
market determine the career paths of employees (Doeringer & Piore, 1971). From
this perspective, individual's personal attributes are not believed to have as strong
an effect on their career paths as individualistic models argue. Instead, career
advancement takes place in MILMs, which influence the advancement processes. In
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
1 0
Likelihood of
Being Promoted
Tournament Factors
Overall Career Velocity
Number of Passovers at Current Level
Number of Passovers at Previous Level
Relational Demographic Factors
* Age Similarity to CEO
* Educational Affiliation Similarity to CEO
* Outside Status Similarity to CEO
* Functional Background Similarity to CEO
Structural Factors—
Managerial Internal Market Factors
Vacancy Rate at Higher Level
Vacancy Rate on Job Ladder
Cross-Job Ladder Movement
Location in Managerial Hierarchies
Individual Factors—
Human and Social Capital Factors
* Educational Level
* Age
* Organiztional Tenure
* Managerial Knowledge Base
* Pay Increase
* External Networking Ability
* Internal Network Centrality
Figure 1. An Integrative Model of Executive Promotion
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
fact, managerial jobs are characterized as having more features of internal labor
markets than professional or technical jobs (Baron, Davis-Blake, & Bielby, 1986):
Because managerial jobs tend to be linked more closely to the firm's core technology
(Kanter, 1984), managers are not as easily replaced from outside as other employees.
Literature on internal labor markets developed in occupational/
organizational sociology and labor economics defines internal labor markets as fixed
ports of entry, internal promotions as the main way of filling vacancies, and
normatively approved, established procedures for hiring, firing, and promoting
(Caplow, 1954; Doeringer and Piore, 1971; Kalleberg & Sorenson, 1979). After an
extensive review of die literature on internal labor markets, Althauser and Kalleberg
(1981) conclude that the concept of internal labor markets should include any cluster
of jobs that have diree basic structural features: (a) a job ladder, widi (b) entry only
at the bottom of die hierarchy and (c) movement up this ladder to fill the vacancies,
which is associated widi a progressive development of knowledge and skill. In
many ways, MILMs can be understood in light of the dieories of internal labor
market in general (Pinfield & Berner, 1994). In this section, some salient features of
MILMs— vacancies at each level of managerial hierarchies, vacancies on job ladders,
cross-job ladder movements, and the location in the pyramidal structure— are
examined below that are believed to affect die upward career mobility of managers.
The effects of exits, promotions, and entries at higher ranks Filling job vacancies
through internal promotion is a fundamental feature of MILMs (Pinfield & Berner,
1994). It has an advantage over hiring from external labor market as follows. Filling
job vacancies through internal promotion motivates managers through the hope of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
promotion, reduces tire likelihood that managers will leave for better opportunities
elsewhere, and provides promotion opportunities as an aid in recruiting better
employees (Pfeffer & Cohen, 1984). The actual promotion opportunities are created
when persons leave tire firm or move to a new job, setting in motion chains of
vacancies in an orderly manner (White, 1970; Miner, 1987; Sorensen & Kalleberg,
1981). Vacancies are also created when new positions or duties are added to the
hierarchy (Wlroley, 1985; Miner, 1987; Sorensen & Kalleberg, 1981). Mobility
depends on tire availability of empty positions and tire filling of jobs is
interdependent. A manager within tire system going to that job leaves his or her
previous job vacant and so on. A manager entering a system to take a job ends a
chain of vacancies.
The concept of vacancy chain multiplier has a particular relevance here. The
vacancy chain multiplier measures tire extent of the internal chain reaction of job
opportunities as firms iteratively fill vacancies with internal candidates before
finally going outside (Stewman, 1988). For example, a multiplier of 3 means that
two internal selections are made for every outside hire. Assuming that outside hires
take place only at tire bottom of the hierarchy, a multiplier of 3 means that internal
promotion stops two levels below tire original vacancy. The lower the multiplier in
relation to tire overall height of tire hierarchy, the less likely promotion opportunities
in tire lower levels of tire organization. No matter how the multiplier is determined,
tire highest chain effect is expected between adjoining levels in managerial
hierarchies: Exits and promotions at one level in the managerial hierarchy are most
likely to create and increase tire promotion opportunities for those down at the next
level.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Another important feature of internal labor markets— entry only at the bottom
and no entry above in the managerial hierarchy— may not exist in a strong form in
MILMs. As indicated by internal labor market theorists, jobs may be open above the
lowest entry port to outside the organization (Doeringer & Piore, 1971; Spilerman,
1977). Entry only at the bottom in most internal labor markets serves two functions.
First, it provides job security for employees with experience (Althauser & Kalleberg,
1981): Those with skills and experience can be protected from market competition.
Second, filling die vacancies at die bottom with less skilled or unskilled employees
enables employees with more experience at the higher position on the ladder to train
those with less. However, tiiese two positive effects are sometimes attenuated or
eliminated by otiier organizational or political needs to bring in new managerial
competencies at higher levels. For example, CEOs, die very top positions in
managerial hierarchies, are often brought in from outside to reorient strategic goals
and control mechanisms of die firm (Tushman & Romanelli, 1985). CEOs from
outside may have die motivation and ability to execute organizational change
programs more successfully than CEOs through inside promotions (Nystrom &
Starbuck, 1984; Korn, Milliken, & Lant, 1992). In a similar vein, entries to senior
managerial positions below the CEO are not rare (Keck & Tushman, 1993). If a
pressing need develops for employees with expertise in a new managerial or
technological area, die firm may hire people with the needed knowledge and skills
from outside rather dian attempting to develop the requisite expertise internally
(Milgrom & Roberts, 1992). Therefore, a certain degree of entry at all managerial
levels appears to be necessary. Firms can establish new technical competencies by
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
14
bringing in new managers from outside, and stimulate diverse opinions and
consideration of strategic change (Wiersema & Bantel, 1992).
It should be noted, though, that the practice of hiring executives from outside
does not necessarily eliminate the presence of internal promotions. There is
substantial empirical evidence indicating that internal promotion practices are
common for top managers in firms (Beatty & Zajac, 1987; Dalton & Kesner, 1985;
Forbes & Piercy, 1991; Swinyard & Bond, 1980; Wholey, 1985). In management
succession studies, CEO succession from outside was only 12 to 16 percent of the
entire sample of the succession events (Dalton & Kesner, 1985; Beatty & Zajac, 1987).
In another empirical study, Forbes and Piercy (1991) reported that about twenty to
twenty five percent of senior executive positions were filled by personnel outside
die firm. Wholey (1985) found diat the odds of an internal versus an external
selection of a partner in law firms were about 4 to 1. Swinyard and Bond (1980) also
showed diat, among about 11,000 newly appointed executives as vice presidents or
higher positions, less tiian twenty percent of the executives joined their firms in
recent years. These findings are consistent with Stewman's (1988) summary on
vacancy chain multiplier of 4 in various MILMs. In sum, past empirical findings
indicate diat a majority of top executive positions, about 80 percent or more, are
filled through internal promotions. Nonetheless, any filling of vacancies from the
outside reduces the promotion opportunities for the managers at lower levels, and
forecloses opportunities for promotion at the immediately lower level of the
organization. Vacancy-chain effects stop at that level. This leads to the following
hypotheses:
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
15
Hypothesis 1-a: The vacancy rate caused by exits at the immediately higher
rank will be positively associated with the likelihood of promotion of a
manager at the immediately lower rank.
Hypothesis 1-b: The vacancy rate caused by promotions at the
immediately higher rank will be positively associated with the likelihood of
promotion of a manager at die immediately lower rank.
Hypothesis 1-c: The rate of entry from outside the firm at the immediately
higher rank will be negatively associated with the likelihood of promotion
of a manager at the immediately lower rank.
The effects of exits, promotions, and entries on managerial job ladders Another
distinctive structural feature of internal labor markets is the vertical job ladder of
positions; that is, vacancies are filled through internal promotion within a defined
job ladder. Several researchers have reported the existence of multiple job ladders in
managerial hierarchies (e.g., Baron et al., 1986; DiPrete, 1987; Forbes and Piercy,
1991; Kanter, 1984; Stewman & Konda, 1983). While mobility is possible between
job ladders as will be discussed later, managers are much more likely to have
promotion opportunities within a given job ladder than outside it (Althauser, 1989;
DiPrete, 1987). Like a physical ladder, a job ladder poses barriers to entry except at
the bottom and constrains individuals on any particular ladder to remain on the
same ladder. Jumping from one ladder to another is a risky trick on job ladders, just
as it is on physical ladder (Rosenbaum, 1984). Consequently, all individuals at a
given position will have a similar career history on the same ladder, and as a result,
career history becomes irrelevant, net of present position, in predicting future
attainments (Rosenbaum, 1984). Along the career line, managerial jobs progress
with complementary skills and experience (Althauser & Kalleberg, 1981). A
manager's future career path is primarily determined by tire job ladder on which he
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
or she enters. With the presence of job ladders, the vacancies are more likely to be
filled by someone who is on the same job ladder than someone outside the job
ladder.
Explanation of the advantage of managerial job ladders in internal labor
markets can be convincingly provided by transaction cost perspectives. First, job
ladders may arise due to idiosyncracy. Along a job ladder, an orderly sequence of
firm-specific joint investments between firms and employees is made that effectively
lock them into an ongoing relationship (Aoki, 1988; Williamson, 1975). In an
ongoing employment relationship, opportunistic or strategic behaviors are reduced
since there is much surplus to be lost or significant fixed costs to be incurred in
terminating or restarting tire relationship (Wachter & Wright, 1990). In other words,
firm-specific managerial skills make managers a valued asset not easily replaced
from the outside, thus making turnover costly to both the firm and the managers
(Williamson, 1975). From a firm's point of view, job ladders therefore allow a firm
to obtain a well-trained workforce at a minimum cost. Additionally, given
opportunism and uncertainty in organizations, job ladders allow firms to monitor
managers' performance over an extended period of time and benchmark their
performance (DiPrete, 1987). From managers' point of view, job ladders confer
promotion opportunities that are protected from market competition. Also, by
residing on a given job ladder, managers are guaranteed a certain degree of job
security.
Second, job ladders may partially resolve die problem of information
impactedness (Williamson, 1975). If labor markets function perfectly, information
about workers should flow freely across and within firms. In this situation, there is
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
no need to have internal labor markets or job ladders in a given organization since
the market can identify the best candidates for any given job. However, there has
been a significant amount of research indicating market imperfections and
information impactedness within organizations (Williamson, 1975; March & Simon,
1958; Watcher & Wright, 1992). Because of this, information regarding performance
evaluation of a manager on a given job ladder sometimes may not be transmitted to
another division, department, or job ladder (DiPrete, 1987). Organizations may not
w ant to risk person-job mismatch by picking up somebody from other job ladder
they do not have enough information about.
Again, moving up the hierarchy on a given job ladder is likely to occur
through vacancies caused by either exits or promotions at higher levels, most likely
at one level higher. Those on the same job ladder will have higher probability of
filling the vacancy than those whose job is not located on the same job ladder. On
the other hand, entries at one level on the same job ladder are likely to reduce the
promotion opportunity at lower levels, most likely at one level below. Based on the
theoretical argument above:
Hypothesis 1-d: The vacancy rate caused by exits at the immediately
higher rank on a given job ladder will be positively associated with the
likelihood of promotion of a manager at the immediately lower rank.
Hypothesis 1-e: The vacancy rate caused by promotions at the immediately
higher rank on a given job ladder will be positively associated with the
likelihood of promotion of a manager at the immediately lower rank.
Hypothesis 1-f: The rate of entry from outside the firm at the immediately
higher rank on a given job ladder will be negatively associated with the
likelihood of promotion of a manager at the immediately lower rank.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Cross-job ladder movements While a large portion of career movements occur on a
given job ladder, it is also possible to observe between-job ladder mobility in
MILMs. Even though it is not always clear under what circumstances and in what
directions these between-job ladder movements take place in MILMs, firms move
managers from one function to another in certain cases. For example, a vice
president of marketing can move to a vice president of operations, or can be
promoted to senior vice president of corporate planning, while it is also possible for
him or her to be promoted to senior vice president of marketing and sales along the
same job ladder. Rather than promoting managers through narrow career routes
into higher executive positions, organizations sometimes strongly emphasize
breadth of experience as a requirement for promotions, and lateral or upward
moves through a diversity of jobs and departments are viewed as providing the
required preparation (Raghuram, 1992; Forbes & Piercy, 1991). Cross-functional job
rotations are also viewed as an aid to increase general rather than special learning
capabilities of employees (Aoki, 1988). More frequent and effective coordination
and integration caused by the exchange of personnel among different functions,
departments, and divisions are additional benefits (Aoki, 1988). Lastly, cross
functional or cross-job ladder movements can be an adaptation mechanism to
environmental changes (Kanter, 1984). With rapid expansion and technological
change, firms need to reorganize their organizational structure and create new
positions more frequently, thereby making career lines less meaningful. Under
these circumstances, lateral-cross unit advancements, generalist tracks, extra-
hierarchical leaps, dual ladders, and high-level analyst career paths are commonly
observed (Kanter, 1984, pp.118-124).
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
In this sense, MILMs, especially at the executive level, may deviate from
traditional functional-line ladder career systems that dictate strict movements along
a given job category. These variants of career systems can demonstrate strength at
the executive level, because the scope of executive tasks tends to be much broader
than those at middle or lower management. As Rosenbaum (1984) noted, while
some jobs are better than others for aiding an individual's promotion chances, the
range of advancement routes to a high-level position is likely to be far broader than
a single job ladder. Moreover, movements to a different job ladder are a critical
juncture or testing point for managers to demonstrate their ability to handle new
challenges and to keep themselves in contention for advancements (Martin &
Strauss, 1959). Aldiough not so frequent as within-job ladder movements, cross
functional or cross-ladder moves in MILMs provide a manager with an opportunity
to acquire broader skill bases and experiences. And that will help a manager to
move further up the hierarchy. Therefore:
Hypothesis 1-g: Lateral or upward movement to a different job ladder
during service as an executive in the firm will be positively associated with
the likelihood of promotion.
Location in the managerial hierarchy The shape of organizational hierarchies
imposes unavoidable constraints on the possibilities of career advancement (Baron
et al., 1986; Rosenbaum, 1984). Since most organizational hierarchies are shaped like
pyramids with declining numbers of jobs at progressively higher grades or levels
(Kanter, 1979; DiPrete, 1987; Stewman & Konda, 1983; Stewman, 1988), an
organizational pyramid must inevitably present barriers to the advancement of an
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
employee. To illustrate this, imagine that a firm has an executive structure looking
like bowling pins, 1 CEO at die top, 2 executive vice presidents, 3 senior vice
presidents, and 4 vice presidents on the bottom. Also assume that promotion takes
place from one level only to the next upper level and that the probability of exit at
each level is constant. Under this hierarchical structure, with 100% vacancies at die
upper level, die promotion probability of a vice president to a senior vice president
will be 0.75, and tiiat of an executive vice president to CEO will be 0.5. Declining
chances of promotion in tiiis structure can be easily observed. Although other
pyramidal structures may have die opposite distribution of promotion probability
along die location of structure, the example of the structure of the managerial
hierarchy described here is believed to be typical in MILMs and can take a form of
Ventiiri Tubes (Rosenfeld, 1992; Stewman, 1988; Stewman & Konda, 1983). Venturi
Tubes are a special form of die hierarchical structure where stacking the multiple
grade ratios (MGRs), calculated from die relative sizes of adjacent grade levels and
vacancies' arrivals at a level, show the shape of promotion opportunities for the job
structure. Given all the otiier factors being constant, the example of balling pins has
an increasing MGRs witii the increasing rank in the managerial hierarchy.
The pyramidal structure of managerial hierarchies is often characterized as a
"tournament structure" where many vice presidents compete against each other to
move up to higher positions (Lazear & Rosen, 1981; Lazear, 1991). Simple logic
dictates that die average manager's advancement must slow down as he or she
climbs up the hierarchy. Also, die higher stability of chief executive positions at the
summit of managerial hierarchies tiian diat of other managerial positions reduces
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
21
the promotion chance of managers at lower levels to this top position (Boeker, 1992).
Therefore:
Hypothesis 1-h: The lower die manager is located in managerial
hierarchies, the higher die likelihood of promotion.
Individualistic Models—
Human Capital And Social Capital Factors and Promotion
Individualistic career models are the dominant career models in the Unites
Sates. Individualistic models contend that individuals are the main agents
determining their own careers, and diat employers are highly responsive to
individuals' efforts (Rosenbaum, 1989). Individualistic models reflect America's
most prominent cultural ideology— individualism. In this model, careers depend
more on people knowing their own strengdis and interests, than on their knowing
much about organizations or career systems. Individuals may take actions to
change their skills and strengths and thereby affect the direction of their career path.
Examples of this approach are abundant in organizational behavior literature (Hall,
1976; Schein, 1964; Wanous, 1975). In this section, two major theoretical arguments
from human capital theory and social capital theory that have been developed in
different research traditions will be discussed. First, human capital theory, that has
been primarily developed in labor economics, will be examined. Second, as a
distinct concept differentiated from human capital, social capital theory will be
discussed. This theory has recendy emerged as a powerful theoretical paradigm in
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
sociology to explicate a variety of organizational phenomena. The focus of the
analysis is to investigate whether these approaches have particular relevance in
explaining career advancement among top executives in major U.S. firms.
Human Capital:
The human capital model derives from the tradition of labor economics
pioneered by Becker (1975). The human capital model is a widely studied and
accepted model in a variety of theoretical ramifications in social sciences.
However, direct empirical connections between human capital and promotion
have not been established (Sicherman & Galor, 1990). Human capital theory
posits that labor markets offer open opportunity and that managers'
attainments are largely a function of human capital which is defined as a
person's effort, ability, education, and training (Becker, 1975; Milgrom &
Roberts, 1992; Sicherman & Galor, 1990). Like physical capital, human capital is
increased by investment. Education and training are the primary investments
individuals can make in their human capital. And the cost of accumulating
human capital and the return to the investment can be shared by workers and
employers. By sharing this investment, the parties reduce the likelihood of
either party unilaterally terminating die employment relationship and imposing
on the other party a loss in his or her return (Becker, 1975; Hashimoto, 1981).
From the individual's point of view, human capital can be increased by making
investments of time and money in education and training (Becker, 1975; Blaug,
1976; Mincer, 1974). The human capital model assumes individuals' rationality
to judge if they have enough ability to justify their investments of time and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
23
tuition (Rosenbaum, 1989). Consequently, individuals will efficiently maximize
their returns. The human capital model therefore perfectly represents the
individualistic ideal that "individuals control the investments that affect their
careers (Rosenbaum, 1989)."
The human capital model, like other neoclassical economic theories, also
assumes efficient labor markets. Employers will pay what their employees are
worth; otherwise, employees will move to another employer who is willing to pay
their human capital value. The structural characteristics of managerial hierarchies
are mere representations of the imperfections of labor markets which will be
eliminated eventually by market processes (Rosenbaum, 1984). It should be noted,
however, that efficient market is possible only when human capital assumes such a
general nature that labor mobility across firms is costless. When human capital is
specific to tire firm, inter-firm labor mobility becomes difficult. With the presence of
firm-specific human capital, there are ways that firms and employees can devise an
efficient internal employment contract. Given market imperfections and specificity
of human capital, internal labor markets from the perspective of human capital
theory can be a rational choice between firms and employees.
The human capital model suggests that promotions serve two roles in an
organization. First, they help assign people to the roles where they can best
contribute to the organization's performance and success. Higher level jobs in
the hierarchy typically demand higher levels of knowledge or skill and involve
assignments and responsibilities where successes are more significant and
mistakes or failures are more costly (Milgrom & Roberts, 1992). Second,
promotions serve as incentives and rewards. Firms use promotion as a key
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
incentive mechanism for employees, since promotion is likely to be
accompanied by significant pay increases. This is due to the widespread
practice among U.S. firms of linking pay to a job or a job level, not to a person
(Baker et al., 1994a). Firms tend to compress within-job pay variance, especially
by compressing the pay increase for top performers in a given job level. Rather
than increase pay, firms tend to promote top performers in a job level to the
next level (Gibbs, 1994). Hierarchy is thus viewed as an intermediate variable
for translating tire value of an individual's hum an capital into earnings (Strober,
1990).
With certain exceptions in worker's internal labor markets (Carmichael,
1983; Abraham & Medoff, 1985), ability is tire most commonly used criterion in
U.S. firms in determining promotion, while seniority is the most common for
layoffs (Doeringer & Piore, 1971; Althauser & Kalleberg, 1981). This is
particularly true with managerial jobs. Given the way managerial jobs are
usually structured, it is crucial to have more productive people in higher-level
jobs. In standard economic language, the marginal returns to increasing either
the effort provided by a manager or the level of talent, ability, and relevant
experience that he or she brings to bear should be greater, the further up the
manager is in tire hierarchy (Baker et al., 1993). Firms want to put harder-
working and more talented managers in higher level positions, with the best
people at the top. Managers who have a high human capital value in terms of
skills, experience, training, and innate ability advance to the higher assignments
in the hierarchy (Milgrom & Roberts, 1992). Relative to the important
theoretical linkage between human capital and promotion, there have been very
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
25
few empirical studies to test the connection. The primary reason for the lack of
empirical effort appears to lie in the extreme difficulty of obtaining the objective
measures of employees' human capital value.
Human capital can be categorized into two types: general and firm-specific
human capital. General human capital includes education, training, and experience
that serve a general purpose. It is valuable in a broad set of employment
relationships, and so the employee must be fully compensated for the increased
productivity that it brings. In contrast, firm-specific human capital is valuable only
in the particular employment relationship (Becker, 1975; Prescott & Visscher, 1980;
Williamson, 1975,1985). Education and age, two general human capital variables,
are discussed first.
Education Human capital model proposes a specific rationale for the positive
correlation between time-dependent variables and productivity. First, human
capital model proposes that education level is positively correlated with income. It
specifies a particular mechanism through which this correlation results: Education
increases skills, and diis in turn increases productivity. Higher productivity is then
rewarded through higher earnings and higher status (Hartog, 1987; Strober, 1990;
Sicherman & Galor, 1990). One convincing advice from human capital model can be
summarized as "If you want a good job, get a good education (Stober, 1990)" There
is a large volume of research to support the prediction that high ranking managers,
especially top executives, are far better educated than ordinary Americans (Forbes &
Piercy, 1991; Kurtz et al., 1989; Swinyard & Bond, 1980; Useem & Karabel, 1986). For
example, in their extensive study on more than 11,000 newly promoted executives of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
26
the major U.S. firms, Swinyard and Bond (1980) reported that 89 percent of
executives in their sample graduated from college, and that 41 percent had earned
graduate degrees. Other studies report similar results (Forbes & Piercy, 1991; Kurtz
et al., 1989; Useem & Karabel, 1986). A suggested correlation between education
level and managerial productivity leads to die following hypotheses:
Hypothesis 2-a: Educational level will be positively associated with the
likelihood of promotion.
Age Human capital theory proposes mixed predictions for the relationship between
age and productivity, earnings, and promotions. One argument is that people who
are older have acquired more general and firm-specific experiences (Hartog, 1987;
Strober, 1990). Age is highly collinear with job experience and is sometimes used as
a proxy for experience (Felmlee, 1984; Sorensen, 1974; Tuma, 1976). A positive
correlation is expected between age and managerial productivity, and thus
promotion. Another argument posits the opposite prediction: Managerial
productivity declines with age due to the gradual decline in individuals' self
investments (Rosenbaum, 1984). A negative correlation is expected between age and
managerial productivity. This contradiction, however, appears to indicate an
inverted U-shape relation: A manager's marginal product increases with age to a
certain point, and then declines with age diereafter.
When managers are young, they can continuously invest their time and effort
in learning so that they can benefits later. After a certain age, however, investments
in human capital decline because expected returns decline. This inverted U-shape
relationship between age and productivity has been suggested and tested by several
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
27
researchers in human capital theory (Hartog, 1987; Medoff & Abraham, 1980). If
managerial productivity leads to promotion, then it can also be argued that there is
an inverted U-shaped relationship between age and the likelihood of being
promoted. Based on the discussion, the following hypothesis is presented:
Hypothesis 2-b: Age will have an inverted U-shape relationship with the
likelihood of promotion: The likelihood of promotion will increase with age
to a certain age, and then decrease with age thereafter.
A more important aspect of human capital can be found when it takes a firm-
specific nature. Firm-specific human capital is valuable only in the particular
employment relationship, but is less useful if managers leave to join other firms
(Becker, 1975; Williamson, 1975,1985). It is obtained through on-the-job training,
job-rotations, in-house education programs, company sponsored education
programs, and by-product learning through assignments to different positions.
Firm-specific human capital can take various forms: Knowledge of heuristics,
routines, operating procedures in the firm; knowledge of locally specialized
language usage, and of customers, suppliers, coworkers, and subordinates;
idiosyncratic skills in tasks peculiar to die firm; and membership in the social
networks within the firm and between tire firms and its suppliers and customers
(Milgrom & Roberts, 1992).
Other things being equal, managers who have acquired firm-specific skills
and experience will be more efficient in their current employment than elsewhere
(Becker, 1975; Williamson, 1985; Milgrom & Roberts, 1992). In this dissertation,
three different aspects of firm-specific human capital will be examined:
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
28
organizational tenure, criticality and breadth of managerial task knowledge, and
managerial productivity reflected in pay raises.
Organizational tenure Whereas age is considered as a proxy for general human
capital variable, organizational tenure is one of the widely used firm-specific human
capital variable (Strober, 1990). With tenure in the firm, managers acquire firm-
specific knowledge, skills, and experience. In studies testing human capital theory,
wage grows with tenure in the firm, because employee productivity grows with the
firm-specific training and experience (Abraham & Medoff, 1985; Hashimoto, 1979;
Jung & Magrabi, 1991; Koike, 1988; Topel, 1991). In other words, upward-sloping
tenure-earnings profiles reflect upward-sloping tenure-productivity profiles
(Abraham & Medoff, 1985). In other studies, however, a negative correlation was
suggested and supported in predicting promotion (DiPrete & Soule, 1986). As in
age, numerous studies demonstrate a non-monotonic relationship between
organizational tenure and the likelihood of being promoted (Medoff & Abraham,
1981; Miner, 1987; Sicherman & Galor, 1990; Zenger, 1992).
These results indicate strongly an inverted U-shape relationship between
organizational tenure and the likelihood of being promoted. It appears that the
same argument with age can also be made for the relationship between
organizational tenure and managerial productivity. Individuals' self-investments
and their return increase with tenure up to a certain point, but decrease thereafter.
Farther along the organizational tenure, the amortization of the investment in
human capital should be much higher because of imminent retirement. Both firms
and managers have less incentive to invest in human capital, since both parties may
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
29
not enjoy the return for the investment in the short-run before the retirement of the
managers. Therefore, managerial productivity has an inverted U-shape relationship
with organizational tenure. Likewise, tire likelihood of being promoted should also
have an inverted U-shape relationship with organizational tenure. Therefore:
Hypothesis 2-c: Organizational tenure will have an inverted U-shape
relationship with promotion likelihood of a manager, that is, the likelihood
of being promoted will increase with organizational tenure to a certain
point, and then decrease with organizational tenure thereafter.
Managerial task knowledge While education, age, and organizational tenure have
been studied as major human capital variables, the content of managerial task
knowledge has largely been neglected. As noted earlier, organizational success
depends on managers' ability to deal with environmental contingencies (Crozier,
1964; Hickson, Lee, Schneck, & Pennings; 1971; Mintzberg, 1983; Tushman &
Romanelli, 1983). And this coping ability with environmental contingencies is clearly
one feature of managers' human capital (Milgrom & Roberts, 1992). Several
components of its task environment create uncertainty for an organization, such as its
customers, suppliers, competitors, and the government (Porter, 1980; Thompson,
1967). Managers who have acquired knowledge base to effectively deal with
uncertainties created by elements of task environment will likely be assigned to higher
or more powerful positions, since those with relevant expertise may have significant
influence on strategic choices of a firm (Finkelstein, 1992).
Two aspects of managerial task knowledge are worth mentioning: criticality
and breadth. Experience in critical task areas provides managers with an
opportunity to learn about the key success factors of organizations (Sheridan et al.,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
30
1990). This is particularly so at higher levels of managerial hierarchies where
decision making, supervision, and leadership have more direct impact than those at
lower levels (Milgrom & Roberts, 1992). Criticality in turn depends on what
elements of the task environment the organization finds most problematic (Kanter,
1977).
Another dimension is breadth of a manager's knowledge in task
environments (Finkelstein, 1992; Forbes, 1987). Kotter (1988) identified breath of
knowledge as one of the most important attributes to be a senior manager. It
includes broad industry and organizational knowledge, broad relationships in the
firm and in the industry, and a broad functional track record. This breadth might be
enhanced by a number of different career experiences (Martin & Strauss, 1959). In
their policy capturing study, Forbes and Piercy (1991) found that breadth of
experience accounted for 69 percent of tire variance in the promotion ranking of
executives. Based on tire discussion on criticality and breadth of managerial task
knowledge, it can be argued that a manager with broad and critical knowledge base
in firm's task environments will advance to a higher managerial position. Hence:
Hypothesis 2-d: A manager's knowledge base in firm's task environment
will be positively associated with tire likelihood of promotion.
Pay raise Employee productivity has been studied as a key variable in determining
pay level and promotion in human capital theory. Although problems are
acknowledged, several organizational economists have attempted to measure
employee productivity. The most commonly used measure of productivity in these
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
studies is performance ratings (see Medoff & Abraham, 1981; Gibbs, 1994). Since
performance evaluation data are nearly impossible to obtain for upper-echelon
managers, certain compromises need to be made. Gibbs (1994), in his case study of
incentive mechanisms of an organization, used pay increase to predict promotion as
a proxy for employee's ability, which is accumulated human capital over time.
According to Gibbs, pay raises should reflect most closely the employee's market
value. Otherwise, he continues, the managers who receive the compensation below
market value will leave die firm. Consistent with this hypothesis, Gibbs
demonstrated that those with higher average increases were promoted both more
often and faster. Consistent with Gibbs (1994), the following hypothesis is
presented:
Hypothesis 2-e: The past pay increases of a manager will be positively
associated with tire likelihood of promotion.
Social Capital:
Social capital is conceptualized as the network of relationships among purposeful
social actors that are embedded in the social structure. Sociologists of networks
label this social capital which is distinctly differentiable from human capital
(Bourdieu, 1977; Coleman, 1988; Burt, 1992; De Graff & Flapp, 1988). Social capital is
distinct in two ways from financial and human capital. First, while financial capital
and human capital is owned by an individual or a firm, social capital is embedded
in a social actor's relationship with others (Coleman, 1988). Second, social capital
concerns rate of return in tire market production equation, while human capital
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
32
concerns the investment term in the market production equation. In other words,
while financial and human capital gets invested to create production capabilities,
social capital through relations with colleagues, friends, clients, government, and
leaders in other firms transforms financial and human capital into profit (Burt, 1992).
Social capital has been related to individual's organizational or professional
advancement in the labor market (Bourdieu, 1977; Granovetter, 1974; De Graff &
Flapp, 1988; Kanter, 1979; Useem & Karabel, 1986). Useem and Karabel (1986) assert
that social capital, primarily accumulated by upper-class background through elite
education, net of educational credentials, confers an important advantage in
climbing the corporate ladder. They hypothesized that those individuals who
possess line greatest amount of die relevant forms of capital, scholastic and social
capital, will be the most likely to gain access to the core of the American business
elite. Kanter (1979) contends that social capital, dirough formal and informal social
networks, allows an individual to gain access to promotion and organizational
resources.
External networking ability Networking ability refers to a manager's capability to
connect direcdy or indirectiy through social networks with the dominant actors in
business community. Networking ability, together witit network centrality as will
be discussed below, should be distinguished conceptually from human capital, even
though it sometimes interacts or correlates with certain aspects of human capital.
According to Kanter, a manager who is well connected to the dominant parties
outside the firm will have greater visibility to insiders by often bringing up crucial
information about activities outside die firm. Also if one has network ties with high
status people, one will sooner reach a high status position by obtaining pre-
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
33
socialization information to reach the position, or by learning the necessary
negotiation skills (De Graff & Flapp, 1988). Through this process, one can gain a
"little" but sometimes crucial help from those whom one knows. De Graff and
Flapp (1988) demonstrated that social capital has large effects on income attainment
independent of human capital. Other studies have also documented the
independent effect of social capital on job-related outcomes (Coleman, 1988; Useem
& Karabel, 1986). In sum, those who are better equipped with social capital will
more frequently employ formal or informal channels to enhance their socio
economic attainments (Boxman, De Graff, & Flapp, 1991).
Despite a widespread scholarly consensus that social capital is a powerful
mechanism for a manager to continue moving up the managerial hierarchy,
empirical evidence directly linking social capital variables to promotion is scarce.
The empirical studies discussed above were conducted in the context of job search,
hiring, and socio-economic attainment primarily measured by income and top
managers' cross-sectional profiles. Nonetheless, theory and evidence suggest that
networking ability, a key social capital indicator, can help a manager advance to
higher and more important positions in organizations. Thus, the following
hypothesis can be presented:
Hypothesis 2-f: External networking ability of a manager to dominant
actors, individuals and organizations in the business community, will be
positively associated with the likelihood of promotion.
Internal network centrality Another dimension of social capital is network
centrality. Whereas networking ability refers to a manager's capability to take
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
advantage of social networks outside organization, network centrality in this study
refers to a manager's position in a network structure inside organization. Network
centrality is defined as a function of a point' structural position within a set of
relations (Mizruchi & Bunting, 1981). Managers who are centrally located in the
network can take advantage of his social capital embedded in the relationships with
others better than those who are not. Centrality in a network should depend on
three criteria: (1) number of links to other points; (2) the intensity of the links; and (3)
the centrality of those with whom one is linked (Bonacich, 1972; Katz, 1953). Since
die level of analysis in diis study is an individual, network centrality of an
individual in the organization appears adequate as adopted by several researchers
in diis theoretical area (Brass & Burkhardt, 1992; Ibarra, 1993; Nelson, 1989).
Various of types of networks exist widiin organizations— workflow,
communication, and friendship (Brass & Burkhardt, 1992; Tichy, Tushman, &
Fombrun, 1979). Since widiin-firm friendship network for the managers is
inaccessible, it will be beyond die scope of diis study. Instead, two other types of
network will be discussed. Centrality in a workflow or communication network in
organizations increases a manager's power. A manager's power tends to be
increased via links witii powerful others (Brass & Burkhardt, 1992). For example,
direct links witii die CEO or widi die dominant coalition in an organization may be
positively related to one's power (Cyert & March, 1963). More specifically, being
central or getting closer to the center of network benefits a manager in the following
manner. First, network centrality increases a manager's visibility to the top manager
(Kanter, 1979). Second, centrality in workflow network decreases one's dependence
on others by having many alternative sources for acquiring inputs or distributing
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
35
outputs. Also, mediating the flow of work may provide a manager with control
(Brass & Burkhardt, 1992). Third, managers who are centrally located in the
communication network have potential access to, and control over, information and
thus are potentially powerful (Pfeffer, 1981). A manager in a central position in
networks in an organization can therefore mobilize, control, and direct resources
and information to his or her advantage in such important situations as promotion.
Those who are located in central positions in their workflow or communication
network in organization should have better chance to move up the hierarchy than
those who are not. Hence:
Hypothesis 2-g: A manager's network centrality in the current firm will be
positively associated with die likelihood of promotion.
The Tournament M obility M odel and Promotion
What principle of competition best describes the career system in
organizations? Can the organizational career system be characterized as random
process suggested by the garbage can model (Cohen, March, & Olsen, 1972), or can
it be better described as a rational process suggested by the tournament model
(Rosenbaum, 1984)? Some organization theorists take an anarchistic perspective on
organizational structure (Cohen et al., 1972; March & Olsen, 1976). Goals of
organizations are frequently problematic, technologies are frequently ambiguous,
and the performance of organizational personnel is often difficult to assess (Cohen et
al., 1972; March & Olsen, 1976). From this perspective, promotion should occur on a
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
36
random basis and past career history should have no predictive power. Promotion
can occur at any point of time regardless of individuals' ability, motivation, or their
winning or losing history.
From the opposite camp, as a proponent of rational process model,
Rosenbaum (1979,1984) has suggested tournament as the dominant competition
principle in organizational career systems. Rosenbaum argues that organizational
career systems are rational processes that resolve different mobility norms in an
efficient way. Rosenbaum's pioneering work on tournament models in career
mobility stems from earlier sociological models of occupational mobility: contest and
sponsored mobility (Turner, 1964). Contest mobility dictates that individuals,
regardless of origins or past positions, should have an equal or fair "opportunity" to
continue to compete for advancement. Human capital theory proceeds from the
same concepts as contest mobility. In contrast, sponsored mobility stresses
"efficiency." It prescribes drat selections occur as early as possible, even before die
competition starts, so drat the system can maximally benefit from the efficiencies of
specialized training and socialization dirough predefined career paths. Individuals
are selected for dieir ultimate careers very early and departures from these early
assigned careers are not permitted. Those who are selected for elite status are
separated from otiiers and are given specialized training and socialization. Internal
labor market tiieory, die most common version of structural theory, posits a
situation compared to that described by sponsored mobility. In reality, firms are
described as struggling between the two equally dominant career philosophies.
Rosenbaum (1979,1984) proposes die tournament model as a rational
resolution of die conflict between "opportunity" and "efficiency" by offering initial
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
"opportunity" to all employees while fostering "efficiency" by repeated selections to
remove employees from the tournament for top positions. In the tournament
mobility model, careers are conceptualized as a sequence of competitions, each of
which has implications for an individual's mobility chances in all subsequent
selections. Winners have the opportunity to compete for high levels, but they have
no assurance of attaining them; losers are permitted to compete only for low levels
or are denied the opportunity to compete any further at all. As in the contest model,
winners must continue competing in order to attain high levels, for there is no
assurance of continual advancement; but, as in the sponsored model, early selections
have irreversible consequences for losers.
Some advantages of tournament career system are noteworthy. First, a
tournament system allows both mobility norms to operate: "opportunity" for
individuals and "efficiency" for organizations. Second, a tournament produces a
signal or reputation of managers' ability by sorting out winners at earlier stage
(Spence, 1973). Early promotion is therefore critical for later promotion because it
clearly has an important objective signal where decision makers in the firm often
work with very subjective information and operate under conditions of bounded
rationality (March & Simon, 1958; Rosenbaum, 1989). This interpretation is echoed
by several economists: Favoring early winners can utilize maximally the information
value of the past performance. Meyer (1991) explains this succinctly. Suppose a
situation exists where two contestants, A and B, compete with one another for two
observation periods and person A wins tire first round. If the two are treated the
same in die second round of observation, independent of which one did better in the
first round, the firm gains nothing from having the extra observation on their
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
38
performance in the first round. Compared to significant monitoring and evaluation
costs, there is no benefits for the firm from observing the first round. The firm
should favor person A without having to wait to see the performance of both
persons in the second round.
Another explanation for signaling is derived from economics of reputation,
primarily developed in game theory (Kreps, 1990). Given that the employment
relationship can be characterized as repeated transactions between employer and
employee, opportunistic behaviors are minimized from both parties to reach an
optimal joint outcome, i.e., a Nash equilibrium (Kreps, 1990). Establishing
reputational capital at tire earliest career stage possible will have the greatest impact
on the employee's future career since that is die time when the length of the
remaining career is longest and so that reputational capital has its highest expected
value (Gibbons & Murphy, 1992). Later, when a person's reputation is stabilized by
repeated observations of performance, it will be harder to alter perceptions of
ability, and the benefit of an improved reputation will be shorter lived (Milgrom &
Roberts, 1992).
Rosenbaum (1979,1984) demonstrated strong support for the tournament
model. He found that those who were promoted early were much more likely to
receive further promotions and reach higher levels than those who were not. In fact,
only those promoted in tire earliest period possible were able to reach middle
management within 13 years. Subsequent empirical studies testing Rosenbaum's
tournament model, however, have found only a weak tournament effect on
promotion. Forbes (1987) found that after controlling for intervening attainments,
early promotions did not have significant effects on later attainments. Sheridan,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
39
Slocum, Buda, and Thompson (1990) found that fast starters attained only slightly
higher levels than the slow starters, and the differences were not statistically
significant.
Two major problems should be pointed out with regard to tournament
theory and its empirical philosophy. First, if revealing a manager's ability in an
early stage of tournament competition were valuable from the organization's point
of view, as Rosenbaum contends further, signaling of a manager's ability at later
stage would have equally or more important information value. Tournament theory
in its original form does not exclude diis possibility. A central principle of
tournament theory involves an important distinction between winners and losers in
each round in die tournament, not only at the earliest round (Rosenbaum, 1984).
However, the empirical testing procedure used by Rosenbaum (1979,1984) and its
results have been interpreted stricdy as an early winning theory. Although the
signaling effect of later success is not explicidy acknowledged, Forbes (1987)
suggests a round robin tournament, a modified version of tournament, indicating that
not only an early record of winning but also the latest attainment counts in
predicting future success. Therefore, it appears critical to examine the effects of later
success to predict future success in order to fully test the tournament model. Second
problem is tournament theory's assumption that competition takes place primarily
within a cohort group. All the empirical studies of tournaments deal with a single
cohort in a single organization (Rosenbaum, 1979,1984; Forbes, 1987; Sheridan et al.,
1990; Wakabayashi, Graen, Graen, & Graen, 1988). While this approach has the clear
advantage of controlling other organizational and exogenous factors, it neglects a
very important reality of competition among managers in organizations: Managers
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
compete with their own cohort, but also witii other cohorts (Vancil, 1987). For
example, a newly appointed vice president, say vice president A, who is the
forerunner of his or her cohort, must compete with other vice presidents at his or her
level. Suppose there is another vice president, vice president B, who is the
forerunner of his or her own cohort 5 years ahead of vice president A. If vice
president A is promoted to senior vice president position while vice president B
stays in vice president position, it can be said that vice president B lost the
tournament at that level, despite die fact that he or she has defeated all the other
competitors in his or her own cohort. At the executive level in particular, winning
the tournament in one's own cohort group does not guarantee further successes and
one m ust continue competing with other executives from other cohort groups.
Tournaments in organizations can thus be better described as different sub
tournaments taking place at each stage of grand tournament. This aspect of
competition must be integrated in to tests of tournament theory.
Witii these considerations in mind, in this dissertation, three tournament
variables are examined: Career velocity, passovers at the present level, and
passovers at the previous level. If the assumptions of the tournament model were
accurate, and if information regarding success at later stages were equally or more
valuable than success at earlier stages, one should expect that those who fall behind
the competition should experience limited promotion opportunities at the current
level and at the next level as well. Overall career velocity, encompassing the
broadest time span, denotes a manager's ability or performance in the past. It is the
only useful variable, especially when there is no possible way of directly tracing the
earlier successes of a manager in his or her cohort group. The number of passovers
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
41
in the current hierarchical level addresses the question, "Does the tournament
competition take place at each level in the managerial hierarchy now?" It simply
tests whether or not there is a tournament competition operating at each level of the
managerial hierarchy. Passovers at the previous level tap into the carry-over effect
of the liability experienced from the previous level (one level below) onto the current
level.
Overall career velocity Managers are viewed as more capable if they pave rapidly
advancing careers or if they are younger than their peers in their status level. Career
velocity has important information value of signal or reputation of manager's ability
as discussed earlier. High-velocity advancements indicate that managers have
quickly won many competitions, presumably because they are more capable (Martin
& Strauss, 1959; Rosenbaum, 1989; Kanter, 1977). Managers compete for career
success, their career histories reveal their successes in past competitions, and only
the most able can win several competitions. Thus career velocity indicates
managers' performance or ability in their past career. When ability is hard to
measure, as in the case of top managers, a manager's job history provides objective
indicators that are interpreted as signals of ability (Rosenbaum, 1989). Also, career
velocity is the only alternative variable of earlier successes when there is no possible
way of knowing who are the winners of their own cohorts at the earliest tournament
round. Managers in high status relative to his or her age or organizational tenure
are viewed as winners of past competitions and, consequently more able. More able
managers characterized by high velocity will have higher promotion chances than
those who have low velocity. Therefore:
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
42
Hypothesis 3-a: The overall career velocity of a manager will be positively
associated with the likelihood of promotion.
Passovers at the present hierarchical level While overall career velocity matters in
predicting the promotion chances of a manager at the current level, tournament
competition at the present level should also affect the promotion prospects of a
manager. A central question here is "Does tournament competition take place at
each level of the managerial hierarchy?" If tournament principle dominates in the
managerial hierarchy, as described earlier, one should expect a decreasing
promotion probability for those who pass over promotions at the current level. If
the probability of promotion is random no matter how many times a manager
misses promotion chances at the current level, it can be said that there is no
information value regarding who is winning or who is losing: Anybody, regardless
of winning or losing, will have an equal chance to move up die ladder, as might be
indicated by the garbage can model (Cohen et al., 1972). On the other hand, if the
tournament principle is an accurate characterization of career systems in
organizations, it dictates diat die more promotion opportunities a manager loses, the
less likely it is that a manager will move up the hierarchy. The principle of
tournament competition therefore leads to the following hypothesis:
Hypothesis 3-b: The number of surpasses by other managers at the present
level will be negatively associated with the likelihood of promotion of a
manager.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Passovers at the previous hierarchical level Do managers carry die liability of
being passed over for promotion from the previous level to the present level?
Passovers in the previous hierarchical level should matter in a round robin
tournament, a modified version of tournament in which not only an early record of
winning, but also the latest attainment counts in predicting future success. This is
consistent with Rosenbaum's (1984) tournament model. If the firm uses promotions
to sort managers on the basis of their ability, then those promoted quickly once
should be promoted quickly again (Gibbs, 1994). On the other hand, a manager who
is passed over for promotion at the previous level should have the lesser chance of
promotion at the present level. Therefore:
Hypothesis 3-c: The number of surpasses by other managers at the
previous level will be negatively associated with the promotion likelihood
at tire present level.
The Relational Demography Model and Promotion
Organizations can be characterized not only as an economic system where
efficiency is the primary concern, but also as a socio-psychological system where
relations, perceptions, and emotions among members of organization play an
important role (Katz & Kahn, 1981). Sociological and social psychological literature
have consistently documented how relational properties among social actors and
their perceptions and emotions affect various decisions in organizations. People
routinely classify themselves and others based on social categories such as age,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
44
gender, race, and status and form strong preferences for groups based on these
social categories, and these social categorizations, in turn, affect various
organizational behaviors and attitudes (Tajfel & Turner, 1986). Demographic
characteristics such as age, race, gender, education, and organizational tenure have
been related to selection (Mclntire, Moberg, & Posner, 1980), job performance
(Waldman & Avolio, 1986), communication (Zenger & Lawrence, 1989), and
turnover (O'Reilly, Caldwell, & Barnett, 1989; Wagner, Pfeffer, & O'Reilly, 1984).
Basically, from this perspective, demographic similarity provides a salient basis for
group membership and affects favorability of decisions in various settings.
The concept of organization demography in the extant literature, however,
primarily refers to the composition of group in terms of the distribution of basic
demographic attributes. In this dissertation, a central focus is given to a unique
aspect of demography, dyadic relational demography between the CEO and a
manager. Relational demography is defined as comparative demographic
characteristics of members of dyads or groups who are in a position to engage in
regular interactions in organization. Given that the CEO is generally the most
powerful decision maker and the highest of all employees in the firm, the CEO can
exert a significant influence on promotion decisions of executives. In deciding
whom to promote, the dyadic comparative demographic similarities or
dissimilarities between the CEO and a manager may play a crucial role through
similarity-attraction (Byrne, 1971) and social categorization processes (Tajfel &
Turner, 1986).
The similarity-attraction connection has been well documented by a number
of researchers in social psychology. Interpersonal attraction is affected by
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
45
demographic similarity (Byrne, Clore, & Worchel, 1986) and by pastime preference
(Werner & Parmelee, 1979). As relating concepts, social integration, and frequency
of interaction or communication (Lincoln & Miller, 1979; Zenger & Lawrence, 1989)
are influenced by demographic similarity. In sum, relational demography can affect
perceptions and attitudes through both interpersonal attraction and the frequency of
interactions. Positive affect or liking is a natural outcome of similarity in
demographic backgrounds between members of a dyad.
Although the effects of relational demography have been studied in other
settings such as marriage practices (Guttentag & Secord, 1983), crime rates (Maxim,
1985), and public attitudes (Glenn, 1969), none of the studies has focused on
promotion. The closest examples can be found in hiring or performance evaluation
decisions. In performance evaluation and hiring contexts, which are similar settings
to promotion decision, relational demographic similarities are associated with
favorable biases. In experimental and field research on hiring decisions, studies
demonstrated a positive relationship between applicant-rater similarity and the
perceive quality of die applicant. In studies on performance evaluation,
demographic similarity was associated with higher performance ratings and higher
personal attractions on the part of superiors for subordinates. Westphal and Zajac
(1995) found, in tiieir study on new director selection of major firms in the U.S., that
powerful CEOs appointed new board members who are demographically similar
and therefore more sympathetic to the CEO.
In promotion decision contexts, die similarity-attraction and social
categorization principles dictate that CEOs should prefer demographically similar
managers for higher positions. The promotion of demographically similar managers
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
46
can also facilitate interaction and communication (Zenger & Lawrence, 1989). Easy
communication is particularly important when disagreement or conflict arises
among top managers. Interestingly, this aspect of maintaining good relationships
has been observed as a promotion criterion in white-collar work by internal labor
market researchers (Osterman, 1984; Pinfield & Berner, 1994).
Whether they are aware of the relational demographic literature or not, top
management succession researchers document predecessor-successor similarity
(Smith & White, 1987). Top management team researchers also addresses the role of
"interpersonal dynamics" in top management structure (Hambrick & Mason, 1984).
Extrapolating from Barnard's (1938) work, Michel and Hambrick (1992) argue that
top managers are selected and promoted on the basis of "fitness" between their
backgrounds and the general condition of personnel characteristics rather than
solely of formal competence. The fitness question, they continue, involves such
matters as education, experience, age, sex, race, personal distinction, prestige,
nationality. The general method of maintaining an executive organization is to
select and promote executives such that a general condition of compatibility of
personnel is maintained. Perhaps often, and certainly occasionally, managers
cannot be promoted or selected because they cannot function, because they "do not
fit," where there is no question of formal competence. Relational demography
theory indicates this "interpersonal dynamics" is greatly influenced by CEO-
executive similarity in many ways.
Consistent with the theoretical arguments and supporting empirical findings,
four demographic characteristics are selected due to their salience in the context of
dyadic CEO-manager demographic similarity: age, elite school, functional
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
background, and outside status similarity. These demographic characteristics have
been widely studied in organization research.
Age similarity Age is one of the most studied demographic attributes in
organizational demography research (Hambrick & Mason, 1984; Jackson et al., 1991;
Wiersema & Bantel, 1992). Theory and research suggests that cognitive flexibility,
willingness to change, and risk taking decrease with age (Wiersema & Bantel, 1992).
Also, managers from same age cohorts tend to hold common work-related and non
work related experiences that lead to shared attitudes and belief structures.
Through the attraction and categorization process that is triggered by age similarity
as a salient demographic cue, one may argue that other factors being equal, CEOs
will select similar aged managers to higher positions.
Functional background similarity Executives with common primary experience in
a particular functional area tend to develop a similar schema, dominant logic, or
"belief structure" relevant to strategic decision making (Dearborn & Simon, 1958;
Hambrick & Mason, 1984; Hitt & Ireland, 1985). Walsh (1988) argues that functional
background influences how the problem is recognized and information is processed.
Given common knowledge bases and work experiences, individuals with similar
functional orientations should diagnose strategic issues in a comparable maimer and
prefer similar solutions (Bantel & Jackson, 1989). For these reasons, functional
background similarity will affect attraction and categorization process. In addition,
CEOs will prefer functionally similar executives in higher positions in order to
execute strategies in a more cohesive way.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Educational affiliation similarity Similar attitudes, philosophies and cognitive
styles may be shaped by similar educational backgrounds (Holland, 1973).
Executives sharing common school affiliations may posses similar leadership and
communication styles. Common school affiliation may indicate similar attitudes
regarding such basic issues as tire direction of organizational change and the
allocation of organizational resources (Hambrick & Mason, 1984). For example,
Collins (1979) found that graduates from elite schools such as Ivy League
institutions are more similar on relevant dimensions such as behavioral styles,
values, membership in social and professional networks than those who are not
from Ivy League. Attendance at an Ivy League school provided a meaningful basis
for psychological group membership. Thus, again, educational affiliation similarity
triggers attraction and categorization process drat leads to the favoring of such
managers in promotion decisions.
Outside status similarity Executives with litde experience in the firm may share a
common identity as relative "outsiders," even drough they may differ in their prior
experiences (Westphal & Zajac, 1995). Moreover, executives and CEOs hired from
outside the firm may recognize opportunities for change in existing strategies and
operational norms (Dalton & Kesner, 1985). New CEOs from outside are more able
and likely to execute dramatic changes of organizational structure and strategic
orientations dian CEOs promoted from within (Tushman & Romanelli, 1985; Keck &
Tushman, 1993; Wiersema & Bantel, 1992; 1993). Given that executives with long
tenure in the firm are likely to be more resistant to change (Finkelstein & Hambrick,
1990; Wiersema & Bantel, 1992,1993), CEOs may want to form a political coalition to
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
49
execute their new plans and visions with executives from outside who are more
amenable to change. Thus, CEOs from outside may have an incentive to promote
newly hired executives to higher positions.
Based on die arguments developed above, it can be hypodiesized that CEOs
favor executives in the higher positions who possess similar demographic
characteristics across all four dimensions. Therefore:
Hypothesis 4-a: A manager's age similarity with the CEO will be positively
associated with the likelihood of promotion.
Hypothesis 4-b: A manager's functional background similarity with the
CEO will be positively associated with the likelihood of promotion.
Hypothesis 4-c: A manager's educational affiliation similarity with the
CEO will be positively associated with the likelihood of promotion.
Hypothesis 4-d: A manager's outside status similarity with the CEO will
be positively associated with the likelihood of promotion.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
50
CHAPTER 3: DATA AND METHODS
Sample
The sample for the empirical analyses consists of 1576 top managers collected
from 92 manufacturing firms drawn from 3 industries: 32 firms from the computer
industry, 34 firms from the chemical industry, and 26 firms from the natural gas
industry. The list of die firms in the sample is in Appendix 1. The data were drawn
from a population of the largest firms in each industry for up to 13 fiscal years
between 1978 to 1990 as long as the data were available. The total number of
observations was 8978; diat is, diere were 8978 executive-year datapoints. The
sample was confined to relatively large firms because data on top managers of
smaller firms is often inaccessible, and a relatively homogeneous set of firms were
required to ensure comparability. Because of these choices, however, empirical
results may reduce generalizability of the findings.
All corporate officers holding vice president or higher positions in each firm
were included in the sample so long as their demographic and career data were
available from die secondary data sources noted below. In rare cases, directors or
general managers of a division were also included in the sample if they are also on
the list of corporate officers in die firm's or public data sources such as proxy
statement or Dun and Bradstreet's Reference Book of Corporate Management. Top
managers who cannot expect any more promotions by the nature of their positions
were excluded from die sample. These are (a) CEOs; and (b) Chairmen who had
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
stepped down as CEOs due to retirement. The criterion of including corporate
officers holding vice president or higher positions was chosen for three reasons: (a)
corporate officers at such high level positions are a selective group of people in
executive suite who are distinctively separated from other middle or lower level
managers in terms of legal rights and responsibility; they have to fulfill different
career expectations and tend to move on separate career ladders than lower-level
managers; (b) from vice president or above, corporate officers are eligible to be
elected for the CEO who is ultimate leadership of die business enterprise (Lazear &
Rosen, 1981); and (c) practically, it is impossible to include all the managers below
vice presidents due to data unavailability.
Data on firm characteristics were collected from Compustat, Standard and
Poor's Register of Corporations, Directors, and Executives, and The Million Dollar
Directory of Corporations. Data on top managers were collected from annual
corporate proxy statements, Dun and Bradstreet's Reference Book of Corporate
Management, Standard and Poor's Directory of Directors, Who's Who in Finance and
Industry, Who's News in die Wall Street Journal Index, and corporate annual reports.
Stock data were drawn from CRSP to estimate die stock option values granted to top
managers. When stock data were unavailable from CRSP, other secondary data
sources were counted on, including Standard and Poor's Stock Guide and Compustat.
Measures: Dependent Variable
Promotion What constitutes "a promotion" is always unclear (Martin & Strauss,
1959; Rosenfeld, 1992; Baker et al., 1994a). Lateral movements with pay increase or
better looking new titles, assignments to a position with higher responsibility and
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
52
wider span of control, or any type of movement that creates better feelings for the
manager may be considered as a promotion (Rosenbaum, 1984; Veiga, 1973).
However, because of self-reporting biases, any perceptual or self-report measure of
promotion is problematic (Veiga, 1973). In this study, a more strict and objective
measure of promotion was used. Promotion was identified exclusively on the basis
of status enhancement in terms of rank or level in the managerial hierarchy to avoid
any confusion or ambiguity (Rosenbaum, 1984; Baker et al., 1994a). A promotion
event in this study was measured as the dichotomous variable, 0 or 1, indicating
whether or not a manager's rank or level in managerial hierarchies was enhanced at
the end of each year during the period 1978-1990. While it is possible to measure
promotion in different ways, this operationalization has clear advantages: (1) the
rank or level categories in organizations often have extremely concrete referents
(Rosenbaum, 1984); and (2) the rank or level categories provide the most important
status and authority distinctions within organizations (Evans, 1975).
A great amount of care was taken to make sure if the identification of
promotion events was accurate. First, all die official tides are sorted in the
descending order of importance as follows: CEO, chairman, president, vice
chairman, chief operating officer, executive vice president, senior vice president,
group vice president, vice president, and director or general manager at divisional
level. Since each firm's executive hierarchy is different from one another, not all the
sample firms have above 10 level categories.
Promotion events were identified from annual corporate proxy statements or Dim and
Bradstreet's Reference Book of Corporate Management. The wording or sentences for the
event in proxy statements were also analyzed. Any event in which a manager
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
attained a new level in an organization was determined as a promotion if proxy
statements contained any wording that used "promotion," or "promoted" for a
given manager's new assignment. Most of the promotion events were
straightforward to determine. One example would be an obvious upward move
from vice president to president of the company. In all, 626 events of promotion
were determined by this process alone.
In cases, ambiguity arose regarding whether the movement was actually
lateral rather than upward. In these instances, additional secondary data sources—
Who's Nexus published in Wall Street Journal Index, Standard and Poor's Register of
Corporations, Directors, and Executives, or corporate annual reports— were used for
cross-validation. Who's Nexus in the Wall Street Journal Index was particularly useful
to cross-validate promotion events, since it contains firms' official announcements of
executive promotions. Of die 47 events that were ambiguous, 25 of them were
identified as promotions after this second process. The remaining 22 cases were
eliminated from die sample due to unresolved ambiguity. As a result, a total of 651
promotion events was identified from tiiis procedure and included in the sample.
Measures: Control Variables
Industry Since it is possible tiiat some industries observe more frequent promotions
than otiiers, I controlled for possible industry effects by creating industry dummy
variables. Two dummy variables were defined for the two industries in the study.
Using firms in die computer industry as die base case, one dummy was created for
the chemical industry, and anotiier for natural gas industry.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Firm performance Firm performance can affect promotion opportunities and
influence executive career advancement. It has been well explored in the
management succession literature (Harrison, Torres, & Kukalis, 1988; Reiganum,
1984; Puffer & Weintrop, 1991; Wiersema & Bantel, 1993). In times of poor
performance, managers associated with that performance may become scapegoats
and be replaced (Grusky, 1961; Boeker, 1992). Executives will have more promotion
opportunities if the vacant positions are filled with internal promotions. If vacancies
are filled with outside recruits, then executives may not benefit from these
vacancies. On the other hand, when performance is good, they may have a better
opportunity of getting promoted as a reward for their achievement. Therefore, no
specific direction of die effect of firm performance on promotion is determined. In
diis study, firm performance was measured using accounting Return on Assets
(ROA). Return on assets was computed as die percentage of earnings before
extraordinary items to total assets and lagged one year. These data were obtained
from the Compustat.
Firm size Firm size has been suggested as a correlate of internal promotion rates for
executives (Dalton & Kesner, 1985; Guthrie & Olian, 1991). Organization size has
also been suggested and tested as one of the determinants of internal labor market
which increases the probability of promotion within (Pfeffer & Cohen, 1984; Baron et
al., 1986; Davis-Blake, 1986). Firm size was computed from the log of sales reported
in Compustat and was adjusted for 1987 dollar value.
Firm growth Past empirical studies have consistendy shown that firm growth
provided additional promotional opportunities (Wlioley, 1985; Stewman & Konda,
1982; Rosenbaum, 1979,1984). Firm growth was measured as the five year average
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
55
percent of sales growth, computed from sales data reported in Compustat and
adjusted for 1987 dollar value.
New CEO Chief executive succession tends to be associated with a fundamental
change in strategy and organizational structure including replacing and promoting
the managers (Tushman & Romanelli, 1985; Keck & Tushman, 1993). Higher
promotion opportunities are expected with the new CEO. A dummy variable was
created to indicate a chief executive succession in a given year.
Left-censoring Consistent with Baum and Oliver (1991), a dummy variable for left-
censored managers was created. In order to examine if the left-censored
observations in the dataset had systematically different promotion rates as a result
of having assumed executive position before the observation period, a dummy
variable was created, coded 1 for managers who began to serve vice president or
higher positions prior to 1978, and 0 for managers who began to appear in the risk
set since 1978 during the observation period.
Measures: Independent Variables
Exit, promotion, and entry rates at the immediate higher rank Vacancy rate caused
by the exits at the immediate higher rank was computed from the number of exits
divided by the number of positions available at that level. Vacancy rate caused by
the promotions and entries at the next higher rank were computed the same way:
the number of promotions or entries divided by the number of positions at that
level.
Exit, promotion, and entry rates at the immediate higher rank on a job ladder Jobs
in most internal labor markets are structured around the functions (Milgrom &
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Roberts, 1992; DiPrete; 1987). In order to identify job ladders at executive level, the
present function of a manager is first determined. Eight present functional
categories were identified and coded in the data: production/ operation, human
resources/industrial relations, accounting/finance/data processing,
marketing/ sales/purchasing, legal, R&D, corporate planning/general management,
and government/public relations. In order to identify narrowly defined job ladders,
these eight functions were not collapsed into three functional backgrounds as done
by researchers in tire past (e.g., Hambrick and Mason, 1984; Smith & White, 1987;
Westphal & Zajac, 1995). Instead, a more precise coding scheme of eight functions
was maintained. If a vacancy is created in the same functional position in the
immediate higher rank, it is assumed to have taken place in the same job ladder. Job
movement across different functions can exist, but the data show significantly fewer
such job moves. Among tire promotion events identified, less than 6% were
identified as moving up to different functional positions. Therefore, using
functional categories to identify job ladders above appears to be valid.
Vacancy rates caused by exits in the next higher rank in a given job ladder
were defined as the ratio of the number of exits to the number of same functional
positions available at the next higher rank. Vacancy rates due to promotions and
entries at the next higher rank in the given job ladder were calculated the same
way — the ratio of tire number of promotions or entries to the number of the same
functional positions available at tire next higher rank.
Location in managerial hierarchies This variable was computed as the percentage
of individuals in a firm's executive hierarchy with higher official titles than a focal
executive (Finkelstein, 1992). The CEO was rated 0 on this measure, and the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
executive at the lowest rank was rated highest. Given the pyramidal nature of the
managerial hierarchy, one should expect a positive correlation between this variable
and the likelihood of promotion.
Educational level Educational level was coded into three categories: college,
master, and doctorate degree (Medoff & Abraham, 1981; Cannings, 1988; Zenger,
1992). Since more than 94% of die managers in the sample had college or higher
education, high school education category was less meaningful for a separate
category and therefore was excluded from the analytic models. Inclusion of high
school education did not change the results. Two dummies were defined for the
two educational levels above college education. Using individuals with college
education as tire base case, first dummy was created for master's degree and second
dummy for doctoral degree.
Age Date of birth was recorded and then subtracted from the current yearend to
compute age. To test nonmonotonic relationship with the likelihood of promotion, a
squared term of age also was added to die model.
Organizational tenure The year the person first joined die company was recorded
and used to determine lengdi of tenure with die organization at the end of each year
(Hartog, 1987; Medoff & Abraham, 1981; Miner, 1987). As with age, a squared term
of organizational tenure was added to the model to test nonmonotonic relationship
with the likelihood of promotion.
Managerial task knowledge This variable is a combined measure of criticality and
breadth of managerial knowledge base. Criticality of managerial knowledge was
computed as follows on the basis of die measure developed by Finkelstein (1992).
First, four major types of environmental requirements were specified that
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
correspond to different conditions in task environment of organizations: inputs,
outputs, throughputs, and regulatory concerns. Key environmental requirements
were assessed by counting the number of articles over the time period of interest
cited in die Funk & Scott Predicasts that emphasize each of the four categories of
environmental requirement. Second, all functional areas were identified that
managers had direct experience in using Dun and Bradstreet's Reference Book of
Corporate Management and corporate annual proxy statements. Third, criticality of
die managerial knowledge was assessed by matching functional experience with
environmental requirements: inputs— personnel, exploration, purchasing; outputs—
sales, marketing, and product R&D; tiiroughputs — operations, accounting, process
R&D; and regulatory concerns — governmental service, law (Hambrick, 1981). Then
the actual indicator of criticality of managerial knowledge was calculated by
summing the proportions of citations in each environmental requirement area in
which a focal manager had corresponding functional experience. The range of this
indicator is from 0 to 1.
Breadth of managerial knowledge was computed by counting die numbers of
different functional areas a focal manager had experience in (Finkelstein, 1992). This
indicator is a broader measure of experience that does not limit itself to only those
functional areas deemed important by the Funk & Scott criterion. This indicator
ranged from 1 to 5. Both criticality and breadth of managerial knowledge were then
standardized (mean=0, std. dev.=l) and averaged to produce an overall measure of
managerial task knowledge. Crombach's alpha reliability index of this variable was
0.78.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Pay increase In order to compute pay increase, the total compensation of managers
had to be determined. There are several components to a top manager's
compensation, such as salary, bonus, stock options, and long-term incentive plans.
In this study, total compensation was defined as the sum of salary, bonus, and the
value of stock options granted to the managers. Stock options were valued using the
simplified version of Black-Scholes' (1973) model (Jensen & Murphy, 1990; Kerr &
Kren, 1992) as the following.
Option Value = price X shares X [exp(-df)N(Z) - exp(-rt)N(Z - st)],
where
price = the exercise price of an option,
shares = the number of shares granted,
d = the average dividend yield over the previous 5 years,
t = the time to the expiration of the option, either 5 or 10 years,
when the expiration date was not available, 10 years was assumed,
N = the standard normal probability distribution function,
r = the risk-free interest rate, based on 5 and 10 year average yields
on U.S. government securities,
s = the stock return variance for the previous 5 years, and
Z = (r-d+s2/2)X(f/sf).
All die relevant data were drawn from annual corporate proxy statements and
CRSP. To make the comparison possible, all compensation and other dollar
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
measures were expressed in 1987 dollars using the Consumer Price Index as a
deflator. Pay raises were then computed as the average percentage increase in total
compensation for the last two years. Then, each manager's pay raises were adjusted
for other managers' pay raises by subtracting the average pay raise for all the other
managers from the focal manager's pay raise to give the relative pay raise. This was
chosen because most firms link a certain portion of executive pay to firm
performance, so that unadjusted pay may significantly inflate or deflate in extreme
years. This adjustment method also avoids biases due to sudden fluctuations in pay
for all the executives in a given year for other reasons than pay-performance
linkages, and the artificial inflation of relative pay raise of a focal manager if a ratio
measure is taken from an average pay increase of near 0 values.
External networking ability The networking ability of a manager was computed as
the average score of three indicators: corporate board membership, average board
rating, and elite school connection. First, corporate board membership was counted
as the total number of corporate boards of directors a manager sat on (Finkelstein,
1992). Research on directorship has suggested that managers may use board
memberships to establish and maintain ties with other important people in the
business elite (Allen, 1984). This network ties to outside dominant actors in the
business community, either organizations or individuals, will benefit a manager by
allowing better access to information regarding firm's task environment that serves
to strengthen his or her power base, by learning negotiation skills for a higher
position, or by enhancing firm's legitimacy in the institutional environment.
Second, a manager's networking ability also derives from the qualitative
aspect of the network ties he or she is connected to. For this indicator, average
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
board rating was computed by die average stock rating from Standard & Poor's Stock
Surveys for all corporations of which a manager was an external director. This
indicator also refers to die prestige of the board membership (Finkelstein, 1992). A
more prestigious directorship usually means a network connection through board
membership to the firms tiiat are larger, higher-performing, and better known to the
public.
Third, another important aspect of networking ability stems from a
manager's educational background (Zwigenhart, 1992). Attending prestigious
colleges and universities is a form of scholastic capital, but it also means that one has
developed potentially valuable social contacts, and thus has accumulated social
capital as well as scholastic capital (Useem & Karabel, 1986; Zwigenhart, 1992). Elite
school attendance serves as die basis for later friendship networks, networks that
not only allow for comradeship and companionship, but contacts that are at times
beneficial to one's career (Zwigenhart & Domhoff, 1991). A dummy variable was
created for tins diird indicator on the basis of whether a manager graduated from an
elite undergraduate or a graduate school. A list of elite educational institutions was
in Appendix 2 on the basis of Finkelstein (1992). Three indicators were then
standardized (mean=0, std=l) and averaged to produce an overall measure of
networking ability of a manager. Crombach's alpha reliability of this variable was
0.67.
Internal network centrality Network centrality is measured as dummy variable
whether an executive is also sitting on die board of directors in the firm. Although
there could be other ways of measuring network centrality, board membership of
die present firm was chosen for die following reasons. First, board members are
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
representative of the utmost inner circle of the dominant coalition in organizations
(Cyert & March, 1963; Mace, 1971; Finkelstein & Hambrick, 1990). Located at the
heart of organization, as the ultimate decision body, board members interact with
one another to make key strategic decisions on a regular basis. Second, more than
any other group of employees, board member-corporate officers have a direct link to
the CEO. Linkage to the most powerful actor in the network of the organization is
one of the key determinants of network centrality (Burt, 1992). Third, board
membership is an objective clear-cut criterion that differentiates board members
from all the other managers. Strong identification, loyalty, and intensive relationship
with other members in the board are the necessary results. Intensity of the links at
the highest level of organization is another key measure of network centrality (Burt,
1992; Bonacich, 1972; Katz, 1953). Fourth, board members can get an access to other
central actors in organization by taking advantage of their positional power of
mediating information from other nodes in the workflow or communication
networks to the ultimate decision body, the board.
Overall career velocity Consistent with Kanter (1977) and Rosenbaum (1984), career
velocity was computed from the ratio of managers' total number of positions held in
the firm divided by their firm tenure. In the total number of positions, lateral
assignments as well as upward assignments are included, since lateral assignments
also constitute a certain image of "fast tracker" — a sign of success (Veiga, 1981,
1985). Another measure of career velocity as the ratio of total number of promotions
divided by firm tenure was tested. It produced the same results only with slightly
less significance.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
63
Number of surpasses by other managers at present level This measure tests the
question, "Does a tournament exist at the present level?" Measuring tournament is
tricky at the executive level. While constraining the analysis to within-cohort group
has an advantage in terms of controlling other exogenous and organizational factors,
it is less meaningful and unrealistic in explaining the tournament at the executive
level. Executives m ust compete not only with their own cohort, but also with other
cohorts at each level.
To address this aspect of tournament competition, a new measure of winning
(or losing) die tournament had to be developed. The total number of promotions by
other managers who have stayed shorter than, or equally long, as a focal manager at
die present rank is counted. Constraining promotions of other managers who have
stayed shorter than, or equally long, as a focal manager makes a more precise
comparison of winning (or losing) possible. For example, if manager A has been
serving as vice president for two years, while manager B is promoted to senior vice
president in his or her fifth year, it is not clear whether manager B wins the
competition. One need to wait to see if manager A is promoted until his or her fifth
year. On the other hand, manager A who is promoted in his second year can be
viewed as a winner than manager B who is staying at die current rank for five years.
Therefore, die number of surpasses by diose who stayed shorter, or as equally long,
as a focal manager can be regarded as equivalent to the number of losses in
tournament at the present level: The more one loses, the less one is likely to move up
the hierarchy.
Number of surpasses by other managers at previous level While the number of
surpasses by other managers at die present level measures the existence of
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
tournament competition at the present level, the number of surpasses by other
managers at the previous level refers to the tendency of winning in the immediate
past competition. It was measured as the total number of surpasses by other
managers who stayed shorter, or equally long, as a focal manager at previous level.
It is expected that the higher die value of diis variable, the less the likelihood of
promotion.
Age similarity For age similarity, the absolute value of the difference between a
manager's age and die CEO age is computed. This measure, I Si - Sj I , is a special
form of the original entropy ratio [ 1 / m E (Si - S j)2 J1 /2 used by relational demography
researchers (O'Reilly, Caldwell, & Barnett, 1989). When there are only two
individuals in the formula (in this case, the CEO and a manager), the original
formula becomes I Si - Sj | (m in the formula should be 1, because n is the number of
individuals except a focal member in the team). All the similarity measures here are
a special form of more general demographic distance formula when there is only
two individuals in the team.
Educational affiliation similarity In this study, educational affiliation similarity
was based on whether or not a manager graduated from elite schools including Ivy
League. A dummy variable is created to indicate whether both a manager and CEO
had an elite educational background, either in undergraduate or graduate. It was
coded 1 if a manager and CEO both had elite educational background, 0 otherwise.
A list of elite school is in Appendix 2 according to Finkelstein (1992).
Outside status similarity To measure outside status, a dummy variable was created
and coded as 1 if a manager and the CEO are both outsiders, 0 otherwise. Outside
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
65
status was defined as a status of not employed by the company or any of its
subsidiaries prior to joining the company (Westphal & Zajac, 1995).
Functional background similarity Functional background was categorized into 8
areas: production/operation, human resources/industrial relations,
accounting/finance/data processing, marketing/sales/purchasing, legal, R&D,
corporate planning/general management, and government/public relations.
Functional backgrounds of each manager were coded up to 5 functions to derive a
more precise similarity measure. Consistent with the method of computing
similarity of proportions in different categories (Farjoun, 1994), an Euclidean
distance (or dissimilarity) measure of functional background was defined as the
following.
D = V £ (MGRpj - CEOpj)2 /2
i
where
MGRpj: proportion of a manager's total career spent in functional area j,
0< MGRpj <1,
CEOpj: proportion of CEO's total career spent in functional area j,
0< CEOpj <1,
j = 1, 2, 3,..., 8 (number of possible functions to compare),
The distance measure D can vary from 0 to 1, 0 indicating perfect similarity
and 1 indicating no-similarity. Then, die similarity measure S is defined as the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
negative value of D to denote the similarity. The similarity measure S is a
continuous measure that varies from -1 to 0, -1 indicating no-similarity and 0
indicating perfect similarity. More explanation of the distance measure D and the
similarity measure S may further aid understanding. For example, if a manager has
spent 20 years in production/operation, the proportion of the manager's experience
in this function is 1, and 0 for the other functions. If the CEO has spent 20 years in
production/operation and 5 years in marketing/sales, the CEO's proportion of
experience in production/operation is 0.8 and that in general management is 0.2.
Now the Euclidean distance of functional background between a manager and the
CEO will be V((1.0-0.8)2 + (0-0.2)2)/2 = 0.2. In this example, the similarity measure S
is -0.2, indicating relatively a high level of similarity between a manager and the
CEO since the value is close to 0.
This measure has definite advantages over the functional background
similarity measures used in other studies (e.g., Westphal & Zajac, 1995) for the
following reasons. Whereas previous measures used only one dominant function to
determine similarity, this measure allowed up to five functions, almost all possible
functions a manager can go through his or her entire career. Additionally, whereas
previous measures were categorical, this measure is continuous, thus, allowing more
variations in the measure. Reducing 8 functional backgrounds used above to 3
categories according to Hambrick and Mason (1984) produced similar in pattern but
weaker results. 8 category system in this study showed a significant improvement
in predictive power.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Data Analysis
67
Event history analysis was employed to model the occurrence of promotion.
Event history models can be carried out with either continuous-time models or
discrete-time models. The problem of this study is well-suited to the discrete-time
methods described by Allison (1984), because data are reported in yearly intervals
and the exact timing of the events is relatively unimportant. As long as appropriate
controls are introduced, discrete event history models can be estimated with logistic
regression equations using maximum likelihood methods (Allison, 1984;
Yamaguchi, 1991).
Discrete-time models with maximum likelihood methods offer a adequate
approximation of continuous-time models, which estimate the instantaneous hazard
rates, where the conditional probability of event occurrence is small as in this study
(Clogg & Eliason, 1987). The mean of the promotion rate in this study is slightly
higher than 0.07, which well satisfies Clogg and Eliason's criterion value of 0.1.
Further, this technique is able to deal with large number of ties without introducing
bias in parameter estimation and adequately takes into account right-censored
observations (Allison, 1984; Yamaguchi, 1991). The hazard rate of being promoted
can be modeled in die following logistic regression form:
!°g[Pi(t) / (1- Pi(t)] = a + PkXik (t),
where
pi(t): conditional probability of being promoted in year t,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
X ikS : time-varying covariates hypothesized to influence the "risk" or
likelihood of being promoted,
PkS : estimated parameters (notations in bold represent vectors).
The left hand side of the equation represents a logit, or log-odds. The model
assumes that for any manager in the population, the odds of experiencing the event
of promotion at time t are exp[pkXik (t)] times baseline group log-odd exp“, which is
equal to log[po(t) / (1- po(t)]. p;(t) increases monotonically with PkXik (t) and can
assume any value between zero and one. In this model, an executive's likelihood of
being promoted is updated over time as the values of control variables and time-
dependent covariates change. To deal with the potential biases caused by left-
censored observations, a dummy variable was created for the managers who began
to serve vice presidents or higher positions before the beginning of the observation
period, i.e., before 1978 in this study. To deal with potential autocorrelation
problems among time-varying covariates for a given executive, dummy variables
were created for each year during the observation period.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
69
CHAPTER 4: RESULTS
General Description of MILMs
Before I proceed to report empirical results from event history analyses, I will
discuss some basic features of MILMs found in this study. Several aspects of
executive mobility were briefly addressed in this section regarding entries, exits, and
promotions at each level of managerial hierarchies. Table 1 reports frequencies and
percentages of entries, promotions, and exits at each level. '
A total of 380 entries from outside was identified into all levels of executive
hierarchy in the sample. Therefore, a strict form of internal labor market, which
allows entry only at the bottom, was not found. Executive internal labor markets of
the firms in this study show a certain degree of openness to external labor markets at
all levels. There were more frequent entries into lower levels than into higher levels.
To see if this pattern is simply due to the pyramidal nature of managerial
hierarchical structure which is characterized by disproportionately larger number of
positions available lower down in the hierarchy, entry rates of higher versus lower
level subgroups were computed. Higher ranked subgroup was defined as a group
of executives holding tire title of COO or higher (Jackson et al., 1991). This group
includes CEOs, chairmen, vice chairmen, presidents, and COOs. Lower ranked
subgroup was defined as a group of executives holding the title of executive vice
president or lower. This group includes all levels of vice presidents, executive vice
presidents, senior vice presidents, group vice presidents, vice presidents, division
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
70
managers, and subsidiary presidents. Entry rates of each subgroup were computed
as die ratio of total number of entries over the total number jobs available in each
subgroup.
The entry rate into the higher ranked subgroup was 2.2%, and that into the
lower ranked subgroup was 4.1%. This statistic indicates that positions at higher
levels in MILMs are better protected from external labor markets than those at lower
levels. Positions at lower levels tend to be more exposed to external market
competition. Vacancies at the higher levels are more likely to be filled by internal
promotions than vacancies at the lower levels. Although a strong form of internal
labor market closure was not observed in tire sample, a lower rate of entry with
higher levels indicates that firms operate a semi-strong form of internal labor market
closure for the executives. The top echelons of tire managerial hierarchy are better
protected from market competition than the lower echelons.
On the other hand, promotions occur at all levels of managerial hierarchies
with equal probability: There was no significant difference in promotion rate
between higher and lower ranked subgroups: The promotion rate of the lower
ranked subgroup was 7.0%, whereas that of die higher ranked subgroup was 7.3%.
This result indicates that tenure at lower levels is shorter, other factors being
constant, taking outside hiring into account. A more rigorous test will be conducted
later in event history analysis since the simple comparison of promotion rates
between the two groups can be misleading: There may be others factors to
determine promotion rate including exits, promotions, and newly created jobs at
higher levels.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
71
T ab le 1. F re q u e n cies a n d P ercen tag es o f E n tries, P ro m o tio n s, a n d E xits by H ierarch ical L evel
T y p e of M obility
Levels
Entries Prom otions Exits
Total N u m b er
of Jobs
(Executive-Year)
CEO
C h airm an
P resident
Vice C h airm an
C O O
19 (1.8%)
1 (0.5%)
21 (4.4%)
6 (2.1%)
1 (1.2%)
48
(2.2%)
1 (0.5%)
59 (12.3%)
13 (4.5%)
12 (14.0%)
85
(7.3%)
57 (5.4%)
46 (23.0%)
38 (7.9%)
52 (17.9%)
1 (1.2%)
194
(8.8%)
1055 (100%)
200 (100%)
479 (100%)
291 (100%)
86 (100%)
Exec. VP
Sen. VP
G ro u p VP
VP
D ivision M gr. or
L ow er
48 (4.1%)
90 (3.7%)
22 (3.7%)
171 (4.4%)
1 (1.1%)
332
(4.1%)
85 (7.2%)
104 (4.3%)
64 (10.9%)
291 (7.5%)
6 (6.9%)
566
(7.0%)
142 (12.1%)
267 (11.1%)
49 (8.3%)
289 (7.5%)
6 (6.9%)
753
(9.3%)
1178 (100%)
2411 (100%)
588 (100%)
3858 (100%)
87 (100%)
Total 380 651 844 10233
One interesting pattern emerges from the comparison between entries and
promotions in tire sample. Entries are consistently less frequent than promotions
across all levels. The odds of choosing an outside candidate against an inside
candidate are about one-to-two (37%), meaning that for every three vacancies, one
vacancy is filled by outside personnel. These odds are much higher in comparison
with one-to-four odds (25%) that have been commonly found in past research on
vacancy chain multiplier in MILMs (e.g. Rosenfeld, 1992; Wholey, 1985; Forbes &
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Piercy, 1991). This strongly indicates that executive internal labor markets are more
open to external market than suggested by past research. Although the final
conclusion is dangerous to be made, several explications are possible for this
finding. First, firms in the sample have grown relatively rapidly with 11% annually
during the observation period. They may have needed to secure necessary new
managerial skills from outside into the growing or new areas. Second, managerial
capital especially at the executive level may have a more general scope and may be
transferred more easily across organizational boundaries than that at the lower
management. Examples of outside hire are abundant in management succession
literature.
Exits occur equally frequently across all levels only with a slight variations.
Overall, exits occur more frequently than entries and promotions. This is partly due
to the inclusion of all die exits of top managers in the sample regardless of the type
of exit whether it was retirement, voluntary, or non-voluntary turnover. However,
die total number of vacancies filled by entries and internal promotions (N=1030)
exceeds the total number of exits (N=844). This result indicates that, during the
observation period in this study, die firms in this sample created more jobs than
filling out vacancies caused by exits. Some of them are considered to be job
upgrades as indicated by Wholey (1985). Firm growth, reorganization, and firms'
practice of upgrade promotion may be able to explain this finding.
Table 2 reports the distribution of promotion by its type, frequencies and
percentages. Type of promotion was categorized into two forms, incremental or
non-incremental promotion, depending upon whether a manager is promoted to the
immediately higher position in an orderly manner, or is promoted to higher
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
positions skipping the intermediate levels. Incremental promotion refers to a
manager's status upgrade of one rank at a time in the managerial hierarchy. There
is no skipping of intermediate levels to advance to a higher position in incremental
promotion: Managers advance only one level up at a time. On the other hand, non-
incremental promotion refers to a manager's status upgrade of at least two ranks.
With non-incremental promotions, managers advance to new positions at least two
levels higher than their previous positions, skipping the intermediate levels.
In Table 2, titles in rows denote positions before promotion while titles in
columns denote the positions after promotion. To help understanding, one example
of promotion is selected. In the case of promotions from vice president to executive
vice president, there are 37 incremental promotions and 23 non-incremental
promotions. From Table 2, one may conclude that all the promotions in this cell are
non-incremental since promotions from vice president to executive vice president
skipped group vice president and senior vice president. This is simply not true.
Since each firm's executive hierarchy is different from one another, a promotion
from vice president to executive vice president in some firms is an incremental
promotion with only one step up at a time, whereas the same promotion in other
firms is a non-incremental promotion with at least two steps up skipping group vice
president or senior vice president. Therefore, promotion from vice president to
executive vice president could be either incremental or non-incremental depending
upon what hierarchical system a firm has. The frequencies at divisional manager or
lower levels are remarkably low, since this category is not fully represented in the
sample. Only those listed in the public data sources are included.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 2. Frequencies and Percentages by Types of Promotion
Previous Position
Promoted Position CEO Chairman President V. Chm. COO Exec. VP Sen. VP Group VP VP Total
Incremental
1 (100%) 1 (100%)
Chairman Non-lncremental
0 0
Incremental 58(100% ) 1 (100%) 59(100% )
President
Non-lncremental 0 0 0
Incremental 5(56%) 2(100% ) 2 (100%) 9 (69%)
Vice Chairman
Non-lncremental 4 (44%) 0 0 4(31% )
Incremental 3 (75%) 0 7 (100%) 1 (100%) 11 (92%)
COO
Non-lncremental 1 (25%) 0 0 0 1 (8%)
Incremental 4 (67%) 1 (50%) 49 (100%) 26(100% ) 8(100% ) 88 (97%)
Exec. VP
Non-lncremental 2 (33%) 1 (50%) 0 0 0 3 (3%)
Incremental 1 (33%) 0 11 (92%) 1 (100%) 8 (89%) 79(100% ) 100(96% )
Sen. VP
Non-lncremental 2 (67%) 0 1 (8%) 0 1 (11%) 0 4 (4%)
Incremental 0 0 4 (80%) 3(100% ) 1 (100%) 19(56% ) 20 (100%) 47 (73%)
Group VP
Non-lncremental 1 (100%) 0 1 (20%) 0 0 15(44% ) 0 17(27% )
Incremental 0 0 5 (50%) 1 (100%) 2(100% ) 37 (62%) 165 (91%) 47 (100%) 247 (85%)
VP
Non-lncremental 0 0 5 (50%) 0 0 23 (38%) 16(9% ) 0 44(15% )
Division Manager
Incremental 0 0 0 0 0 0 0 0 4 (80%) 4 (67%)
or Lower
Non-lncremental 0 0 0 1 (100%) 0 0 0 0 1 (20%) 2 (33%)
T otal Incremental 72 (88%) 4 (80%) 78 (92%) 32 (97%) 19 (95%) 135 (78%) 185 (92%) 47 (100%) 4 (80%) 576 (88.5%)
Non-lncremental 10(12% ) 1 (20%) 7 (8%) 1 (3%) 1 (5%) 38 (22%) 16 (8%) 0 1 (20%) 75(11.5% )
As in Table 2, among a total of 651 promotion events included in this study,
576 promotions (89.5%) were identified as incremental, and 75 promotions (11.5%)
were identified as non-incremental. About nine out of ten promotion events follow
an orderly pattern of status enhancement of one level. This pattern is consistent
across all levels of the executive hierarchy, although incidences of non-incremental
promotion tend to rise at lower levels. Table 2 provides one important feature of
career mobility of top managers in internal labor markets~an orderly movement up
die hierarchy associated with a progressive development of knowledge and skill. It
also taps into a basic notion chain effect in MILMs. When there is a vacancy caused
by the creation of a new position, exit, or promotion, the vacancy is more likely to be
filled by somebody at the immediate lower level than somebody from farther down
the hierarchy. Because Table 2 simply shows die general patterns of executive
promotion in MILMs, a more rigorous test for this orderly pattern of career
advancement of top managers will be conducted by using event history analysis
later.
Results of Event History Analyses
Table 3 presents means, standard deviations, and correlations among the
study variables. Hypotheses were tested in a series of logistic regressions, an
adequate analytic technique for discrete-time events such as promotions in this
study. For a base model, in Table 4-a, Model 1 includes all the control variables,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 3. Means, Standard Deviations, and Correlations for all Variables in the Study (N=8978)
Variables Mean S.D. 1 2 3 4 5 6 7 8 9 10 11 12 13 14
I . Promotion (Dummy) 0.07 0.26
2 . Firm Size (Log Sales) 7.25 1.29 0.01
3 . Firm Performance (ROA) 5.39 6.27 0.02 0.02
4 . Firm Growth Rate (5 Year Average) 11.12 30.21 0.03 -0.19 0.13
5 . New CEO 0.09 0.28 0.14 -0.02 -0.15 -0.01
6 . Censoring Dummy 0.60 0.49 0.03 0.04 0.09 0.04 -0.02
7 . Exit Rate at Immediate Higher Level 0.05 0.13 0.06 0.03 -0.10 -0.04 0.09 -0.02
8 . Promotion Rate at Immediate Higher Level 0.06 0.22 0.05 0.01 -0.02 0.00 0.35 0.01 -0.08
9 . Entry Rate at Immediate Higher Level 0.03 0.14 0.00 -0.06 -0.08 0.03 0.01 -0.02 0.01 0.03
10 . Exit Rate at Immediate Higher Level 0.03 0.19 0.08 0.05 -0.05 -0.03 0.07 0.01 0.52 -0.04 0.01
on Job Ladder
1 1 . Promotion Rate at Immediate Higher Level 0.02 0.14 0.06 0.05 0.01 0.00 0.21 0.00 -0.04 0.47 -0.01 -0.01
on Job Ladder
12 . Entry Rate at Immediate Higher Level 0.01 0.09 0.00 -0.01 -0.05 0.01 -0.01 -0.02 0.00 -0.02 0.47 0.01 -0.01
on Job Ladder
13 . Cross-Job Ladder Movement 0.05 0.21 0.05 -0.02 -0.04 -0.04 0.03 0.02 0.01 0.01 -0.01 0.04 0.03 0.01
14 . Location in Managerial Hierarchy 34.35 19.27 0.02 0.04 -0.02 0.02 0.01 -0.13 0.04 0.00 0.04 0.00 0.03 0.01 -0.12
(% Higher Positions)
15 . Master's Degreee (Dummy) 0.39 0.49 0.01 0.10 0.00 -0.02 -0.01 -0.06 0.01 0.00 0.00 -0.03 0.01 -0.02 0.02 0.03
* Correlations greater than 0 .0 3 are significant at p < 0.0 5
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 3. Continued
Variables Mean S.D. 1 2 3 4 5 6 7 8 9 10 11 12 13 14
16 .Ph. D. Degree (Dummy) 0.06 0.24 0.00 0.07 0.01 -0.03 -0.01 0.00 -0.01 -0.01 -0.03 -0.01 -0.03 0.01 0.01 -0.03
17 . Age 53.01 7.34 -0.04 0.19 0.00 -0.14 0.00 0.34 0.03 0.01 -0.05 0.03 0.00 -0.01 0.00 -0.11
18 . Age Squared/10 28.63 7.70 -0.05 0.18 0.00 -0.14 -0.01 0.33 0.03 0.01 -0.05 0.03 0.00 -0.01 0.00 -0.10
19 . Organizational Tenure 18.51 11.37 -0.01 0.22 0.10 -0.16 0.00 0.30 0.01 0.01 -0.08 0.02 0.02 -0.05 0.06 -0.06
20 . Organizational Tenure Squared/10 47.23 46.79 -0.01 0.22 0.09 -0.14 0.00 0.26 0.01 0.00 -0.06 0.03 0.02 -0.03 0.05 -0.06
21 . Managerial Task Knowledge -0.06 0.88 0.07 0.13 0.05 0.00 0.02 0.07 0.00 0.01 -0.02 0.03 0.05 0.00 0.17 -0.13
22 . Relative Pay Raise 0.06 19.08 0.08 0.02 0.02 -0.01 0.00 -0.01 0.02 -0.02 0.00 0.01 0.00 -0.02 0.04 -0.03
23 . Networking Ability (Outside) -0.11 0.64 0.09 0.25 -0.01 -0.05 0.02 0.09 0.03 -0.01 -0.02 0.03 0.00 -0.01 0.08 -0.14
24 . Network Centrality (Inside) 0.27 0.44 0.07 0.09 0.02 0.00 0.02 0.20 0.01 -0.02 -0.04 0.04 -0.01 -0.01 0.10 -0.29
25 . Overall Career Velocity 0.28 0.20 0.05 0.04 -0.04 0.04 0.02 -0.26 0.02 0.00 0.10 0.04 0.02 0.06 0.03 -0.04
26 . Number of Surpasses by Other Managers 0.81 0.90 -0.06 0.09 -0.03 -0.06 0.05 -0.05 0.04 0.06 -0.03 0.04 0.04 -0.02 -0.04 0.27
at Present Level
27 . Number of Surpasses by Other Managers 0.75 1.07 -0.07 0.20 0.04 -0.09 -0.02 0.19 0.01 0.05 0.01 0.01 -0.01 0.05 0.00 0.15
at Previous Level
28 . Age Similarity to CEO 8.12 6.60 0.02 -0.14 0.03 -0.02 0.10 -0.17 -0.02 0.05 -0.01 -0.02 0.03 -0.01 0.01 0.08
29 . Functional Background Similarity to CEO -0.33 0.14 0.09 0.11 0.05 -0.04 0.08 0.03 0.03 0.01 -0.03 0.10 0.09 0.04 0.06 -0.14
30 . Elite School Similarity to CEO 0.59 0.74 0.02 0.07 0.08 0.09 0.00 0.00 0.02 -0.02 0.00 0.00 -0.02 -0.04 0.03 0.03
31 . Outside Status Similarity to CEO 0.003 0.05 0.06 -0.03 -0.08 -0.01 0.17 -0.05 0.03 0.04 0.03 0.03 0.02 -0.01 0.00 0.01
* Correlations greater than 0 .0 3 are significant at p < 0.05
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 3. Continued
Variables 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
16 . Ph. D. Degree (Dummy) -0.21
17 . Age -0.07 0.01
18 . Age Squared/10 -0.07 0.01 0.99
19 . Organizational Tenure -0.07 0.01 0.49 0.49
20 . Organizational Tenure Squared/10 -0.08 -0.07 0.50 0.5 0.96
21 . Managerial Task Knowledge -0.04 0.01 0.07 0.07 0.19 0.17
22 . Relative Pay Raise 0.01 -0.01 -0.09 -0.1 -0.02 -0.03 0.01
23 . Networking Ability (Outside) 0.33 0.09 0.09 0.09 0.00 0.00 0.09 0.01
24 . Network Centrality (Inside) 0.00 0.04 0.21 0.21 0.08 0.08 0.08 0.03 0.28
25 . Overall Career Velocity 0.06 0.06 -0.33 -0.33 -0.48 -0.38 0.06 0.05 0.07 0.00
26 . Number of Surpasses by Other Managers 0.00 0.02 0.09 0.09 0.09 0.08 -0.05 -0.08 -0.09 -0.17 -0.16
at Present Level
27 . Number of Surpasses by Other Managers 0.02 -0.03 0.05 0.05 0.15 0.12 0.08 -0.04 -0.01 -0.19 -0.09 0.17
at Previous Level
28 . Age Similarity to CEO 0.04 -0.05 -0.52 -0.49 -0.22 -0.22 -0.07 0.05 -0.05 -0.03 0.17 -0.05 -0.07
29 . Functional Background Similarity to CEO -0.06 -0.03 0.10 0.09 0.14 0.16 0.35 0.00 0.10 0.12 0.04 -0.08 0.05 -0.10
30 . Elite School Similarity to CEO 0.03 0.03 -0.03 -0.04 -0.07 -0.07 0.00 0.03 0.05 -0.05 0.06 -0.04 -0.01 -0.01 -0.03
31 . Outside Status Similarity to CEO 0.00 -0.01 -0.05 -0.04 -0.07 -0.05 -0.03 0.00 0.01 0.01 0.15 -0.03 -0.02 0.06 0.03 -0.03
* Correlations greater than 0.0 3 are significant at p < 0 .0 5
oo
79
industry, size, performance, growth, and new CEO. Size, performance, growth rate,
and new CEO were positively associated with the likelihood of promotion in the
expected direction. No industry effects were found in this base model.
In Table 4-a, the effects of all the structural characteristics of MILMs
hypothesized in the study were tested. In Model 2 in Table 4-a, Hypothesis 1-a
through Hypothesis 1-c were tested. Controlling for other industry and firm
variables, exit rate and promotion rate at the immediate higher level were positively
associated with the likelihood of promotion as predicted in Hypothesis 1-a and
Hypothesis 1-b. Entry rate at the immediate higher level was not significantly
associated with promotion in Model 2. Hypothesis 1-c was not supported. The
result that entries at the immediate higher level did not affect promotions at the
immediate lower level may suggest that entries may not be linked with exits. Firms
may hire managers from external market not for the purpose of filling the vacancies
caused by exits, but for the positions newly created. Firms may operate career
systems for the managers in a way that they do not create disadvantage for the
managers who are already in the managerial hierarchies by hiring somebody from
outside.
In Model 3, all the vacancy factors of MILMs are included to test Hypothesis
1-a through Hypothesis 1-f simultaneously. Exit rate and promotion rate which
were significant in Model 2 became insignificant with job ladder vacancy variables.
Significance of exit rate (p<0.001) in Model 2 was reduced greatly (p<0.05) in this
simultaneous testing. Promotion rate which was significant in Model 2 became
insignificant in Model 3. Instead, exit and promotion rate on the same job ladder
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
80
Table 4-a. Logistic Regressions of Executive Promotion (N=8978)
V a ria b le s M o d e l 1 M o d el 2 M o d el 3 M o d el 4
Intercept -3.63*** -3.76*** -3.68*** -3.92***
(0.31) (0.31) (0.31) (0.33)
C ontrol V ariables
Chem ical Industry (Dummy) -0.14 -0.16 -0.16 -0.18
(0.09) (0.10) (0.10) (0.10)
N atural G as Industry (Dummy) -0.17 -0.16 -0.15 -0.18
(0.12) (0.12) (0.12) (0.12)
Firm Size (Log Sales) 0.08* 0.08* 0.07* 0.07*
(0.03) (0.03) (0.03) (0.03)
Firm P erfo rm an ce (ROA) 0.02** 0.03*** 0.03*** 0.03***
(0.007) (0.007) (0.007) (0.008)
Firm G ro w th (5 Y ear A verage) 0.002* 0.002* 0.002* 0.002*
(0.001) (0.001) (0.001) (0.001)
New CEO 1.17*** 1.16*** 1.09*** 1.11***
(0.11) (0.12) (0.12) (0.12)
In d ep en d en t V ariables
C ensoring Dum m y 0.10 0.10 0.11
(0.09) (0.09) (0.10)
(Internal Labor Market Factors)
Exit R ate a t Im m ediate H igher Level 1.31*** 0.73* 0.71*
(0.26) (0.34) (0.34)
Prom otion R ate a t Im m ediate Higher Level 0.71* -0.27 -0.39
(0.35) (0.34) (0.36)
Entry R ate a t Im m ediate H igher Level 0.18 0.26 0.22
(0.28) (0.32) (0.33)
Exit R ate a t Im m ediate H igher Level 0.72*** 0.70***
on Jo b Ladder (0.20) (0.21)
Prom otion R ate a t Im m ediate Higher Level 0.54* 0.53*
on J o b Ladder (0.21) (0.21)
Entry R ate a t Im m ediate H igher Level -0.28 -0.25
on Jo b Ladder (0.55) (0.55)
C ro ss-Jo b Ladder M ovem ent 0.78***
(0.17)
Location in M anagerial H ierarchy 0.007**
(% H igher Positions) (0.002)
C hi-square 276.57*** 312.34*** 337.51*** 365.80***
-2 Log Likelihood 4427.98 4396.24 4383.85 4346.32
U n standardized reg ressio n co efficien ts reported; standard errors in p a ren th eses.
*** p < 0 .0 0 1 ; ** p < 0 .0 1 ; * p < 0 .0 5 ; + p < 0.1
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
were significant (p<0.001 and p<0.05 respectively). Hence Hypothesis 1-d and
Hypothesis 1-e were supported in Model 3. Entry rate at the immediate higher level
was not significant, although it was in the expected direction. Hypothesis 1-f was
not supported. Again, firms may hire managers from outside for the newly created
positions, not for the vacancies caused by exits. Firms may operate career systems
for the managers in a way that they do not create disadvantage for the managers
who are already in the managerial hierarchies by hiring somebody from outside.
Summarizing the results from Model 3, it seems that firms operate a strong form of
job ladder career systems in the executive hierarchy. Firms tend to promote
managers along a given job ladder structured around functions. While vacancies at
the higher level increase promotion chances for the managers at lower levels, the
vacancies are more likely to be filled by someone on the same job ladder defined by
functions than someone on the other ladders.
In Model 4, Hypothesis 1-g and Hypothesis 1-h were tested. Hypothesis 1-g
that lateral and upward movement to a different job ladder during the service as an
executive in tire firm will be positively associated with the likelihood of promotion
was supported (p<0.001), controlling for other internal labor market factors.
Although these cross-functional movements are not so common relative to within-
functional movements, firms seem to move top managers to different functions to
help them build necessary skills and experience for higher positions in the future.
Hypothesis 1-h that the lower the manager is located in managerial hierarchies, the
higher the likelihood of promotion was supported (p<0.01). This result validates the
existence of pyramidal hierarchical structure at executive level.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Hypothesis 2-a through Hypothesis 2-h were tested in Model 5 through
Model 8. The results are listed in Table 4-b. All the general and specific human
capital variables were tested in Model 5 with other industry and firm variables. In
Model 5, there was a weak positive association between graduate education and
promotion (p<0.1). Ph.D. degree did not affect promotion positively. Getting a
Ph.D. degree did not help the managers moving up the hierarchy, probably due to
its specificity of the knowledge base. Since top managers are required to be
equipped with a certain level of breadth of knowledge, acquiring an expert
knowledge in a narrow area would not help advancing to higher top positions. In
this study, managers with Ph.D. were most likely to be found in R&D functional
areas. As predicted in Hypothesis 2-b, age had an inverted U-shape relationship
with promotion (p<0.001). The age level with the highest promotion likelihood was
53. Promotion chance increased with age up to 53, and then decreased thereafter.
Contrary to Hypo diesis 2-c, organizational tenure had a U-shape relationship with
promotion (p<0.01). Probability of promotion decreased with organization tenure
up to 20 years, and then increased thereafter. It seems like that top managers enjoy
the benefit of being new to the firm and start to bear the liability of being old in the
firm. The finding indicates that firms maintain a certain degree of openness to
external labor market by hiring and favoring new managers in order to
accommodate new technologies or managerial skills tiiat are deemed necessary for
the successful operation of the firm. This finding, however, should be interpreted
with caution. Organizational tenure in this study may have a spurious relationship
with outside status similarity as will be discussed in Model 14. Hypothesis 2-d that
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 4-b. Logistic Regressions of Executive Promotion (N=8978)
83
Variables Model 5 Model 6 Model 7 Model 8
Intercept -13.84*** -21.27 -3.22*** -13.62***
(2.20) (51.48) (0.32) (2.20)
Control Variables
Chem ical Industry (Dummy) -0.06 -0.33* -0.18 -0.08
(0.10) (0.15) (0.10) (0.10)
N atural G as Industry (Dummy) -0.09 -0.11 -0.17 -0.07
(0.12) (0.18) (0.12) (0.12)
Firm Size (Log Sales) 0.06+ 0.01 0.01 0.002
(0.03) (0.05) (0.03) (0.03)
Firm P erfo rm an ce (ROA) 0.02** 0.02* 0.02** 0.02**
(0.007) (0.01) (0.007) (0.008)
Firm G row th (5 Y ear A verage) 0.002+ 0.001 0.002+ 0.001
(0.001) (0.001) (0.001) (0.001)
New CEO 1.16*** 1.62*** 1.14*** 1.11***
(0.11) (0.15) (0.11) (0.12)
Independent Variables
C ensoring Dum m y 0.30** 0.23 -0.01 0.18+
(0.10) (0.17) (0.09) (0.10)
(Human and Social Capital Factors)
M a ste r's D egree (Dummy) 0.16+ 0.22 + -0.01
(0.09) (0.13) (0.09)
Ph. D . D egree (Dummy) 0.02 -0.09 -0.16
(0.18) (0.26) (0.18)
Age 0.44*** 0.30* 0.46***
(0.09) (0.14) (0.09)
Age S q u ared /1 0 0.04*** 0.03** 0.05***
(0.008) (0.01) (0.008)
Organizational T enure -0.04** -0.01 -0.04**
(0.01) (0.02) (0.02)
O rganizational T enure S q u a re d /1 0 0.01** 0.001 0.01**
(0.004) (0.001) (0.004)
M anagerial T ask K now ledge 0.28*** 0.28*** 0.24***
(0.05) (0.07) (0.05)
Relative Pay In crease 0.01***
(0.003)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
84
Table 4-b. Continued
Variables Model 5 Model 6 Model 7 Model 8
N etw orking Ability (O utside) 0.34***
(0.06)
0.35***
(0.07)
N etw ork C entrality (Inside) 0.41***
(0.09)
0.50***
(0.10)
C hi-square 356.13*** 293.34*** 353.67*** 435.91***
-2 Log Likelihood 4284.99 1961.92
4358.90
4213.15
N=3777
U nstandardized reg ressio n co efficients reported; stan d ard errors in p a ren th eses.
*** p < 0 .0 0 1 ; ** p < 0 .0 1 ; * p < 0 .0 5 ; + p < 0.1
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
85
the criticality and breadth of a manager's knowledge base in firm's task
environment will be positively associated with the likelihood of promotion was
supported (p<0.001). Managers with broad knowledge base critical to the firm's
task environment had a better chance to advance to higher positions.
In Model 6, the effect of relative pay raise as a proxy for managerial
productivity was tested. As predicted in Hypothesis 2-e, relative pay raise was
positively associated with promotion. Due to the unavailability of compensation
data for some managers, the total number of N's was reduced to 3777. With the
reduced sample, relative pay increase was positively associated with the likelihood
of promotion (p<0.001). Managers who were more productive as reflected in
compensation were able to advance to higher positions than those who were less
productive.
Model 7 lists the results of Hypothesis 2-f and Hypothesis 2-g testing the
effects of two social capital variables. Networking ability outside the firm and
network centrality inside the firm were significant positively associated with the
likelihood of promotion (p<0.001 for both hypotheses). In Model 8, all the human
capital and social capital variables were tested simultaneously. All the relationships
were consistent with the results of the separate testing for the human and social
capital variables in Model 5-Model 7.
Hypotheses regarding tournament factors were tested in Model 9 through
Model 12 as in Table 4-c. In Model 9, as predicted in Hypothesis 3-a, overall career
velocity was positively associated with the likelihood of promotion (p<0.001). The
faster a manager moved along their career line in the past, the more likely he or she
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 4-c. Logistic Regressions of Executive Promotion (N=8978)
Variables Model 9 Model 10 Model 11 Model 12
Intercept -4.05*** -3.79*** -15.23 -15.90
(0.32) (0.31) (367.7) (369.4)
Control V ariables
C hem ical Industry (Dummy) -0.12 -0.16 -0.27 -0.32
(0.09) (0.10) (0.21) (0.22)
Natural G as Industry (Dummy) -0.10 -0.26* -0.59* -0.60*
(0.12) (0.12) (0.24) (0.25)
Firm Size (Log Sales) 0.07* 0.09** 0.06 0.05
(0.03) (0.03) (0.07) (0.07)
Firm P erform ance (ROA) 0.02** 0.02** 0.02+ 0.03*
(0.008) (0.007) (0.01) (0.01)
Firm G row th (5 Y ear Average) 0.002* 0.002* 0.004+ 0.005*
(0.001) (0.001) (0.002) (0.003)
N ew CEO 1.17*** 1.23*** 2.02*** 2.07***
(0.11) (0.11) (0.20) (0.21)
In d ependent V ariables
C ensoring D um m y 0.24* 0.13 -0.01 0.06
(0.10) (0.09) (0.19) (0.20)
(Tournament Factors)
Overall C areer V elocity 0.91*** 2.89***
(0.20) (0.84)
N um ber of S u rp a sse s by O ther M anagers -0.43*** -0.78***
at P resen t Level (0.06) (0.21)
N um ber of S u rp a sse s by O ther M anagers -0.31** -0.22*
a t Previous Level (0.11) (0.11)
C hi-square 297.65*** 327.33*** 172.82*** 196.14***
-2 Log Likelihood 4404.30 4367.83 953.60 918.93
N = 1 9 3 0
o
C O
o
It
z
U nstandardized regression coefficients reported; stan d ard errors in p a ren th eses.
*** p < 0 .0 0 1 ; ** p < 0 .0 1 ; * p < 0 .0 5 ; + p < 0.1
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
is able to move up die hierarchy. In Model 10, the total number of surpasses by
other managers in the same rank or lower was negatively associated with promotion
as predicted in Hypothesis 3-b (p<0.001). The more often a manager passes over the
promotion chance, the more difficult it becomes for the manager to advance to
higher positions. This result responds to the very question posed by tournament
theory researchers (e.g., Rosenbaum, 1979,1984), "Does tournament competition
take place at all levels of managerial hierarchies?" Therefore, random process of
career systems suggested by garbage can model is not supported. One can see that
managers compete against one another at each tournament round. In Model 11,
Hypothesis 3-c that the number of surpasses by other managers at the previous level
will be negatively associated with the promotion likelihood at the present level was
supported (p<0.01). Managers who did not compete successfully at the previous
level also seem to bear the disadvantage at the current level. The number of N's in
this model was reduced to 1930 due to lack of data for the managers' tournament
competition records at the previous level. In Model 12, all the hypotheses for
tournament factors were tested simultaneously in one equation. The purpose of
simultaneous testing is to determine the relative importance of each tournament
variable. Again, all the hypotheses were supported with the reduced sample of
1930. The most salient effect was found with the record of the latest winning or
losing the tournament at tire current level. Winning or losing at the previous level
had the smallest effect of all the tournament variables (p<0.05).
In Model 13 in Table 4-d, all the relational demographic variables were
tested. As predicted in Hypothesis 4-a, age similarity was positively associated with
promotion (p<0.05). Although the significance level was not high, CEOs promoted
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 4-d. Logistic Regressions of Executive Promotion (N=8978)
Variables Model 13 Model 1 4
Intercept -3.07*** -13.43***
Control Variables
(0.35) (2.33)
C hem ical Industry (Dummy) -0.16 -0.21 +
(0.10) (0.11)
N atural G as Industry (Dummy) -0.14 -0.16
(0.12) (0.13)
Firm Size (Log Sales) 0.07* 0.03
(0.03) (0.04)
Firm P erfo rm an ce (FtOA) 0.02** 0.03***
(0.008) (0.008)
Firm G row th (5 Y ear A verage) 0.003** 0.001
(0.001) (0.001)
N ew CEO 0.99*** 0.93***
(0.11) (0.13)
Independent Variables
C ensoring Dum m y 0.15 0.28*
(0.10) (0.11)
1 Internal Labor Market Factors)
Exit R ate a t Im m ediate Higher Level 0.85*
(0.35)
Prom otion R ate at Im m ediate Higher Level 0.03
(0.36)
Entry R ate a t Im m ediate Higher Level 0.22
(0.34)
Exit R ate a t Im m ediate Higher Level 0.56**
on J o b Ladder
(0.22)
Prom otion R ate a t Im m ediate Higher Level 0.44*
on J o b Ladder (0.22)
Entry R ate a t Im m ediate Higher Level -0.46
on Jo b Ladder
(0.57)
C ro ss-Jo b Ladder M ovem ent 0.40*
(0.18)
L ocation in M anagerial Hierarchy 0.02***
(% H igher Positions)
(0.002)
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 4-d. Continued
V a r ia b le s M o d el 1 3 M o d el 1 4
(Human and Social Capital Factors)
M a s te r's D egree (Dummy) 0.01
(0.10)
Ph. D. D egree (Dum m y) -0.05
(0.19)
A ge 0.43***
(0.09)
A ge S q a u re d /1 0 -0.05***
(0.008)
O rganizational T enure -0.02
(0.02)
O rganizational T en u re S q u ared /1 0 0.08+
(0.04)
M anagerial T ask K now ledge 0.18***
(0.04)
R elative Pay In crease
N etw orking Ability (O utside) 0.32***
(0.07)
N etw ork C entrality (Inside) 0.53***
(0.11)
(Tournament Factors)
O verall C areer V elocity 0.16
(0.27)
N um ber of S u rp a sse s by O ther M anagers
a t P re se n t Level
-0.38***
(0.07)
N um ber of S u rp a ss e s by O ther M anagers
a t Previous Level
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 4-d. Continued
Variables Model 13 Model 14
(Relational Demographic Factors)
Age Sim ilarity to CEO 0.01* -0.001
(0.006) (0.008)
Functional B ackground Sim ilarity to CEO 2.13*** 1.43***
(0.29) (0.33)
Educational Affiliation Sim ilairty to CEO 0.05 0.06
(0.06) (0.06)
O utside S ta tu s Similarity to CEO 2.28** 1.17*
(0.46) (0.46)
C hi-square 325.58*** 598.52***
-2 Log Likelihood 4304.62 4027.52
U nstandardized reg ressio n co efficients reported; stan d ard erro rs in p aren th eses.
*** p < 0 .0 0 1 ; ** p < 0 .0 1 ; * p < 0 .0 5 ; + p < 0.1
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
91
managers similar in age more favorably than those who are not. Also, Hypothesis 4-
b, testing the effect of functional background similarity on promotion, was
supported (p<0.001). Managers with similar functional background to CEO had an
advantage in advancing to higher positions. Contrary to Hypothesis 4-c,
educational affiliation similarity was not significant in explaining the likelihood of
promotion. As predicted in Hypothesis 4-d, outside status similarity was positively
associated with promotion (p<0.01). CEOs newly hired from outside seemed to
favor newly hired managers in deciding promotion.
In Model 14, the effects of all study variables, internal labor market, human
capital and social capital, tournament, and relational demographic factors, were
simultaneously tested. Relative pay increase and the total number of surpasses at
the previous level were excluded from the model due to the problems associated
with the reduced sample size and possible biases from it.1 Overall, the result of the
all inclusive testing is consistent with those of separate testings for each set of
variables in Model 1-Model 13, with a few exceptions. Notable deviations were
organizational tenure, career velocity, and age similarity. They became insignificant
when tested with all the other independent variables. Insignificance of
organizational tenure is particularly interesting. When tested only with other
human capital and social capital variables, the likelihood of promotion decreased
with tenure to a certain point, then increased thereafter. This finding was contrary
to the hypothesized direction. Managers seemed to have advantage of being new to
the firm, and this advantage diminished with longevity of service. In Model 14,
1 A lthough n o t reported in the T ab le, separate an aly ses w ere co n d u cted to ex a m in e if th e
relatio n sh ip s held w ith the reduced sam ple. T h ere w ere no sig n ific an t ch an g es in th e results.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
92
organizational tenure became weak; organizational tenure became insignificant and
organizational tenure squared was weakly positively correlated with promotion
(p<0.1). Given that outside status similarity remained significant in this inclusive
model, the effects of organizational tenure seemed to be attenuated greatly by
outside status similarity. It can be argued that organizational tenure effect may be a
mere representation of outside status similarity. It may not be that newly hired
managers were favored despite their low firm-specific human capital value, but that
they were favored due to their similarity with the CEO in terms of being new to the
firm.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
93
CHAPTER 5: DISCUSSION A N D CONCLUSION
In this section, empirical results will be interpreted in light of several research
paradigms. The implications of career advancement of top managers explored in
this study will also be discussed. Then, the study limitations will be addressed and
conclusions will be drawn from the discussions of the findings. In doing so, a new
set of questions will be proposed for future research.
The first set of results provides evidence supporting the importance of the
structural effect of MILMs on promotions. Firms in the study sample seem to
operate a certain form of internal labor markets, characterized as follows. First,
firms predominantly use internal promotions to fill vacancies occurring at higher
levels in die managerial hierarchy. The ratio of filling vacancies from outside into
the higher ranked subgroup was almost a half of those into the lower ranked
subgroup. This lower entries into the higher positions should lead to higher
promotion rates for die insiders. Better shield from outside market competition at
higher positions is one of die internal labor market characteristics. Second, the
presence of job ladders was supported with empirical evidence that vacancies were
filled with managers at the lower level in die same job ladder. Third, promotions in
MILMs seem to occur in a remarkably orderly manner. As discussed in the results
section, promotions to higher levels skipping die intermediate steps were rare.
Among die executives in the study sample, only about one every 10 promotions fell
in this category. It seems diat firms maintain highly stable hierarchical systems to
train and develop managerial skills through systematic and orderly internal
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
promotion processes. By moving managers through incremental promotions in an
orderly manner, firms can also reduce the possible larger risk of mismatches
between a person and a job caused by allowing managers to bypass job ladder steps
(DiPrete, 1987). Fourth, firms seem to operate a pyramidal structure in their
managerial hierarchies. Although not as strongly as it was hypothesized, this
pyramidal shape affected the promotion chances of managers. The lower in the
hierarchy a manager was located, the higher the manager's likelihood of promotion.
This result indicates that firms run a tournament among managers competing with
each other in order to move up the hierarchy and achieve associated higher pay and
status (Lazear & Rosen, 1981). The efficiency of such tournaments will be discussed
later in this section.
It should be noted, however, that there were certain characteristics of the
MILMs deviating from what have previously been suggested to be key aspects of
internal labor markets. First, this study finds that ports of entry were not found only
in the lower echelons of tire firms studied. On the contrary, sample firms revealed a
significant degree of openness to the external labor market. As reported in Table 1,
firms hired managers from outside into all levels of managerial positions. Tire odds
of hiring from outside against internal promotion were about one-to-two, meaning
that for every three vacancies, firms filled out one vacancy by hiring outside
personnel. This proportion is substantially higher than those found in past research
(e.g., Wholey, 1985; Rosenfeld, 1992). The reasons for the openness of internal labor
markets for managers seem obvious. First, firms can secure managerial resources in
new managerial or technological areas by hiring outside personnel (Milgrom &
Roberts, 1992). Second, firms can expand learning capabilities among present
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
95
managers in the firms by bringing in managers with new skills and experiences
(Keck & Tushman, 1993; Virany, Tushman, & Romanelli, 1986). The relatively high
average growth rate of 11% among sample firms during the observation period and
its positive correlation with the average entry rate of outside hires into the firms
(p<0.01) confirms this reasoning. Frequent technological and environmental
changes also contribute to the openness of MILMs.
A high degree of openness, though, poses certain threats to the career
systems that firms operate. First, it may disrupt the notion of developing
idiosyncratic managerial capabilities, or firm-specific managerial skills in the firm
(Becker, 1975). If internal labor markets are not shielded from market competition,
managers would have less motivation and incentive to align their careers with the
current firm, leading to more frequent strategic behavior and to higher turnover
rates which will both incur significant costs to the firm (Williamson, 1985). Second,
given the imperfect information in external labor markets, firms may wrongly
choose personnel who are less productive or do not match their new jobs.
Responding to this puzzle, firms in the sample in this study seem to make a subtle
but logical compromise between going outside and hiring from inside. First, filling
vacancies through internal promotion is twice as frequent as outside hiring. Second,
as shown in Table 1, entry rates in upper levels are half of those in lower levels. Top
managers in upper structure in the managerial hierarchy are better shielded from
outside competition than those in lower structure. Therefore, firms can still
motivate managers to pursue their career interests in the firm. Third, as tested in
Model 1 to Model 4, entry rates at the immediate higher rank did not reduce the
promotion opportunity for managers at die immediate lower rank. It seems likely
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
96
that firms hired outside personnel for positions that were newly created. Vacancies
caused by exits or promotions were more likely to be filled by insider promotions
than outside hires (correlation between exit rate and entry rate was not significant
with p=0.05).
A second aspect of MILMs that may deviate from traditional notions of
internal labor markets is the firms' use of cross-job ladder movements among their
top managers. While between-job ladder movements are rare compared to within-
job ladder movements in the sample (only 37 such movements were identified in the
sample), firms certainly favored managers in determining promotion who had
cross-ladder experiences. This practice stems from die particularity of managerial
tasks. Managerial jobs as indicated by several researchers should encompass a
broad range of experience and skill base (Martin & Strauss, 1959; Kanter, 1984). By
moving managers to different functions or different job ladders, firms provide the
managers with required preparation to qualify for higher positions. Firms may also
use cross-functional and cross-ladder movement of their managers as an adaptation
mechanism to environmental changes (Kanter, 1984).
A second set of variables, human capital and social capital factors, were
associated with promotion in die expected direction. General and firm-specific
human capital variables, age, managerial task knowledge, and managerial
productivity reflected in relative pay increase, were correlated with promotion as
predicted in the hypotheses. The only variable that was not supported was
organizational tenure in Model 5-Model 8. With all the other variables, though,
organizational tenure in Model 14 seemed to support firm-specific human capital
prediction. Two key social capital variables, networking ability and network
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
97
centrality, were positively associated with promotion. Firms seem to operate
sophisticated career systems that sort out managers in terms of human capital and
social capital value.
A third set of factors tested whether firms run tournament mobility systems
for their managers. All three aspects of tournament, overall career velocity, losing
(or winning) the present tournament round, and losing (or winning) the latest
previous round, were associated widi promotion as predicted in separate
hypotheses. These findings strongly suggest that firms maximally utilize all aspects
of career information of their employees. Past winning or losing career history is
particularly relevant in predicting future career success. Given information
impacteness, bounded rationality, and opportunism in organizations, the
tournament system has clear advantages. First, the record of each employee's
winning or losing career history is valuable organizational capital under conditions
of non-transferability of career information across organizational boundaries
(Prescott & Visscher, 1980). By favoring early winners, firms can reduce
opportunistic behaviors of managers at early stage of their careers since that is when
the return for reputational capital has its highest value (Gibbons & Murphy, 1992).
Second, firms can also reduce opportunism and strategic behavior of the managers
at later stages of career, by running continuously tournament beyond the earlier
career stage. Managers should compete against each other at each level of
managerial hierarchies. Arid the competition, as discussed earlier, includes not only
a manager's own cohort group, but also other managerial cohort groups.
Tournaments in managerial hierarchies can best be described as different sub
tournaments in which different cohort groups compete with one another, as sub
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
98
stages of a grand tournament. This important aspect of tournament was validated
by the empirical findings in this study. Tournament at every level, with increased
costs associated with monitoring and performance evaluation, may also reduce the
strategic behavior of employees at the later stage of their careers. This aspect of
tournament operation can be particularly useful if firms link disproportionate pay
increase to higher levels as suggested by several researchers (Capelli & Cascio, 1991;
Baker et al., 1994a; Lazear & Rosen, 1981). Managers will try their best to move up
the hierarchy to achieve higher status and pay.
Finally, a fourth set of variables, relational demographic factors, were also
associated with promotion. Except in the case of educational affiliation similarity,
all the other dyadic relational demographic characteristics between the CEOs and
the managers affected the likelihood of promotion. The evidence of relational
demographic effects on promotion provides an important basis for coalition
formation. The CEOs at the apex of managerial hierarchies seek to expand their
coalition of support by promoting people with similar demographic group
membership (Westphal & Zajac, 1995). As a result, demographic similarity both
reflects and reinforces the existing power distribution (Pfeffer, 1981).
Institutionalization of present power is perpetuated through selecting
demographically similar managers to higher positions.
Some important issues remain to be examined and resolved in this study.
First, in this study, certain sets of other organizational factors were not examined.
Past research on executive succession or career theory suggests that organizational
and environmental factors such as technology, strategy, and industry characteristics
affect the career patterns of managers (e.g., Smith & White, 1987; Vardi & Hammer,
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
1977; Forbes & Piercy, 1991; Hitt & Ireland, 1985). These variables are suggested to
influence certain structural characteristics of organizations, and hence, career
mobility of managers. Technological or industry characteristics have been related to
the favoring managers with certain functional background and demographic
characteristics (Rajagopalan & Datta, 1995; Vardi & Hammer, 1977). Future research
should take these factors into account as main effects and also as interaction effects
with all the characteristics examined in this study. One possible set of hypotheses
would be regarding the interaction effects between managerial discretion and
demography similarities. I would propose that promoting managers on the basis of
demographic similarity will be more prevalent in a low discretion environment
where non-efficiency based promotion decision is more possible due to the lack of
urgent economic goals.
Second issue relates to generalizability the finding. This study examined
career advancement of top managers in multiple industry settings. However,
generalizability of the finding is limited due to the limited number of industries.
Future research should test external validity of the findings of this study.
Third issue relates to other individual factors which are not examined in this
study. Literature on social stratification and distribution of inequality consider s
gender and race as one of the most important variables. In this study, they were
excluded not because their importance was ignored, but because there were no
possible ways of obtaining these variables from secondary data sources. Future
research will contribute to the literature of managerial promotion by looking at
whether or not, or to what extent managers with different gender and ethnic
backgrounds are constrained from moving up the hierarchy. Therefore, the
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
100
examination of interaction effects between various factors in this study and gender
or race will produce a meaningful contribution to the literature.
Fourth issue relates to the political perspective on careers. As suggested by
Pfeffer (1981,1989), the political theory of careers can be an alternative perspective
on understanding careers and career processes (Pfeffer, 1989, p.380). Numerous
evidence has been documented how managers obtain and exercise power to
influence organizational processes (Pfeffer & Salancik, 1974; Hickson et al., 1971).
Even in internal labor markets, it has been demonstrated that certain groups of
people have unequal chances of moving up in the hierarchy from the power
perspective (Pfeffer & Cohen, 1984; Baron et al., 1984; Baron & Bielby, 1986). One
should be cautious, though, in interpreting career processes merely from political
perspective. Power should be differentiated from other organizational and
individual factors that affect career processes in organizations. In this study, certain
individual and social factors can be interpreted as power variables. For an example,
networking ability and network centrality have been presented as sources of power
in past research (Finkelstein, 1992; Burt, 1992; Brass & Burkhardt, 1992). It needs to
be understood, however, that these network properties may not be the results of
purely political behaviors of individuals. They may be embedded in the social and
economic relational structure that an individual cannot fully manipulate.
Managerial task knowledge is another example. Although it is possible that one can
gain and exercise one's expert knowledge, it also may be the end-product of human
capital investment in that particular knowledge base. Therefore, the power
construct should be differentiated from other theoretical constructs in order to
examine its individual effect on promotion. I believe this will prove extremely
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
difficult. Empirical findings of this study suggest that firms operate efficient and
systematic career systems. In order for the power perspective to gain explanatory
"power," it needs to go beyond and above what is explained by "efficiency" theories
as suggested in this study.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
102
Appendix 1. List of the Sample Firms
Chemical Industry Computer Industry
Air Products & Chemicals
Allied Chemical Inc
American Cyanatnid Co
Big Three Industries
Cabot Corp
Celanese Corp
Chemed Corp
Clorox Co
Crompton-Knowles Corp
Dexter Corp
Essex Chemical Co
Ethyl Corp
Ferro Corp
H. B. Fuller Co
Great Lakes Chemical
Hercules Inc
Int’l Flavor & Fragrance
International Mineral Co.
Koppers Co
Loctite Corp
Lubrizol Corp
Monsanto Co
Nalco Chemical Co
N at’l Distiller & Chem
National Service
Olin Corp
PPG Industries Inc
Pennwalt
Petrol ite Crop
Rohm & Haas Co
Sherwin & W illiams Co
Stauffer Chem Co
Sun Chemical (Sequa)
Union Carbide Co
Witco Corp
Advanced Micro Device
Amdahl Corp
Apple Computer Inc
Burroughs Co
Computervision
Control Data Corp
Cray Research
Data General
Datapoint Corp
Dataproducts Corp
Diebold Inc
Digital Equipment
General Automation
Gerber Scientific
Gould Inc
Harris Corp
Hewlett-Packard Co
Honeywell Inc
Int’l Business Machine
Mohawk Data (Qantel)
NCR Corp
Nashua
Paradyne Corp
Pitney-Bowes Inc
Recognition Equipment
SCI
Savin Corp
Storage Technology
Tandem
Telex
Wang Laboratories
Xerox Corp
Natural Gas Industry
American Natural Resources
ARKLA Inc
Atlanta Gas Light Co
Bay State Gas
Brooklyn Union Gas
Coastal State Gas
Columbia Gas Systems
Connecticut Natural Gas
El Paso Company
Enserch Corp
Gas service
Houston Natural Gas Corp
Natural Fuel Gas Co
N icor Inc
Northwest Natural Gas Co
ONEOK Inc
Pacific Enterprise
Pacific Gas & Electric
Piedmont Natural Gas Co
Pioneer Corp
Southern Union Co
Texas Eastern Corp
Transco Energy Co
United Cities Gas Co
Washington Gas Light Co
W ICO R Inc
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Appendix 2. Elite Educational Institutions
Amherst College
Brown University
Carleton College
Columbia University
Cornell University
Dartmouth College
Grinnell College
Harvard University
Haverford University
Johns Hopkins University
MIT
New York University
Northwestern University
Oberlin College
Pomona College
Princeton University
Stanford University
Warthmore College
United States Military Academy
United States Naval Academy
University of California, Berkeley
University of California, Los Angeles
University of Chicago
University of Michigan
University of Pennsylvania
Wellesley College
Wesleyan University
Williams College
Yale University
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
104
REFERENCES
Abraham, K. G., & Medoff, J. L. 1985. Length of service and promotions in union and
nonunion work groups. Industrial and Labor Relations Review, 38:408-420.
Allen, M. P. 1974. The structure of intraorganizational elite cooptation: Interlocking
corporate directorates. American Sociological Review, 39: 393-406.
Allison, P. D. 1984. Event history analysis: Regression fo r longitudinal event data.
Newbury Park, California: Sage Publications.
Althauser, R. P. 1989. Internal labor markets. Annual Review o f Sociology, 15:143-161.
Althauser, R. P., & Kalleberg, A. L. 1981. Firms, occupations and the structure of labor
markets: A conceptual analysis. In I. Berg (Ed.), Sociological perspectives on labor
m arkets: 119-149, New York: Academic Press.
Anderson, J. C., Milkovich, G. T., & Tsui, A. 1981. A model of intra-organizational
mobility. Academy o f Management Review, 6:529-538.
Andrews, K. R. 1971. The concept o f corporate strategy. Homewood, Illinois: Irwin.
Aoki, M. 1988. Information, Incentives, and Bargaining in the Japanese Economy.
Cambridge, MA: Cambridge University Press.
Argyris, C. 1957. Personality and organization. New York: Harper & Row.
Baker, G., & Gibbs, M., & Holmstrom, B. 1993. Hierarchies and compensation.
European Economic Review, 37: 366-378.
Baker, G., & Gibbs, M., & Holmstrom, B. 1994a. The internal economics of the firm:
Evidence from personnel data. Quarterly Journal o f Economics, 109: 881-919.
Bantel, K. A., Jackson, S. E. 1989. Top management and innovations in banking: Does
the composition of the top team make a difference? Strategic Management
Journal, 10:107-124.
Barnard, C. 1.1938. The functions o f the executive. Cambridge, MA: Harvard
University Press.
Baron, J. N. 1984. Organizational perspectives on stratification. Annual Review o f
Sociology, 10: 37-69.
Baron, T . N., & Bielby, W. T. 1986. The proliferation of job title in organizations.
Adm inistrative Science Quarterly, 31: 561-585.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
105
Baron, J. N., Davis-Blake, A., & Bielby, W. T. 1986. The structure of opportunity: How
promotion ladders vary within and among organizations. Administrative Science
Quarterly, 31: 248-273.
Baum, J. A. C., & Oliver, C. 1991. Institutional linkages and organization mortality.
Adm inistrative Science Quarterly, 36:187-218.
Beatty, R. P., & Zajac, E. J. 1987. CEO change and firm performance in large
corporations: Succession effects and manager effects. Strategic Management
Journal, 8:305-317.
Becker, G. S. 1975. Human Capital: A theoretical and empirical analysis, w ith special
reference to education. Chicago: University of Chicago Press.
Bird, A. W. 1988. Career advancement in Japanese top management teams: Who gets
to the top? Doctoral dissertation, University of Oregon.
Blaug, M. 1976. The empirical status of human capital theory: A slightly jaundiced
survey. Journal o f Economic Literature, 14: 827-855.
Boeker, W. 1992. Power and managerial dismissal: Scapegoating at the top.
Administrative Science Quarterly, 37:400-421.
Bonacich, P. 1972. Factoring and weighting approaches to status scores and clique
identification. Journal o f Mathematical Sociology, 2:113-120.
Bourdieu, P. 1977. Cultural reproduction and social reproduction. In J. Karabel and A.
H. Halsey (Eds.), Power and ideology in education.: 487-511. New York: Oxford
University Press.
Boxman, E. A. W., De Graff, P. M., & Flapp, H. D. 1991. The impact of social and
human capital on the income attainment of Dutch managers. Social Network, 13:
51-73.
Brass, D. J., & Burkhardt, M. E. 1992. Centrality and power in organizations. In N.
Nohria & R. G. Eccles (Eds.), Networks and organizations: 191-215. Boston, MA:
Harvard University Press.
Burt, R. S. 1992. The social structure of competition. In N. Nohria & R. G. Eccles (Eds.),
Networks and organizations: 57-91. Boston, MA: Harvard University Press.
Byrne, D. 1971. The attraction paradigm. New York: Academic Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
106
Byrne, D., Clore, G. L., Jr., & Smearton. G. 1986. The attraction hypothesis: Do similar
attitudes affect anything? Journal o f Psychology and Social Psychology, 4: 220-
224.
Cannings, K. 1988. Managerial promotion: The effects of socialization, specialization,
and gender. Industrial and Labor Rela tions Review, 42: 77-88.
Caplow, T. 1954. The sociology o f work. New York: McGraw-Hill.
Cappelli, P., & Cascio, V V . F. 1991. Why some jobs command wage premiums: A test of
career tournament and internal labor market hypotheses. Academy of
Management Journal, 34: 848-868.
Carmichael, L. 1983. Firm-specific human capital and promotion ladders. Bell
Journal o f Economics, 251-258.
Child, J. 1972. Organization structure, environment, and performance: The role of
strategic choice. Sociology, 6: 2-22.
Chinoy, E. 1955. Automobile workers and the American dream. New York: Random
House.
Clogg, C. C., & Eliason, S. R. 1987. Some common problems in log-linear analysis.
Sociological Methods and Research, 15: 8-44.
Cohen, M. D., March, J. G., & Olsen, J. P. 1972. A garbage can model of organizational
choice. Administrative Science Quarterly, 17:1-25.
Coleman, J. S. 1988. Social capital in the creation of human capital. American Journal of
Sociology, 94: S95-S120.
Cotton, J. L., & Tuttle, J. M. 1986. Employee turnover: A meta-analysis and review with
implication for research. Academy of Management Review, 11: 55-70.
Crozier, M. 1964. The bureaucratic phenomenon. Chicago: University of Chicago Press.
Cyert, R. M., & March, J. G. 1963. A behavioral theory o f the firm. Englewood Cliffs,
New Jersey: Prentice-Hall.
Dalton, D. R., & Kesner, I. F. 1985. Organizational performance as an antecedent of
inside/outside chief executive succession: An empirical assessment. Academy of
Management Journal, 28: 749-762.
Dalton, M. 1951. Informal factors in career achievement. American Journal o f
Sociology, 56:407-415.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
107
Davis-Blake, A. 1986. Career Contingencies: Organizational effects on individual
earnings and promotions. Doctoral dissertation, Stanford University.
De Graff, N. D., & Flap, H. D. 1988. With a little help from my friends: Social resources
as an explanation of occupational status and income in West Germany, the
Netherlands, and the United States. Social Forces, 67:452-472.
Dearborn, D., & Simon, H. A. 1958. Selective perception: A note on the departmental
identification of executives. Sociometnj, 35:38-48.
DiPrete, T. A. 1987. Horizontal and vertical mobility in organizations. Administrative
Science Quarterly, 32:422-444.
DiPrete, T. A., & Soule, W. T. 1986. The organization of career lines: Equal
employment opportunity and status advancement in a federal bureaucracy.
American Sociological Review, 51: 295-309.
Doeringer, P. B., & Piore, M. J. 1971. Internal labor markets and manpower analysis.
Lexington, MA: Lexington Books.
Donaldson, G., & Lorsch, J. W. 1983. Decision making a t the top: The shaping o f
strategic direction. New York: Basic Books.
Dreyfuss, C. 1938. Occupation and ideology o f salaried employee. New York:
Columbia University Press.
Farjoun, M. 1994. Beyond industry boundaries: Human expertise, diversification, and
resource-related industry groups. Organization Science, 5:185-199.
Felmlee, D. H. 1984. A dynamic analysis of women's employment exits.
Demography, 21:171-183.
Finkelstein, S. 1992. Power in top management teams: Dimensions, measurement, and
validation. Academy of Management Journal, 35:505-538.
Finkelstein, S., & Hambrick, D. C. 1990. Top-managemenl-Leam tenure and
organizational outcomes: The moderating role of managerial discretion.
Adm inistrative Science Quarterly, 35:484-503.
Forbes, J. B. 1987. Early intraorganizational mobility: Patterns and influences.
Academy o f Management Journal, 30:110-125.
Forbes, ]. B., & Piercy,}. E. 1991. Corporate m obility and paths to the top: Studies
fo r human resource and management development specialists. New York:
Quorum Books.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
108
Galbraith, J. K. 1971. The new industrial state. (2nd ed.) New York: Mentor.
Gibbons, R., & Murphy, K. 1992. Optimal incentive contracts in the presence of career
concerns. Journal o f Political Economy, 100: 468-505.
Gibbs, M. 1994. Incentive compensation in a corporate hierarchy. Working Paper,
University of Chicago.
Glenn, P. E. 1969. Aging, disengagement, and opinionation. Public Opinion Quarterly,
33:17-33.
Granovetter, M. S. 1974. Getting a job: A study o f contacts and careers. Harvard
University Press.
Grusky, 0 . 1961. Corporation size, bureaucratization, and managerial succession.
American Journal o f Sociology, 67: 263-269.
Grusky, 0 . 1966. Career mobility and organizational commitment. Adm inistrative
Science Quarterly, 10: 489-502.
Guthrie, J. P., & Olian, J. D. 1991. Does context affect staffing decisions? Case of
general managers. Personnel Psychology, 44:263-292.,
Guttentag, M., & Secord, P. 1983. Too many women: The sex ratio question. Beverly
Hills: Sage Publications.
Hall, D. T. 1976. Careers in organizations. Pacific Palisades, CA: Goodyear.
Hambrick, D. C., & Finkelstein, S. 1987. Managerial discretion: A bridge between
polar views on organizations, hi L. L. Cummings & B. M. Staw (Eds.), Research in
Organizational Behavior, Vol. 9: 69-406. Greenwich, CT: JAI Press.
Hambrick, D. C. 1981. Environment, strategy, and power within top management
teams. Administrative Science Quarterly, 26: 253-276.
Hambrick, D. C., & Mason, P. A. 1984. Upper echelons: The organization as a
reflection of its top managers. Academy o f Management Review, 9:193-206.
Harrison, J. R., Torres, D. L., & Kukalis, S. 1988. The changing of the guard: Turnover
and structural change in the top-management positions. Administrative Science
Quarterly, 33: 211-232.
Hartog, J. 1986. Earnings functions: Beyond human capital. Applied Economics, 18:
1291-1309.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
109
Hashimoto, M. 1979. Bonus payments, on-the-job training, and lifetime employment
in Japan. Journal o f Political Economy, 87:1086-1104.
Hashimoto, M. 1981. Firm-specific human capital as a shared investment. American
Economic Review, 71:475-482.
Helmich, D. L., & Brown, W. B. 1972. Successor type and organizational change in the
corporate enterprise. Administrative Science Quarterly, 17:371-381.
Hickson, D. J. 1987. Decision-making at the top of organizations. Annual Review o f
Sociology, 13,165-192.
Hickson, D. J., Lee, C. A., Schneck, R. E., & Pennings, J. M. 1971. A strategic
contingencies theory of intraorganizational power. Adm inistrative Science
Quarterly, 16:216-229.
Hitt, M. A., & Ireland, R. D. 1985. Corporate distinctive competence, strategy, industry
and performance. Strategic Management Journal, 6: 689-713.
Holland, J. 1973. M aking vocational choices: A theory o f careers. Englewood Cliffs,
NJ: Prentice-Hall.
Hrebiniak, L. G., & Joyce, W. F. 1985. Organizational adaptation: Strategic choice and
environmental determinism. Administrative Science Quarterly, 30:336-349.
Ibarra, H. 1993. Network centrality, power, and innovation involvement
Determinants of technical and administrative roles. Academy ofManagetnent
Journal, 36:471-501.
Jackson, S. E., Brett, J. F., Sessa, V. I., Cooper, D. M., Julin, J. A., & Peyronnin, K. 1991.
Some differences make a difference: Individual dissimilarity and group
heterogeneity as correlates of recruitment, promotions, and turnover. Journal o f
Applied Psychology, 76: 675-689.
Jensen, M. C., & Murphy, K. J. 1990. Performance pay and top management incentives.
Journal o f Political Economy, 98: 225-264.
Jung, J. H., & Magrabi, F. M. 1991. Work experience, specific human capital, and
earnings. Quarterly Review of Economics and Business, 31:15-27.
Kalleberg, A., & Sorenson, A. B 1979. The sociology of labor markets. Annual Review
o f Sociology, 5: 351-379.
Kanter, R. M. 1977. Men and women o f the corporation. New York: Basic Books.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
110
Kanter, R. M. 1979. Differential access to opportunity and power. In R. Alvarez (Ed.),
Discrimination in organizations: 52-68. San Francisco: Jossey-Bass.
Kanter, R. M. 1984. Variations in managerial career structures in high-technology
firms: The impact of organizational characteristics on internal labor market
patterns. In P. Osterman (Ed.), Internal labor markets, Cambridge, MA: The MIT
Press.
Katz, D., & Kahn, R. L. 1978. The social psychology o f organizations. 2nd ed., New
York: Wiley.
Katz, L. 1953. A new status index derived from sociometric analysis. Psychometrika,
18: 39-43.
Keck, S. L., & Tushman, M. L. 1993. Environmental and organizational context and
executive team structure. Academy o f Management Journal, 36:1314-1344.
Kerr, J., & Kren, L. 1992. Effect of relative decision monitoring on chief executive
compensation. Academy o f Management Journal, 35:370-397.
Koike, K. 1988. Industrial relations in Japan. Tokyo: University of Tokyo Press.
Korn, H. J., Milliken, F. J., & Lant, T. K. 1992. Top management team change and
organizational performance: The influence of succession, composition, and
context. Paper presented at the National Academy of Management Annual
Conference, Las Vegas, Nevada.
Kotter, J. P. 1988. The leadership factor. New York: Free Press.
Kreps, D. M. 1990. Game theory and economic modeling. New York: Oxford
University Press.
Kurtz, D. L., Boone, L. E., & Fleenor, C. P. 1989. CEO: Who gets to the top in
America? East Lansing, Michigan: Michigan State University Press.
Lazear, E. P. 1991. Labor economics and the psychology of organizations. Journal o f
Economic Perspectives, 5: 89-110.
Lazear, E. P., & Rosen, S. 1981. Rank-order tournaments as optimum labor contracts.
Journal o f Political Economy, 89: 841-864.
Levenson, B. 1961. Bureaucratic succession, hi A. Etzioni (Ed.), Complex organization:
A sociological reader. 362-275. New York: Holt.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
I l l
Lieberson, S., & O'Connor, J. F. 1972. Leadership and organizational performance: A
study of large corporations. American Sociological Review, 37; 117-130.
Lincoln, J. R., Miller, J. 1979. Work and friendship ties in organizations: A comparative
analysis of relational networks. Administrative Science Quarterly, 24:181-189.
London, M., & Stumpf, S. A. 1981. Managing careers. Reading, MA: Addison-Wesley.
Mace, M. L. 1971. Directors: M yth and reality. Boston: Harvard University Business
School Press.
March, J. C., & Simon, H. A. 1958. Organizations. New York: Wiley.
Markham, W. T., Harlan, S. L., & Hackett, E. J. 1987. Promotion opportunity in
organizations: Cause and consequences. In K. M. Rowland & G. R. Ferris (Eds.),
Research in Personnel and Human Resource Management, Vol. 5:223-287.
Martin, N. H., & Strauss, A. L. 1959. Patterns of mobility within industrial
organizations. In W. L. Warner & N. H. Martin (Eds.) Industrial man:. 85-101.
New York: Harper & Row.
March, J., & Olsen, J. P. 1976. Ambiguity and choice in organizations. Bergen, Norway:
Universitetsforlaget
Maxim, P. S. 1985. Cohort size and juvenile delinquency: A test of the Easterlin
hypothesis. Social Forces, 63:, 661-681.
Mclntire, S., Moberg, D. J., & Posner, B. Z. 1980. Preselection decisions according to sex
and race. Academy o f Management Journal, 23: 738-749.
Medoff, J. L., & Abraham, K. G. 1981. Are those paid more really more productive?:
The case of experience. Journal o f Human Resources, 16:186-216.
Meyer, B. 1991. Learning from coarse information: Biased contests and career profiles.
Review o f Economic Studies, 58:15-41.
Milgrom, P., & Roberts, J. 1992. Economics, organization, and management.
Englewood Cliffs, New Jersey: Prendce-Hall.
Mills, D. Q. 1985. Seniority versus ability in promotion decisions. Industrial and Labor
Relations Review, 38:421-425.
Mincer, J. 1987. Schooling, experience, and earnings. New York: National Bureau of
Economic Research.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
112
Miner, A. S. 1987. Idiosyncratic jobs in formalized organizations. Adm inistrative
Science Quarterly, 32: 327-351.
Mintzberg, H. 1983. Power in and around organizations. Englewood Cliffs, New York:
Prentice-Hall.
Mizruchi, M. S., & Bunting, D. 1981. Influence in corporate networks: An examination
of four measures. Administrative Science Quarterly, 26:475-489.
Nelson, R. E. 1989. The strength of strong ties: Social networks and intergroup conflict
in organizations. Academy o f Management Journal, 32: 377-401.
Nystrom, P. C., & Starbuck, W. H. 1984. To avoid organizational crises-unleam.
Organizational Dynamics, 12:53-65.
O'Reilly III, C. A., Caldwell, D. F., & Barnett, W. P. 1989. Work group demography,
social integration, and turnover. Administrative Science Quarterly, 34:21-37.
Osterman, P. 1987. Choice of employment systems in internal labor markets.
Industrial Relations, 26: 46-67.
Oswald, A. J. 1984. Wage and employment structure in an economy with internal
labor markets. Quarterly Journal o f Economics, 99: 693-716.
Pfeffer, J. 1981. Power in organizations. Marshfiled, MA: Pitman.
Pfeffer, J. 1989. A political perspective on careers: Interests, networks, and
environments. In M. B. Arthur, D. T. Hall, & B. S. Lawrence (Eds.), Handbook o f
career theory: 380-396.
Pfeffer, J., & Cohen, Y. 1984. Determinants of internal labor markets in organizations.
Administrative Science Quarterly, 29: 550-572.
Pinfield, L. T., & Berner, M. F. 1994. Employment systems: Toward a coherent
conceptualization of internal labor markets. In G. R. Ferris (Ed.) Research in
personnel and human resource management, Vol. 12:41-79.
Porter, L. W., Steers, R. M. 1973. Organizational, work, and personal factors in
employee turnover and absenteeism. Psychological Bulletin, 50:151-176.
Porter, M. E. 1980. Competitive strategy. New York: Free Press.
Porter, M. E. 1990. The competitive advantage o f nations. New York: Free Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
113
Powell, G. N., & Butterfield, D. A. 1994. Investigating the "glass ceiling" phenomenon:
An empirical study of actual promotions to top management. Academy of
Management Journal, 37: 68-86.
Prescott, E. C., & Visscher, M. 1980. Organizational capital. Journal o f Political
Economy, 88: 446-461.
Priem, R. L. 1990. Top management team group factors, consensus, and firm
performance. Strategic Management Journal, 11:469-478.
Puffer, S. M., & Weintrop, J. B. 1991. Corporate performance and CEO turnover: The
role of performance expectation. Administrative Science Quarterly, 36:1-19.
Raghuram, S. 1992. Linking management staffing and training practices w ith business
strategy. Doctoral dissertation, University of Minnesota.
Rajagopalan, N., & Datta, D. K. 1995. CEO characteristics: Does industry matter?
Working Paper, University of Southern California.
Reinganum, M. 1985. The effect of executive succession on stockholder wealth.
Administrative Science Quarterly, 30: 46-60.
Rosenbaum, J. R. 1979. Tournament mobility: Career patterns in a corporation.
Adm inistrative Science Quarterly, 24: 220-241.
Rosenbaum, J. R. 1984. Corporate mobility in a corporate hierarchy. Orlando,
Florida: Academic Press.
Rosenbaum, J. R. 1989. Organization career systems and employee misperceptions.
In M. B. Arthur, D. T. Hall, & B. S. Lawrence (Eds.), Handbook o f career theory:
329-353. New York: Cambridge University Press.
Rosenfeld, R. A. 1992. Job mobility and career processes. Annual Review o f Sociology,
18: 39-61.
Schein, E. H. 1964. How to break in the college graduates. Harvard Business Review,
42: 68-76.
Sheridan, J. E., Slocum, Jr. J. W., Buda, R., & Thompson, R. C. 1990. Effects of corporate
sponsorship and departmental power on career tournaments. Academy o f
Management Journal, 33:578-602.
Sicherman, N., & Galor, 0 . 1990. A theory of career mobility. Journal o f Political
Economy, 98:169-192.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
114
Smith, M., & White, M. C. 1987. Strategy, CEO specialization, and succession.
Administrative Science Quarterly, 32: 263-280.
Sorensen, A. B. 1974. A model for occupational careers. American Journal o f
Sociology, 80:44-57.
Sorensen, A. B., & Kalleberg, A. L. 1981. An outline of a theory of the matching of
persons to jobs. In Berg (Ed.), Sociological perspectives on labor markets: 49-76.
New York: Academic Press.
Spence, M. A. 1973. Job market signaling. Quarterly Journal o f Economics, 83: 355-
374.
Spilerman, S. 1977. Careers, labor market structure, and socioeconomic achievement.
American Journal o f Sociology, 83:, 551-593.
Stewart, L. P., & Gudykunst, W. B. 1982. Differential factors influencing the
hierarchical level and number of promotions of males and females within an
organization. Academy o f Management Journal, 25: 586-597.
Stewman, S. 1988. Organizational demography. Annual Review o f Sociology, 14:173-
202.
Stewman, S., & Konda, S. L. 1983. Careers and organizational labor markets:
Demographic models of organizational behavior. American Journal o f
Sociology, 88: 637-685.
Strober, M. H. 1990. Human capital theory: Implication for HR managers. Industrial
Relations, 29:215-239.
Swinyard, A. W., & Bond, F. A. 1980. Who gets promoted? Harvard Business Review,
Sep-Oct: 6-18.
Tajfel, H., & Turner, J. C. 1986. The social identity theory of intergroup behavior. In S.
Worchel & W. G. Austin (Eds.), Psychology of intergroup relations: 7-24.
Tharenou, P., Latimer, S., & Conroy, D. 1994. How do you make it to the top? An
examination of influences on women's and men's managerial advancement.
Academy o f Management Journal, 37: 899-931.
Thompson, J. D. 1967. Organizations in action. New York: McGraw-Hill.
Tichy, N. M., Tushman, M. L., & Fombrun, C. 1979. Social network analysis for
organizations. Academy o f Management Review, 4: 507-519.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
115
Topel, R. 1991. Specific capital, mobility, and wages: Wage rise with job seniority.
Journal o f Political Economy, 99:145-176.
Tuma, N. B. 1976. Rewards, resources, and the rate of mobility: A nonstationary
multivariate stochastic model. American Sociological Review, 41:338-360.
Turner, R. 1964. Models of social ascent through education: Sponsored and contest
mobility. American Sociological Review, 25: 855-867.
Tushman, M. R., & Romanelli, E. 1983. Uncertainty, social location and difference in
decision making. Management Science, 28:12-23.
Tushman, M. R., & Romanelli, E. 1985. Organizational evolution: A metamorphosis
model of convergence and reorientation. In L. L. Cummings & B. Staw (Eds.),
Research in Organizational Behavior, Vol. 7:177-222. Greenwich, CT: JAI Press.
Useem, M., & Karabel, J. 1986. Pathways to top corporate management. American
Sociological Review, 51:184-200.
Vancil, R. F. 1987. Passing the baton: Managing the process o f CEO succession.
Boston: Harvard Business School Press.
Vardi, Y., & Hammer, T. H. 1977. Intraorganizational mobility and career
perceptions among rank and file employees in different technologies. Academy
o f Management Journal, 20:622-634.
Veiga, J. F. 1973. The mobile manager at mid-career. Harvard Business Review, Jan-
Feb: 115-119.
Veiga, J. F. 1981. Do managers on the move get anywhere? Harvard Business Review,
Jan-Feb: 205-38.
Veiga, J. F. 1985. Plateaued versus nonplateaued managers: Career patterns, attitudes,
and path potential. Academy o f Management Journal, 24: 566-578.
Virany, B., Tushman, M. L., & Romanelli, E. 1985. A longitudinal study of the
determinants and effects of executive succession. Academy of Management
Proceedings o f Annual Meeting, 186-190.
Wachter, M. L., & Wright, R. D. 1990. The economics of internal labor markets.
Industrial Relations, 29: 240-262.
Wagner, W. G., Pfeffer, J., & O'Reilly III, C. A. 1984. Organizational demography and
turnover in top-management groups. Administrative Science Quarterly, 29: 74-92.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
116
Wakabayashi, M., Graen, G., Graen, M., & Graen, M. 1988. Japanese management
progress: Mobility into middle management. Journal o f Applied Psychology, 73:
217-227.
Waldman, D. A., & Avolio, B. J. 1986. A meta-analysis of age differences in job
performance. Journal o f Applied Psychology, 71: 33-38.
Walsh, J. P. 1988. Selectivity and selective perception: An investigation of managers'
belief structure and information processing. Academy o f Management Journal,
873-896.
Wanous, J. P. 1975. Realistic job previews for organizational recruitment. Personnel, 52:
50-60.
Wells, D. L., & Muchinsky, P. M. 1985. Performance antecedents of voluntary and
involuntary managerial turnover. Journal o f Applied Psychology, 70: 329-336.
Westphal, J. D., & Zajac, E. D. 1995. W7io shall govern? CEO/board power,
demographic similarity, and new director selection. Adm inistrative Science
Quarterly, 40: 60-83.
White, H. C. 1970. Chains o f opportunity: System models o f m obility in organizations.
Cambridge, MA: Harvard University Press.
Whitely, W., Dougherty, T. W., & Dreher, G. F. 1991. Relationship of career mentoring
and socioeconomic origin to managers' and professionals' early career progress.
Academy o f Management Journal, 34: 331-351.
Wholey, D. R. 1985. Determinants of firm internal labor markets in large law firms.
Adm inistrative Science Quarterly, 30:318-335.
Wiersema, M. F., Bantel, K. A. 1992. Top management team demography and
corporate strategic change. Academy of Management Journal, 35: 91-121.
Wiersema, M. F., Bantel, K. A. 1993. Top management team turnover as an adaptation
mechanism: The role of the environment. Strategic Management Journal, 14: 485-
504.
Wilensky, H. L. 1956. Intellectuals in labor unions. New York: Free Press.
Williamson, O. E. 1975. Markets and hierarchies: Analysis and antitrust implications.
New York: Free Press.
Williamson, O. E. 1985. The economic institutions o f capitalism. New York: The Free
Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
117
Yamaguchi, K. 1991. Event history analysis. Newbury Park, California: Sage
Publications.
Zenger, T. R. 1992. Why do employers only reward extreme performance? Examining
die relationships among performance, pay, and turnover. Adm inistrative Science
Quarterly, 37:198-219.
Zenger, T. R., Lawrence, B. S. 1989. Organizational demography: The differential
effects of age and tenure distributions on technical communication. Academy of
Management Journal, 32:353-376.
Zweigenhaft, R. L. 1992. The application of cultural and social capital: A study of the
25th year reunion entries of prep school and public school graduates of Yale
College. Higher Education, 23:311-320.
Zweigenhaft, R. L., & Domhoff, G. W. 1991. Blacks in the white establishment? A
study o f race and class in America. New Haven: Yale University Press.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Astrocyte regulation of endothelial haemostasis function via transforming growth factor-beta

PDF

Aspirations For Career Advancement In Public Organizations: The Case Of South Korean Bureaucracy.

PDF

A biogeographic theory of industrial market structure and competitive dynamics

PDF

Causes, consequences, and moderating factors of career plateauing.

PDF

A conceptual framework of distribution intensity

PDF

Cross-sectional tests of tax and regulatory costs of a change in depreciation methods in the railroad industry

PDF

Defining the advice -seeking request in electronic solicitations in boundary spanning environments

PDF

Effects of risk perception on participation in the advertising agency review

PDF

An investigation of the impact of congruent managerial values on appraisal behavior

PDF

A system of participation and allocation preferences in organizations

PDF

Communication practices and organizational change: The structuring of process improvement experiences

PDF

A correlational study of feminist/womanist identity development and depression in women

PDF

An investigation into the regional segmentation of the commercial real estate market in the United States

PDF

A prospective cohort study of morbidity and mortality among middle-aged men in Shanghai, China

PDF

IT -enabled dynamic capabilities in new product development: Building a competitive advantage in turbulent environments

PDF

Correlates of deceit: A cross-cultural examination

PDF

AIDS preventive sexual behavior in college students: An empirical test of the health belief model

PDF

The development and testing of a seminar for increasing the cognitive complexity of individuals

PDF

A study on the content and determinants of close exchange relationships in channels of distribution

PDF

A test of survival-the case of South Vietnam

Asset Metadata

Creator Kim, Yong-Min (author)

Core Title Career advancement in managerial hierarchies in the United States firms: A multi-theoretic model and empirical tests of the determinants of managerial promotion

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Business Administration

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag business administration, management,OAI-PMH Harvest

Language English

Advisor [illegible] (committee chair), [illegible] (committee member), Fulk, Janet (committee member), Rajagapalan, N. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-476434

Unique identifier UC11352665

Identifier 9616977.pdf (filename),usctheses-c17-476434 (legacy record id)

Legacy Identifier 9616977-0.pdf

Dmrecord 476434

Document Type Dissertation

Rights Kim, Yong-Min

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA