><.<.T.
UBRARIES - DEVVPY
HD28
.M414
&
/ ^
Oewey
\
6j^
BOARDS AS NORMATIVE ARENAS:
$*
'
CORPORATE GOVERNANCE AND THE ROUTINES OF
CEO SELECTION
Worldiig Paper No. 3724-94
WILLIAM OCASIO
Massachusetts Institute of Technology
Sloan School of Management
E52-562
50 Memorial Drive
Cambridge, Massachusetts 02139
Tel. (617) 253-8412 E-Mail: wocasio@sloan.mit.edu.
September 1994
;.;assachusetts iNsriTure
OF TECHNOLOGY
DEC 01
1994
LIBRARIES
Boards as Normative Arenas:
Corporate Governance and the Routines of
CEO
Selection'
William Ocasio
Tel:
Massachusetts Institute of Technology
Sloan School of Management
E52-562
50 Memorial Drive
Cambridge, MA 02139
(617) 253-8412 e-mail: wocaslo@sloan.mit.edu
September 1994
I
would
like to
thank Mauro Guillen, Rebecca Henderson, and Nick Ziegler
suggestions, Hyosun Kim, Sandy Raju, and Hee-Jun
the Stanford Graduate School of Business and the
research support.
Rho
for
assistance
MIT Sloan School
of
in
for
comments and
data collection, and
Management
for
Boards as Normative Arenas:
Corporate Governance and the Routines of
CEO
Selection
Abstract
I
develop a conceptual model of corporate boards as normative arenas, self-regulated by rules
embodied
in
the organization's norms and routines.
routinization of
CEO
selection
in
I
apply this theory to explain the
U.S. industnal corporations.
An
analysis of the competing nsks
of insider versus outsider succession provides strong evidence for reliance
informal
CEO
selection routines.
selection routines,
structure,
I
and early
To
test for alternative explanations for
examine the mediating
CEO
on formal and
why boards
effects of performance, firm age,
rely
on
ownership
departures on their persistence. The findings support the view of
boards as normative arenas.
The governance
of U.S. corporations
is
undergoing renewed scrutiny by organization theorists,
strategy researchers, financial economists, and by the general business press. While both
organizational and economic theorists have traditionally viewed corporate boards as
mechanisms
for adapting the organization to
Salancik, 1978;
Fama and
environmental and market
demands
(Pfeffer
and
Jensen, 1983), researchers are increasingly questioning their
effectiveness as instruments of corporate control (Walsh and Seward, 1990). Jensen (1993),
address
his presidential
to the
Amehcan Finance
in
Association, attributed the inability of U.S.
corporations to adapt effectively to changing competitive and financial environments during the
1980s
to the inertia
Kodak provide
and
confronted with
crisis.
to foster
change
in
corporate strategies and structures, e.^ept
Jensen (1993: 863) argues
board ineffectiveness, with
truth
corporate governance. General Motors, IBM, and Eastman
well-publicized instances of the persistent failure of corporate boards during the
1980s and early 1990s
of
failure of
their
that board culture
is
when
an important component
emphasis on "politeness and courtesy
at the
expense
of
and frankness." However, the question remains: why do boards value consensus and
discourage overt conflicts? What causes
inertia in
corporate governance and the
"failure" of
boards of directors to foster corporate adaptation?
I
will
argue that existing explanations
literature
of corporate
and controversies regarding boards
interests in shaping board decisions yet
logic of
governance are incomplete
of directors
much,
if
has focused on the
— most of the
role of
contending
not most, of board actions are governed by a
appropriateness and rule-following (March and Olsen, 1989). Corporate governance
typically involves not instrumental control
over contending interests, but reliance on the
appropriate rules of corporate behavior. This implies that board decisions are often determined,
not by a calculus of the
norms,
rules,
consequences
of alternative
standardized patterns of behavior, conventions, and organizational routines to the
specific situations faced by the organization.
actions
courses of action, but by the application of
become
institutionalized
and norms
Board routines persist through time as
of appropriate beliefs
1
their
and behavior are established
Boards as Normative Arenas
(Zucker, 1977;
March and Olsen, 1989). The
reliance
and the persistence
in
of routines of
board decision-making implies that the formal mechanisms of corporate governance are
characterized by organizational inertia (Hannan and Freeman, 1984).
Adaptationist theorists have highlighted the importance of
mechanisms by which organizations and
CEO
boards are able
their
succession and selection as
to align the organizations with
environmental contingencies and resource dependencies (Pfeffer and Salancik, 1978;
and Romaneli, 1985). Outsider succession,
organizational
in
particular
change (Helmich and Brown, 1982). But
of either insider or outsiders as
informal routines of
CEO
CEOs
selection.
is
has been seen as a trigger of
this
paper
will
contend that the selection
a routinized process, conditioned by the formal and
These norms and
boards and serve
routines of
CEO
accountability
and
they
capacity to foster organizational change and adaptation.
limit their
reliability of
to
selection increase the
channel corporate actions. At the
This paper examines the persistence of routines of corporate governance
selection of insiders versus outsiders as
to
develop a theory of corporate boards
logic of
CEOs. The
(2) to
persistence of formal and informal routines of
is
research on
CEO
time
the context of
normative arenas, which are guided by a
apply the theory to explain the reliance on and
CEO
selection; and, (3) to
examine how the
mediated by the age, economic performance, and ownership structure
of the corporation, as well as by the timing of
theoretical literatures
in
same
objectives of the paper are as follows: (1)
of directors as
appropriateness and rule-following;
persistence of routines
Tushman
CEO
departure. This paper contnbutes both to the
on corporate governance and organizational routines and
to the empirical
selection.
THEORIES OF CORPORATE GOVERNANCE
Boards as Governing Councils
Following Berle and Means' (1967) classic study of the separation of ownership and control
American corporations, theory and research on corporate governance has focused
in
attention on
Boards as Normative Arenas
the question of
who
controls the corporation
managehal, shareholder, and class interests
theoretical
approaches predominate
and
in
its
board of directors and the relative power of
shaping instrumental control. Three alternative
the organizational literature: agency theory imported
in
,
from financial economics (Jensen and Meckling, 1976;
Fama and
Jensen, 1983; Tosi and
Gomez-Mejia, 1989; Zajac, 1990), focuses on the effects of ownership structure and board
control
in
management, the agents,
aligning the interests of
with the interests of the
shareholders, the principals; managerialist theory (Pfeffer and Salancik, 1978; Fredhckson,
Hambhck, and Baumrin, 1986; Boeker, 1992; Cannella and
role of
management
and
political coalition;
Davis,
1
in
controlling board
Lubatkin, 1993), focuses on the
processes to guarantee survival of the firm and
Useem, 1984;
social structural theory (Mizruchi, 1982; Palmer, 1983;
391) focuses on
''
e interests and position of networks of
elite
its
members
board
in
^
shaping board behavior and decision-making. Agency, managenalist, and social structural
theories of corporate governance vary
managers, or networks of
implicitly
elites
in their
predominate
views of whether the interests of shareholders,
in
board decision-making.
All
three perspectives
agree, however, on the disciplinary role of boards of directors as governing councils,
which control and mobilize resources
in
support of contending economic and
Governing councils are social groupings that mobilize both social
for controlling
and
disciplining the behavior of actors
activities (White, 1992).
Examples
and
capital
political interests.
and material resources
identities in production
of governing councils given by
and exchange
White include Hollywood
production, parliaments, cooperative federations, commissahes, and royal councils. Actors
film
in
governing councils are characterized by their dependencies on the rooted interests of the
diverse factions to which they are
maneuverings
compensation,
examples
that
of
tied.
The essence
of
governing councils are the
these factions around substantial outcomes (White, 1992). Executive
CEO
dismissals, and the adoption of takeover defenses are
of the substantive decisions by corporate directors
boards allocate resources
in
in
some
of the
which past studies have found
support of the interests and allegiances that they represent
Boards as Normative Arenas
(Wade, O'Reilly and Chandratat, 1990; Boeker, 1992; Davis, 1991). The
directors as governing councils highlights their strategic role
ignores the routines which characterize
much
of
in
identity of
boards of
firm decision-making, but
what boards do.
Boards as Normative Arenas
Participants
and observers of board decision-making note
corporate actions, seldom
firm's political coalition,
explicitly
boards rarely
that corporate
members
consider the conflicting interests of the
initiate
of the
and almost always reach decisions by consensus (Mace, 1971; Lorsch,
1989; Bowen, 1994). Boards of directors are subject to norms and rules of appropriate behavior
that structure board meetings
and decision-making. While most
governance focus on the instrumental
existing theories of corporate
control function of board of directors, the obse
.
atioiis of
board behavior present a very different view, one that highlights the rule-bound process of
board decision making. Hirsch (1982; 12)
companies found
or organizations
that directors
in their role
do not
interviews of board
in
think of
members
of Fortune
500
themselves as representing any outside interest
as board members but are self-regulated by a "widely shared
normative structure of appropriate behavior and forms of control." To account for these
observations and for the prevalence of routines
in
corporate governance,
I
will
argue that
boards often serve as normative arenas,' where decision are governed not by a logic of
interests or
consequences, but by a
logic of
appropriateness and social identity (March and
Olsen, 1989; March, 1994).
'
I
use the term "normative" with some concern that
connotes
many
some
might be misinterpreted, as
a Parsonian, oversocialized conception of behavior.
As
it
Garfinkel (1967)
and
others have shown, norms and routines both inside and outside of organizations are
extremely
as
for
it
fragile
rigid rules
and subject
to challenge.
followed by "cultural dopes,
social identity to those that enact
"
Norms
are best understood,
but as tools that provide a
them Norms
of corporate
in
my
sense
of
interpretation, not
meaning and
governance are not external
board members, but are actively created and reproduced by participants
in
the arena.
to
Boards as Normative Arenas
Building
upon both White's (1992) theory
of
arena disciplines and March and Olsen's (1989)
analysis of rule following, institutionalized action, and the logic of appropriateness,
I
will
define
normative arenas as social groupings that create and enforce socially constructed rules and
procedures that represent the social
appropriateness of their
activities.
identity of the participants in the
Examples
of normative
systems, and
editorial
of
and material
what
among
is
or
is
in their ability to
activities of the actors in the discipline in
discipline, distinct
select
and exclude the
terms of socially-negotiated rules
not the right thing to do. Normative arenas are charactehzed by a conversation
a community of participants, one
which the vocabulary of motives
in
on normative commitments rather than on material
presuppose a
corporations, appellate court
in
peer review boards. Normative arenas are a form of
from governing councils, whose control function resides
social
arenas include professional
and licensing boards, functional departments
accreditation
arena and the
common
appropnateness
social identity by participants
of the rules
down, and decisions
interests.
will
is
based
^.rimarily
Normative arena disciplines
and shared assumptions regarding the
and standards. OthenA/ise, arena
disciplines are likely to break
be based instead on the concrete economic,
social,
and
political
networks of relationships of participants within the arena. Normative arenas place strong
reliance
on histoncal precedent
to
guide their decisions. The rules, standards, and principles
which take force within the arena are no:
and
I
beliefs of their participants (White,
argue that boards of directors
and appropriateness
rules, standards,
^
and
fixed,
however, but evolve and change with the values
1992)^
typically act
as normative arenas, concerned with the legitimacy
of board decision-making,
principles prevalent
White's (1992) concept of arena
is
in
and with matching corporate decisions
to the
the organization. They exert control over the
much broader than
that of normative
includes other forms of selection processes including exchange markets.
arenas and
Exchange markets
share with normative arenas a concern with self-regulation based on the values established by
participants
in
the arena, but
in
the former case the values are market pnces while
they are normative commitments.
in
the latter
Boards as Normative Arenas
corporation not by "deciding" on
how resources
stakeholders that constitute the firm's
of the activities
are allocated and distributed
political coalition,
among
the various
but by evaluating the appropriateness
and procedures undertaken by the organization. The view
of
boards as
normative arenas emphasizes the active agency and ongoing achievement required on the part
of board
members
things are
done
in
the reproduction of the rules, routines, and conventions that govern
organizations and
in
example, boards of directors
will
how
organizational
members
play by those rules. For
be concerned with monitohng established procedures
articulation of a strategic direction, plans,
how
for the
and objectives by management, and making sure
that
those strategies selected match the then-prevailing standards of appropriate firm practices. But
contrary to what
is
implied
in
the view of boards as governing councils, directors rarely
initiate
corporate strategies or independently evaluate the consequences of the alternatives
considered, and
it
is
a taboo subject to consider openly the effects of such strategies on the
various organizational stakeholders (Lorsch, 1989).
The
identity of
boards as normative arenas highlights
standardization of
prevalent
in
their role in the formalization
modern corporate bureaucracies. Boards are both a symbol
and
of the rationality
corporate structures (Weber [1922] 1978, Meyer and Rowan, 1978), and a
controller of corporate activities to assure their accountability
Freeman, 1984). These functions presuppose a set
of
and
reliability
(Hannan and
agreed-upon norms and routines of
corporate governance, which serve as precedent and guidance for board decisions. Examples
of informal
norms
of corporate
contacting fellow board
governance include
members
limits
on open
criticism of the
CEO,
not
outside of meetings, and for most boards, using a rhetoric of
shareholder interests to explain board decisions (Mace, 1971; Lorsch, 1989). Examples of
routines of corporate governance include the formal reliance on committees of outside directors
for setting executive
of
all
compensation and
capital expenditure decision
for
above a
monitoring corporate audits, the approval by boards
certain
amount, and the standardization of the
succession and selection process. Routines of executive succession and selection,
in
CEO
particular.
Boards as Normative Arenas
embody
the board's role
power and
in
what Weber ([1922] 1978)
authority of the founding chief executive
the formal position, rather than
in
governance embody the mission,
(Selznick, 1957).
The
normative arenas
is
identity,
shaped by both the
normative commitments that board
70).
Norms and
board members
The norms and
is
invested
understandings, and commitments of the corporation
guides the disciplinary function of boards as
and the
legal fiduciary requirements (Clark, 1991)
members
exercise the duties of loyalty, care, due diligence,
their ultimate responsibility to "do the right thing" (Lorsch,
and
1989:
reliability of
accomplishing these fiduciary duties and normative commitments.
FORMAL AND INFORMAL ROUTINES OF CORPORATE GOVERNANCE
The
in
routines of corporate
routines of corporate governance increase the accountability
in
charisma" as the
becomes depersonalized and
the individual occupant.
logic of appropriateness that
good business judgment, with
calls the "routinization of
routinization of organizational activities
tradition in organization theory
and decisions
is
^
an idea with a long history and
(Weber, [1922] 1978; March and Simon, 1958; Cyert and March,
1963; Nelson and Winter, 1982; March and Olsen, 1989; March, 1994). Although the specific
terminology used differs by vanous authors and includes bureaucratic rules, programs,
standard operating procedures, routines, conventions, and organizational norms, the various
terms connote behavioral, cognitive, and
decision making,
In this
paper
cultural persistence in organizational
use the term routines
I
patterns of behavior, cognition, and values
and procedures, as
of the organization
environmental
deschbe the regular and predictable
organizations. Routines include both formal rules
in
well as informal rules, or norms. Routines are history-dependent, socially
constructed programs of action that
memory
to
behavior and
stimuli.
and
its
embody
the knowledge, capabilities, beliefs, values, and
members and which
Routines both
facilitate
are invoked
in
response
and constrain behavior by
(1)
to
conserving on the
cognitive capabilities of individuals (March and Simon, 1958); (2) providing for accountability
Boards as Normative Arenas
and
reliability of
channeling
organizational activities
political conflict in
The Role of Routines
in
(Hannan and Freeman, 1984); and,
Alternative Theories of Corporate
into the
research
persistence of routines has been found
formal
activities is
a
Governance
commonplace
in
the organizational
That they also characterize major decisions of boards of directors
accepted or incorporated
role, e.g.,
and
organizations (Cyert and March, 1963; Nelson and Winter, 1982).
That routines characterize organizational
literature.
(3) limiting
in
literature.
decisions
is
less-well
Although empirical evidence for the
in
which board members play
at least
a
corporate mergers (Amburgey and Miner, 1993), the theoretical literature on
corporate governance has not acknowledged or accounted for the routinization of board
decisions. Furthermore, the concern with instrumental control
and resource mobilization
that
characterizes existing views of boards as governing councils would imply that board routines,
observed, are an ephiphenomena, a reflection of bounded rationality
interests by the board's
dominant
coalition. In particular, the
in
^
if
the pursuit of matenal
view of boards as governing
councils implies that routines of corporate governance are unlikely to persist as the conditions
that led to their establishment vary, as failures
and as the material
interests of controlling board
While the persistence of board routines
governing councils,
Routines
embody
in
it
is
is
performance trigger search
in
new
routines,
members change.
not easily reconciled with views of boards as
an observable implication of the view of boards as normative arenas.
the normative commitments and fiduciary requirements of board
and, as stated above, serve as tools which increase the
members
for
accomplishing their duties. According to
reliability
this view,
members
and accountability
of
board routines are not readily
immediate material interests of board
amenable
to
members
or to their network of connections. Therefore, an examination of persistence
change and cannot be reduced
to the
board
in
the
routines of corporate governance allows us to test empincally the implications of viewing boards
as normative arenas.
Boards as Normative Arenas
Routines of
The
CEO
reliance on
Selection
and persistence of routines
of
CEO
selection provides an important
testing for the implications of the views of boards as normative arenas. While
case
boards
for
may
delegate to
CEOs
new CEOs
provides an opportunity for realigning the corporation with the controlling interests of
as agents important decisions that occur within
the board of directors.
governance
In particular,
their tenure, the selection of
both agency and managerialist accounts of corporate
CEO
highlight the importance of
succession and selection as a mechanism for
adapting the organization to environmental and economic contingencies (eg, Zajac, 1990;
Pfeffer
and Salancik, 1978). These accounts suggest
that
routinized, but will reflect instead the interests of those
The view
1954; Carroll, 1984). At the
or
who
selection
is
unlikely to
same
is
time, the effects of succession
firm profitability and/ or
CEO
a potentially disruptive activity for organizations (Gouldner,
ambiguous, and even the results
decrease
be
control the corporation.
boards as normative arenas suggests a very different perspective on
of
selection. Managerial succession
highly
CEO
its
of
on corporate performance are
academic research on whether
is likely
it
to increase
survival prospects are equivocal (e.g., Carroll, 1984;
Zajac, 1990). Given the potential hazards
and ambiguity surrounding the
CEO
succession and
selection process, boards as normative arenas focus on controlling for the accountability
reliability of
organizational activities. Boards' reliance on
succession and selection provides a sense of order
commitment
to the
in
norms and routines
and
of executive
governance decisions, communicates
corporate mission and programs, and channels
political
dynamics
into
acceptable practices and procedures.
Formal routines of insider selection.
process
in
large U. S. corporations found evidence for the prevalence of two distinct variants of
formal routines of insider
apparent
Vancil's (1987) descriptive study of the succession
in
CEO
selection
The more common form was the existence
of a heir
the form of a president and chief operating officer (COO), distinct from the
CEO.
Boards as Normative Arenas
Cannella and Lubatkin (1993), following a managerialist theory interpretation of Vancil's study,
found that the designation of a heir apparent decoupled the relationship between firm
performance and outsider succession. The second variant was the reliance on a "horse race"
among
internal competitors for the top spot, usually
presidents.
In
both variants the contender(s) for the
board of directors, affording on-the-job training
members
with the opportunity to gain
termed as vice-chairman or executive vice
CEO
were also members
position
of the
and allowing outside board
for the candidates,
knowledge and experience regarding the inside
contenders.
Both types of insider
internal labor
market
CEO
for
selection described by Vancil (1987) constitute a formalization of an
CEOs, which embeds
the routines of executive succession and
selection within the formal authority structure of the corporation. This internal market for
is
characterized, as are other forms, by promotion from within, the salience of formal job
and career ladders, and a reliance on and
making (Doeringer and
when
Piore, 1971).
rigidity of
Boards
formally designated by both their job
will
title
directors, signal to the internal
contenders
for
CEOs,
rely
for the chief
commit themselves
and external communities
When
1:
The formal designation
officer, or vice
Informal routines of
CEO
candidates
either heir apparents
replacement. Job
their designation
that their
titles
of
as inside
occupants are potential
boards
in
likelihood of insider
of executives as directors
chairman increases the
CEO
faced with formalized internal labor markets
on prevailing succession routines and increase the
operating
CEO
the logic of appropriateness and rule-following prevalent
Hypothesis
to insider
and board membership as
and vice chairman, coupled with
executive position.
titles
formal rules and procedures for decision
(Cannella and Lubatkin, 1993), or as potential contenders for
president, chief operating officer
CEOs
and
likelihood of insider
will
CEO
lead
them
to
selection:
either as president, chief
CEO
selection.
selection. Boards as normative arenas are influenced both by the
organization's formal routines and by
its
informal
10
norms and procedures. The reliance on and
Boards as Normative Arenas
persistence of informal routines
executive succession.
organizational
response
in
First,
memory and
is
influenced by several factors, which are not exclusive to
the adoption of an organizational routine increases
increases the likelihood that
future circumstances (March
Amburgey, 1991). Second, the
bargaining and open
will
availability in
be invoked as an organizational
and Simon, 1958; Walsh and Ungson, 1991;
reliance on
political conflict in
it
its
and persistence
of routines
Kelly
and
reduces the costs of
organizations, as routines constitute a form of "truce"
(Nelson and Winter, 1982: 107-112), with stable accommodations and quasi-resolution of
among
conflict
the divergent interests
adoption of routines signals their
become
in
the organization (Cyert and March, 1963).
political viability
and appropriateness.
Third, routines
infused with value, as symbols of the organizational mission and as
institutional
purpose ^Selznick, 1957), increasing the social
The
can
embodiments
of
identification of the organization's
executives.
These
routines
These
and motivational factors
cognitive, political,
may
vary
effects
governance
in
in their
that influence the persistence of informal
strength for routines of either insider and outsider
must be understood
in light
CEO
selection.
of the broader institutional field of corporate
which publicly held corporations operate, where routinization occurs not only
the level of the individual organization but at the organizational
1983). Insider succession has
particularly for larger
become
the (statistical)
norm
(DiMaggio and Powell,
for U.S. publicly held corporations,
companies. While executive succession
effects are likely to matter as experience with insider
field
at
is
a commonplace, firm-level
and outsider selection
will alter
board
decisions through the routinization of charismatic authority and the separation of the firm's
identity
from that of
its
founder(s). Following
Amburgey and Miner
(1992),
I
organizational experience as an indicator of reliance on prevailing routines.
political effects of
outsider selection experience at the firm level
insiders,
however, as
activities
and signals
its
its
adoption both increases
political
its
may be
rely
The
on past
cognitive and
greater than for
availability in the board's repertoire of
appropriateness within the focal organization. The motivational
11
Boards as Normative Arenas
effects of insider
and
CEO
between the board and top executives
implicit contracts
internal labor
selection are likely to be greater, nonetheless, as informal
markets
for
CEOs. The
examine separately the
subsequent
effects of insider
CEO
and motivational effects
selection on
and outsider
firm-level
norm formation lead us
insider selection increases the likelihood of
subsequent
selection.
CEO
selection.
^
likelihood of
of insider succession
subsequent
^i>
'
Note that while views of boards as governing councils could account
1
based on
either the superior
for the reliance
performance of insiders as
CEOs
990) or the power of insiders over the board of directors (Cannella and Lubatkin,
can less
to
succession expehence on
Hypothesis 2b: Past reliance on outsider selection increases the
outsider
of
CEO selection:
Hypothesis 2a: Past reliance on
insider
CEO
more formal
paralleis the effects of
different cognitive, political,
reliance on routines of insider versus outsider
commitments
directly
account
for the
1
on routines
(Zajac,
993), they
existence of both insider and outsider selection routines.
PERSISTENCE OF ROUTINES OF CORPORATE GOVERNANCE
The view
of corporate
boards as normative arenas emphasizes not only
their reliance
on
routines for decision-making, but the persistence of these routines under diverse set of
economic and environmental
conditions. This persistence
is
closely related to the function of
the board of directors as controllers of organizational accountability and
must distinguish between theories
of
bounded
rationality
and
political
albeit
one
limited
by
constraints on rational decision making, and theories of rule-following
the institutionalization of action (March and Olsen, 1989),
structured, socially-constructed
reliability
Here we
(March and Simon, 1958; Cyert and
March, 1963), where routines result from a logic of consequences,
cognitive
reliability.
and accountability
commitments
of the
board
where routines convey
to the organization's
of directors'
12
commitments
and
highly-
purpose and agenda The
to the organization's
mission
Boards as Normative Arenas
entails a high
degree of persistence and
(Hannan and Freennan, 1984). To
adaptability,
routines of
whether
that
CEO selection
CEOs
Effects of
The
and those
inertia in
distinguish
emphasize
the routines of corporate governance
between accounts
their inertia,
I
will
of routines that stress their
evaluate
how the
persistence of
are affected by firm's performance, age, ownership structure, and by
are dismissed or depart prior to normal retirement.
Economic Performance on Persistence and Change
effects of
economic performance on persistence and change
in
governance routines
provide a strong test of the theory of boards as normative arenas and of the logic of
approphateness and rule-following
in
adaptive models of decision-making
programs and routines
in
response
corporate governance.
is
to
An important component
their hypothesis that organizations
change
of
their
*
environmental change and failures of economic
performance (March and Simon, 1958; Cyert and March, 1963). But the view
of
boards as
normative arenas implies a different perspective, one where appropnate behavior under periods
of
economic adversity implies a continuing commitment
and routines.
In
to the boards'
values, norms, beliefs,
discussing boards' commitment to norms and routines of
their social identification with the rules
CEO
and culture of the organization, and not
selection
their
I
mean
personal
allegiances to individual incumbents of the executive positions. Boards as normative arenas
respond
to
economic
whether they should
determining what
is
adversity, not by evaluating the
persist with their past choices or
economic and
political
consequences
of
whether they should change, but by
the appropnate thing to do as determined by the prevailing rules and
routines of the organization.
The
implies that board decisions
will
commitments. This form
of
reliance on rules
be governed by
and routines under economic adversity
their past
experience and organizational
board behavior can be characterized as form of a
threat-rigidity
response (Staw, Sandelands, and Dutton 1981), rather than failure-induced change (March
and Simon, 1958; Cyert and March, 1963).
It
does not mean, however,
13
that
boards
w'll fail to
Boards as Normative Arenas
act under adversity, but rather that their actions
embedded
in
The theory
of boards as normative
be guided by an
will
the culture of the organization and
in its
rules
institutional logic,
and routines (Ocasio, 1995).
arenas and theories of adaptive change consistent with the
view of boards as governing councils provide contrasting hypotheses regarding persistence and
change
in
routines of
CEO
selection:
Hypothesis 3a: Under economic
routines for insider
and outsider
adversity, corporate boards
CEO selection.
Hypothesis 3b: Economic adversity leads
routines of insider versus outsider
CEO
on formal and informal
will rely
(Boards as normative arenas).
to failure-induced
change
in
the organizational
selection. (Boards as governing councils).
Firm Age and the Persistence of Routines
The view
of
boards as normative arenas implies that the persistence of governance routines
increases with the age of the organization. The mission and purpose of the organization
becomes embodied
become
in
the normal and rules of the organization, and organizational routines
taken-for-granted and less subject to question or challenge.
Hannan and Freeman
(1984, 1989) argue that organizational histories generate constraint on structural change by
providing legitimate justifications beyond self interest for opposing change, and precluding
considerations of alternatives repertoires. Consequently, the reproducibility of organizational
routines increases monotonically with age. Furthermore, pressures toward internal consistency
and homogeneity
are
likely to
relationship
Hypothesis
of
members' perceptions also suggest
that organizational
norms and
routines
increase with age (Aldnch and Auster, 1986). These arguments imply a direct
between
4:
firm
age and the persistence
The persistence
of routines of
of routines of
CEO
14
CEO
selection;
selection increases with firm age.
Boards as Normative Arenas
Structure of Firm Ownership and Routines of
One
CEO
Selection
possible explanation for the existence and persistence of routines of
CEO
selection
is
they result from agency problems associated with the separation of ownership and control
U.S. corporations.
boards
In this interpretation,
shareholder interests because neither board
fail
to act
as governing councils
members nor management have
in
the proper incentive structure for board
While
prior
research
vs. outsider
effects
CEO
in
CEO
sufficient stake in
norms
corporate governance results from the lack of
inertia of
members and managers (Walsh and Seward,
selection has
in
protection of
the firm's ownership to act vigilantly (Jensen, 1993). According to this view, reliance of
and routines of decision-making and the
shown
that the structure of
1990).
ownership affects insider
succession (Boeker and Goodstein, 1992; Cannella and Lubatkin, 1993),
on the reliance on routines remains unexplored. The view
arenas suggests that increased
firm
that
of
its
boards as normative
ownership by management and board members, rather
than increasing their reliance on a logic of consequences for board decision, further enhances
their
commitment
to the
norms and values
of the corporation
and
to the organization's routines.
According to the view of boards as normative arenas, any increased commitment to the
corporation resulting from firm ownership
commitment
to the
norms and
is likely
to translate instead into
routines of the corporation.
Two
alternative
increased
hypotheses are
presented to explore how the structure of ownership affects the reliance on routines of
CEO
selection:
Hypothesis 5a: Increased ownership by management increases
reliance
on routines
of
CEO
selection. (Boards as normative arenas).
Hypothesis 5b: Increased ownership by management decreases
selection. (Boards as governing councils).
15
reliance on routines of
CEO
Boards as Normative Arenas
Early
CEO
Departures and Dismissals and Reliance on
CEO
Selection Routines
Political or
managerialist theorists of corporate governance and executive succession often
distinguish
between the
effects of
CEO
dismissals and early departures, and other forms of
succession (Fredrickson, Hambrick, and Baumrin, 1986). Empirical studies have supported
view, finding that the determinants of
retirement
(i.e.,
64 years and
older)
CEO
succession and selection vary for
CEOs
this
facing
and younger executives (Puffer and Weintrop, 1991;
Cannella and Lubatkin, 1993). This managerialist interpretation implies that the persistence of
routines of
of
CEO
CEO
selection
more
likely to
hold for the normal process of retirement.
dismissals and early departures, corporate governance
sociopolitical process,
Alternatively, the
and the view
of
is
CEO
routines that are invoked
the case
boards as governing councils would hold.
CEO
selection
departures as well as to retirements. The norms and repertoires of
may
vary under the two sets of circumstances, however. Note that the
appropriateness and rule-following that underlies the theory of boards as normative
arenas does not imply that these norms and routines are invariant
to the external
contingencies faced by the organization. Different rules and routines
different circumstances. For
dismissals and other for
example, a different set of norms
CEO
may be
may be
retirements. In the former case, boards
of insider succession.
and
may
internal
applied under
appropriate for
reliance on routines of outsider succession. In the latter case, they are
norms
In
best understood as a
view of boards as normative arenas implies that routines of
apply to cases of early
logic of
is
CEO
increase their
more
likely to rely
on
According to the view of boards as normative arenas, norms and
routines are likely to affect board decision making under either circumstance.
To
distinguish whether the view of boards as normative arenas applies only to cases of
retirement, or
more
generally,
Hypothesis 6a: Early
CEO
I
offer the following set of alternative
hypotheses:
departures increases reliance on routines of outsider selection and
decreases reliance on routines
of insider selection
16
(Boards as normative arenas).
Boards as Normative Arenas
Hypothesis 6b: Early
CEO
outsider
CEO
departures decreases reliance on routines of both insider and
selection. (Boards as governing councils).
METHOD
Sample.
A random
sample of 120 U.S.
industrial corporations or
the Moody's Industrial Directory for 1980
population listed
in
sample includes
publicly held
companies
listed in
exchanges and contains a broad representation
of observation
is
ownership data
108
firms.
became
Not
the
company
New
companies
companies had data
publicly held after 1960.
for the
Many
in
York, American, or regional stock
of firm size
and ownership
Given lack of
complete period. Many were
became
The
unit
financial
and
structure.
the original sample, the sample
others merged,
total
selected for the analysis. This
year, covering the years 1960-1990,
for fourteen of the
all
the
was
4.45 percent of the
was reduced
to
found«=>d and/ or
bankrupt, went private, or
otherwise ceased to be publicly held companies duhng the decade of the 1980s. The
methodology
treats
any spells ending
ownership as right-censored
in
bankruptcy, acquisition, and change to private
at that point. This implies that the effects
being measured relate
exclusively to normal forms of succession within the current ownership of the firm.
sample used included 2,287 company-years
The sample was selected as
of
high-techology companies, to
sample selection
bias,
1980
total
of data.
to permit corporations
become
The
founded since 1960, including
part of the sample. This creates
some
potential for
as firms that disappeared between 1960 and 1980 were excluded from
the sample. Sampling as of 1960, however, would have excluded newer firms from the sample,
which then would have been less representative
of firms in 1990.
Sampling
in
1980 was
selected as a compromise solution that would both reduce sample selection bias and produce a
representative sample of industrial firms
Independent Variables and
CEO
in
1990.
Selection Events.
17
Boards as Normative Arenas
CEO
CEO
succession and selection.
succession events were coded from Standard and
Poor's Directory of Corporations. Officers, and Directors based on changes
relevant officers.
CEO
selection
A
total of
216 succession events occurred
Following Cannella and Lubatkin (1993),
.
had been employees
of the
company
for at least
in
the sample. Insider vs. outsider
CEOs were coded
were appointed
employment
CEOs were
in
history.
if
they
reports,
and Who's
some
in
instances,
preparation for becoming
and
Directory of Corporations. Officers,
supplemented by proxy statements, 10Ks, annual
verify prior
as insiders
was undertaken because,
outsiders are appointed to the corporation for a short period of time
Finance to
of the
two years phor to becoming CEO; otherwise
they were classified as outsiders. This procedure
CEOs. Data from Standard and Poor's
names
the
in
Who
in
Industry
and
Outside members of the board of directors
classified as outsiders.
A
total of
was
Directors,
who
157 insider succession events and
59 outsider successions were recorded.
Routines of
CEO
CEO
internal labor markets, for insider
directors, other than the
officer, or vice
board
in
CEO. A
chairman was
succession were coded based on the job
formal ILM
listed
of formal routines, or
was recorded when
as both an officer of the
inside
a president, chief operating
company and
a
member
of the
.
CEO
succession were
in
place
Succession Routines
examining whether the
outsiders, or
whether
it
firm
has
.
I
in
65.4
%
of the
measure past
CEO-years
Directors
.
CEOs
of
in
the sample.
reliance on informal
relied exclusively in the past (since
has experienced both. The
history of
1950 was recorded from Standard and Poor's Directory
new companies were
insiders or outsiders, using the coding
all
CEO
CEO was
selection routines by
1950) on either insiders or
CEOs
in
the
company
of Corporatio ns. Officers,
classified as founders
since
and
Others were classified as
scheme and data sources described above Four
mutually exclusive categories of past succession experience were coded:
current
titles of
Standard and Poor's Directory of Corporations. Officers, and Directors Formal ILMs
for insider
Past
CEOs. The existence
Selection. Formal ILM for
the founder. Past insiders
,
when
18
all
CEOs
None when
,
the
since 1950, other than the founder,
Boards as Normative Arenas
were
insiders, Past outsiders
outsiders,
and Both when
,
sample 52.0%
of the
15.3% as Both
,
when
CEOs
all
CEOs
since 1950 included both insiders and outsiders.
company years were
insiders
since 1950, other than the founder, were
classified
will
the
as Past insiders, 12.6% as Past outsiders,
and outsiders, and 20.1% as None. The
informal succession routines
In
test of the
hypotheses on
focus on the effects of Past insider and Past outsider, and
interaction effects with performance, firm age,
CEO
ownership structure, and
age.
Moderators of Persistence of Routines. Performance. Return on assets (ROA), adjusted
industry average, obtained from
COMPUSTAT, was
measure, following past studies of
(1990) finding that relative
and unadjusted
ROA was
used as a measure
ROA
.94
CEO
and the
on
was
do not change materially
ROA were
averages were also estimated, but a one year
.
used as an economic performance
compensation. The correlation between adjusted
results
of performance. Data
Firm age The firm age
.
firm
CEO
CEO
Directors, proxy statements, lOKs,
"voluntary" departures
CEO
in
year
Standard and Poor's Directory of
initial
founding or incorporation.
.
CEO was
it
and
and Who's
Who
63 years
old or under.
in
Industry
would have been preferable
companies.
Control Variables
19
and
and Finance, and the Wall
theoretically to distinguish
"dismissals", Wall Street Journal
I
Data on the
Directory of Corporations. Officers,
between
and other published sources
departure are either incomplete or unreliable for a large number of firms
particularly for smaller
five
is
Departures Following Puffer and Weintrop (1991),
CEO's age was obtained from Standard and Poor's
.
Two and
ROA
ownership was obtained from Value Line Investment Survey and from
departures as those were the
Street Journal Although
unadjusted
lag provided the best results.
calculated from the date given
corporate proxy statements. Early
classify early
if
lagged one year.
Corporations. Officers, and Directors of the year of the firm's
Ownership Data on
for
succession and selection and Gibbons and Murphy
CEO
affects
its
in
for
the sample,
Boards as Normative Arenas
I
included control variables for firm size, as well as characteristics of the
year of
hire),
directors).
and
The
chairman and proportion of outside
was based on
the findings of recent managerialist
selection of control variables
and Lubatkin, 1993).
in
selection (Boeker
was obtained from COMPUSTAT.
Industry
and Finance Tenure measured the number
.
1960 are subject
to left-truncation,
included, beginning
address the problem of
incumbents
in
measure
we
as
variables prior to 1960
all
CEO-years were
the duration
in
was
first
on phor
remainder of the CEO's tenure was included
the estimates and
in
Size
in
the sample,
full
all
was excluded. To
tenure were recorded for
all
the sample. This procedure reduces the bias
in
the case of the piecewise-exponential model used, leads to consistent
estimates (Tuma and Hannan, 1984; Quo, 1993). The results of
CEOs
appointed prior
the tenure spells began, but the
CEO
Who
the sample for each company, and only the
in
in
and Who's
CEOs
models estimated.
when
and
incumbent has served as
of years the
not recorded. For firms
incumbent
other control
reports,
1960, but the tenure prior to 1960
left-truncation, data
1960, or for the
in all
knovw
All
Directory of Corporations. Officers,
supplemented by data from proxy statements, 10Ks, annual
information on
on
and Goodstein, 1992; Cannella
except employment data, were recorded at the
were obtained from Standard and Poor's
CEO. Tenure was used as
to
CEO
All control variables,
beginning of the year. Data on firm size
Directors,
and
(tenure, age,
of the board of directors (separate board
interpretations of insider versus outsider
variables
CEO,
this
procedure are conditional
having survived to 1960.
was measured as
expectation
is
the loganthm of the
that larger firms
takes the value of zero
if
CEO
the
equals the current age of the
age, age has no effect on
will
is
of
employees
have a greater degree
is
60 years
variable
assumes
to
succession The effects of age on outsider
have a
CEO
positive effect
selection
20
Year
-
that
it
thousands). The
CEO
and Age
succession or selection, but that
expected
(in
of insider
old or under,
CEO. The Age60
CEO
each year afterward. Age60
number
selection..
Age60
60, otherwise,
up
where Age
to the sixtieth
year of
has increasing effects
on the rate of insider
of hire
for
CEO
measured the year the
Boards as Normative Arenas
CEO was
and was intended
hired (minus 1900),
succession.
A
positive trend
was expected,
to capture historical trends in the rate of
taking into account increased pressures
CEO
CEOs
on
during the 1960-1990 period studied, particularly for outsider succession. This variable
assumes
that
that
CEOs
are affected by changing norms and expectations about
were prevalent duhng the beginning
remainder of
of time trend,
number
which assumes that changing norms
of directors.
chairman of the board of directors
Sample
statistics. Table
1
affect
measured the number
Chairman
is
a
dummy
differs
and
much
their tenure. This variable provides a
Proportion of outside directors
total
of their tenure
CEO
succession
that these rules are set for the
better
fit
than an alternative measure
new and incumbent CEOs
of outside board
members
variable that takes the value of
from the CEO, and
0,
equally.
divided by the
1
when
the
otherwise.
presents the sample means, standard deviations, and the Pearson
correlation coefficients for the variables
used
in
the analysis.
The
unit of
observation
is
the
company-year.
Insert
Table
1
about here
Modeling Procedure
I
tested the hypotheses by specifying continuous-time, event history analysis of the competing
risks of insider
versus outsider
transition rates to
CEO
selection. This
and outsider selection and
that the
and
two events are the
trajectories.
CEO
risk
model estimates the
two mutually exclusive events, insider succession and outsider succession.
The competing hsk methodology allows
insider
competing
A
CEO
result of highly routinized
this contention,
selection
the year of turnover, which
is
is
separate mechanisms that yield
consistent with the underlying assumption of this paper
processes with different underlying causes
chi-square contrast of the competing
succession supports
history analysis of
is
for estimating the
and
is
risk
model with the pooled model
available from the author.
preferred to the
more common method
subject to sample selection bias and implicitly
21
The use
of
of
of event
sampling only
assumes
for
Boards as Normative Arenas
equilibrium
using
in
the
maximum
CEO
selection process
(Tuma and Hannan,
TDA
package
likelihood with the software
1984).
The models are estimated
(Rohwer, 1991).
TDA
allows
estimation of models with time-varying covariates and takes right-censoring into account by
using the information provided by the cumulative survival time of censored spells
(Tuma and
Hannan, 1984).
Functional form. The use of event history analysis requires that specification of the functional
forms for the transition rates estimated. The selection of functional form must also account
left-truncation in the
sample, as
CEOs
appointed prior to 1960 are part of the sample. To
achieve consistent estimates with left-truncated data,
(Guo,
199"^).
The hazard
risks of insider
rates
r(t)
of
for
CEO
and outsider succession
I
use a piecewise-exponential model
succession are defined under the two competing
and
r(t)^
r(t)2
,
respectively, as follows:
Piecewise- exponential model:
r(t),
= expfa,,+.... + ap, + p,X, +
r(t)2
= exp
fa,
+....
,
+ a^^ + P,Z,
Note that the independent variables
X and
..+
p^J.
+...+
p^J.
Z, respectively, are not restricted to
piecewise exponential
I
competing
for the
be the same.
tried several alternatives.
The
periods for both insider and outsider succession;
years,
and 17 years and above. There are
three periods, but the fourth and
fifth
Alternative
models estimated.
including
control variables, the
all
performance (ROA), ownership
This
last
I
risks of insider
-
In
and outsider succession
selecting the time periods a,p for the
results presented include five time
2 years, 2
7-12
7 years,
-
statistically significant
years,
12-17
differences between the
first
are not statistically significantly different form the third.
first
estimate a baseline model for comparison purposes
main effects of the moderating variables of
structure,
and an
interaction effect
between
firm age,
ROA
and age 64.
covariate controls for the findings of past studies were performance had limited effect
on succession
for
CEOs
facing retirement.
To
test
22
Hypotheses
1
and
2,
I
then estimate a model
Boards as Normative Arenas
main
that includes the
effects of multicollinearity
separate models
in
for the effects of
interaction effects with the
A full
model
CEOs
performance (Hypothesis
5). In
measures
Formal ILMs, Prior
of
combined
63 years
old
the .10 level
is
and under, and compare the
shows the baseline model
formal and informal routines of
The
include separate
insider. Prior outsider,
is
and Both
the omitted
be
estimate the model for the subsample of
full
sample.
CEO
Selection. Table 2 presents the results of the
for
CEO
CEO
selection
for
and provides
tests for Hypothesis
comparison purposes. Model 2 presents the effects
selection.
The combined
results of
and informal
CEOs, increase
significant at the .001 level.
Model
statistically significant at
of board decision-making as normative arenas,
firm's formal
*
Table 2 about here
2,
1
-5.
of
with a chi-
the .001 level.
governed by the rules
routines.
findings strongly support Hypothesis
form of ILMs
I
results to the
square contrast of 39.7 with the baseline model, are
They sustain the model
and
'
competing nsks of insider versus outsider
in
4),
~
Formal and Informal Routines of
embodied
estimate
I
age (Hypothesis
I
the
limit
also estimated. Finally, to estimate whether the effects
Insert
1
To
selection.
effects of those variable found to
selection routines apply only to retirements,
RESULTS
Model
3), firm
estimating these moderators,
that includes the
statistically significant at
CEO
CEO
and outsider) succession experience. (No succession experience
category).
of
of
evaluating the effects of moderators of persistence,
ownership structure (Hypothesis
(insider
and informal routines
effects of formal
1
t.iat
formal routines of insider
the board's selection of insiders as
CEO
selection, in the
CEOs. These
effects are
Hypotheses 2a and 2b are supported with past routines
selection increasing subsequent insiders, statistically significant at the .05 level,
of insider
and past
outsider routines increasing outsider selection, significant at the .01 level. Informal routines of
23
Boards as Normative Arenas
both insider and outsider selection affect board decision on
stronger
in
Note that
if
size
and
statistical significance for routines of
CEO
selection, with the effects
outsider selection than for insiders.
the firm has had experience with both insiders and outsiders as
CEOs,
this
increases the rate of both subsequent insider and outsider succession. These results are
weaker, however, than
outsiders.
The
in
the case of exclusive reliance on routines of either insiders or
findings suggest that both cognitive
and
political effects of
routines affect board decisions, as past experience with
subsequent
utilization
when
greater
and
political viability
1
that these effects are
either insider or outsider selection dominates.
about here
1
presents graphically the effects of the formal ILMs for
routines on the competing risks of insider versus outsider
are estimated from Model
pattern of findings
show
2,
with
all
independent of each other.
selection.
past
The
CEO
selection
transition rates
of
CEO
of formal
selection
The
have large
and informal routines are
a separate analysis, not presented here but available from the
between the formal and informal routines
not found to be statistically significant).
compahsons
and informal routines
on board decision. The effects
(In
author, the interaction effects
CEO
CEOs and
the control variables estimated at sample means.
that both formal
statistically significant effects
the pairwise
selection
selection routines increases their
and appropriateness, but
Insert Figure
Figure
CEO
CEO
The most
of insider versus
CEO
of
CEO
selection
relevant findings are the differences
selection for
formal and informal selection routines. First note that
in
each
of the eight
the case of no routines
were
between
combinations of
(i.e.,
founders
without internal ILMs) the transition rates for insider and outsider selection are approximately
equal (.021 vs. 018, for insiders and outsiders, respectively). But with an ILM but no
succession experience, the rate of insider selection
.028)
Past insider selection with an ILM
(.109 vs
is
is
almost double that of outsiders (.052
vs.
the condition most favorable to insider selection
040) Past outsider selection without an ILM
24
is
most favorable
to outsiders (.027 vs.
Boards as Normative Arenas
.080). In the
for
CEOs
case of both insider and outsider experience, the presence or absence of an
alter the
are more
likely
dominance
of
subsequent insiders or outsiders: without the ILM, outsiders
(.040 vs. .062), with the rate of Insiders
Insert Figures 2
Effects of Performance
is
higher (.099 vs. .094).
and 3 about here
on the Persistence of Routines. Model 3 analyzes
CEO
performance on persistence of the formal and informal routines of
chi-square contrast between Models 3 and 2
significant at the .05 level.
The
is
Hypothesis 3b, on failure-induced change
outcomes
(Milliken
main effects of
and
likely to persist
ROA and
its
is
CEO
negative and
existing routines.
The
effect of
show
that formal
one must account
ROA
(-.088,
which
and informal
is
the
statistically significant at the .01 level for
for succession, firms facing
findings
no
prior
economic adversity
full
CEO
for both the
selection
effect in the
case of no
outsider selection, and
experience with or formal
will
more
likely
choose outsiders
rather than insiders, consistent with the existence of a generalized
norm
under poor performance. Formal ILM are invanant
effects, with small
statistically insignificant interaction effects with
to
performance
effects of
little
evidence that performance triggers a failure-induced change
routines
CEO
CEO
The
succession
and
selection routines, while not unequivocal,
and are a more consistent with a view
selection decisions.
for outsider
both insider and outsider selection.
The
performance on informal
for
under both positive and negative performance
positive but not significant for insider selection. With
mechanisms
under economic
and provide no support
selection
interaction effects with the formal
main
The
19.21, with 8 degrees of freedom, statistically
Lant, 1991). In interpreting these effects
routines. First note that the
routines)
in
the effects of
selection.
results are consistent with Hypothesis 3a, that
adversity boards rely on dominant routines for
and informal routines are
ILIVI
that
boards
rely
interaction effects of past insider
25
in
show
the application of dominant
on dominant responses
and ROA, negative
for their
for insider
Boards as Normative Arenas
and positive
for outsiders, although not significant, are inconsistent with adaptive
The
insider selection routines.
and
significant for insiders,
to the
main
economic
ROA
is
for past insider
shows
and that when
likely to persist
and past outsider selection routines
that
ROA
interaction effects are best
is
economic adversity increases both
when added
under
understood graphically
Figures 2 and
insider
3,
and outsider
CEO
30 percentage points below the industry average, boards are
with insider selection. Figure 3
performance
of outsider selection, poor
in
in
negative, but not
significant at the .10 level for outsiders, yet
The combined main and
adversity.
respectively. Figure 2
more
and
positive
and
effect yields a negative effect, implying persistence with outsider routines
and are shown
selection
interaction effects of past outsider
change
is
more
shows
likely to
that for firms with
dominant routines
increase persistence with outsider
selection, consistent with Hypothesis 3a.
Insert Figure 4
Effects of Firm
Age on Persistence
the persistence of
persistence of
CEO
CEO
about here
of Routines. Model 4
selection routines.
The
shows the analysis
of
selection routines for the case of past insiders (Hypothesis 4), but
freedom, which
statistically significant is
is
significant at the .05 level in
political
to insider
becomes
is
The mediating
increasingly
10.0,
effect of firm
embedded
in
with 8
is
age on the persistence
is
of
its
consistent with the
The dominance
of insider
the firm's political coalition and institutionalized
of the organization (Selznick, 1957).
become
no
same term becomes
Figure 4, indicates that as the firm grows older
and mission
less likely to
is
only interaction term that
(Note that the
effects of insider selection routines.
within the social structure
outsider routines
in
6).
level.
The
succession increases This increase over time
and motivational
selection
Model
shown
insider selection routines,
commitment
not statistically significant.
past insider, at the .10
age on
results provide limited support for increasing
support for the other cases The chi-square contrast between Models 4 and 2
degrees
of firm
The dominance
of
institutionalized overtime, lacking internal constituents
26
Boards as Normative Arenas
to support
its
interaction effects of
to
statistically significant,
the direction of the
that the effects of formal routines are larger for
because they are more
result both
true of formal routines of selection. While the
age and formal ILM are not
some evidence
effects provide
may
The same may be
perpetuation.
salient
when
there
is
newer
firms. This
no realized succession experience
guide board decision-making and because for rules to become institutionalized over time
not sufficient that they be part of the formal procedures, but
it
is
it
is
also necessary that they be
consistently practiced.
Insert Figure 5
about here
Effects of Board Ownership. Model 5 presents the results of insider ownership on the
adoption of routines of
CEO
selection,
The chi-square
with 8 degrees of freedom, significant at the .05 level.
contrast between
The
Models 5 and 2
is
results support Hypothesis 5a,
20.0,
on
increased reliance of inside selection routines for insider ownership, and no support for
hypothesis 5b. The interaction effects of insider ownership and insider selection routines
positive, albeit not statistically significant, for insider selection
significant at the .10 level for outsider selection (This result
the .01 level
in
Model
6).
and negative and
becomes
is
statistically
statistically significant at
Also note that the interaction of combined insider and outsider
selection experience with insider ownership has a positive effect on insider selection, significant
at
The
the .01 level.
results
do not support Jensen (1993) contention that
cultural persistence in
corporate governance results from the lack of ownership incentives by board
executives.
board
On
the contrary, as
members
in
Figure
5,
in
CEO
comes
the rate of outsider selection, while the rate of insider selection
invariant to insider ownership.
Thus
insider
top
increased ownership by managers and
increases the proportion of insiders selected as CEOs. Note that this
about through a decrease
and outsider
shown
members and
is
ownership decreases the combined rate of insider
selection, thus increasing the relative reliance
27
on routines
of insider
Boards as Normative Arenas
agency theory. No
selection, a set of findings not consistent with
were found on reliance on
either formal
Combined Model. Model
6 combines the
chi-square contrast with Model 2
the .01 level.
When
is
ILMs or routines of outsider
40.4 with 10 degrees of freedom,
even
selection,
if
was found
is
strengthened. This
omitting interaction terms that
effect of past insider
effect of
insiders
and persistence on routines
corporation.
The main
significant, supporting
due
were not
The
3-5.
statistically significant at
significant
I
include
its
effects
Models
except that the
2-5,
statistical
to the reduction in multicollinearity that results
Note that as
statistically significant.
in
Model
4,
from
the main
small, negative, but not significant, while as stated earlier, the interaction
is
age and past
is
be
Models
both terms were not significant. Note that the
results are substantially equivalent to those found in
significance
to
ownership
selection.
statistically significant effects of
a particular interaction term
on both insider and outsider
effects of insider
is
now
significant at the .05 level. This
suggests that the reliance
of insider selection are strongly conditioned
effects of formal
Hypothesis
1
on the age of the
ILMs and outsider selection routines are positive and
and
2b, respectively.
Insert Table 3 about here
Effects of Early
full
CEO
sample, from Table
Departures. Table 3 presents a compahson of Models 2 and 6
3,
with the analogous
Models 2a and 6a,
for the
CEO
years old and under. The results support Hypothesis 6o that early
reliance on informal routines of outsider succession
insider succession.
outsider
CEO
Note that the coefficient
selection,
and
larger
in
the
2a, but large, positive,
rate
and
significant ^t the
is
positive for both insider
is
small, negative,
The
in
is
and not
28
full
of
and
significant in
also positive,
size of these coefficients are also larger than for the
63
either case. For past
01 level for subsequent outsiders.
on combined selection experience on outsider succession
the .05 level.
CEOs
departures increases
although not significant
outsider selection, the transition rate to insiders
of
and decreases reliance on routines
of past insiders
latter,
subsample
for the
Model
The
transition
and
significant at
sample
in
Model
2.
Boards as Normative Arenas
This suggests that
the case of early
in
CEO
departures, whether by dismissal or voluntary,
routines of past outsider selection have a large and significant effect on board decisions to
Model 2a
select outsiders. Also note from
in
the
of firm
full
that formal
ILMs increase insider
sample. Model 6a provides results similar to those of Model
age and past
6.
CEO
The
selection, as
interaction effects
insider are similar as before, albeit significant only at the .10 level.
Control variables. The effects of the control variables are unremarkable with one exception,
the effects of proportion of outside directors. Tenure Both insider and outsider succession were
.
found to increase with tenure, although the effect
the case of
CEOs
is
nonmonotonic
for outsider selection or in
63 years and under. Performance had a negative
and
effect
on both insider and outsider succession. Note however, that without controlling
succession experience, the size of the effects
than 63. The effects of performance on
succession experience, as well as by
significant for insider
that the effects of
Lubatkin, 1993).
CEO
CEO
and negative but not
of hire
was
positive
similar for both,
when
CEO's age
the
age.
Age64*ROA was
positive
and
successors. Chairman
magnitude.
was
It
was
was
is
less
statistically
significant for outsiders. This confirms earlier findings
differs for
and
CEOs
facing retirement (Cannella and
significant for both insider
and outsider
selection, although larger for the latter, suggesting both a general increase
and increasing reliance over more recent
for past
selection are therefore strongly mediated by past
performance on succession
Year
is
significant effect
CEO
in
CEO
succession
cohorts on the selection of outsiders as
positive for both insider
and outsider selection and
statistically significant for insider selection in
all
of similar
models, but the significance
not robust across models for outsider selection. This suggests, consistent with the findings
of Cannella
succession,
and Lubatkin (1993),
it
has
limited effect
size increased insider selection
that while having a separate
Chairman increases
on the selection of insiders versus outsiders. Size Large
and decreased outsider
selection. Firm
age were mediated by past succession expenence with no consistently
Proportion of insider ownership
CEO
No
age The
.
29
effects of firm
significant
statistically significant effects of insider
firm
mdin
effect.
ownership were
Boards as Normative Arenas
found. Boeker and Goodstein (1992) also failed to find statistically significant effects, but they
which
did find effect for the concentration of insider ownership,
Proportion outside ownership
.
The proportion
of outside
owns
The main
were not
findings
effects of proportion of outside directors
in
the two studies
may be
sampled both phvately-held and
1969-1989. Both for
representation
in their
and other forms
interpreted
th'^
publicly held
of
statistically significant,
and
dependencies
for
likely to
(Pfeffer
is
more
the case across the population studied here of
view of board as normative arenas
contrary to
boards
Silicon Valley
represent ties to venture capitalists
of executive succession
activists
board members, the
be applicable. This
is
may have
limited effect
on
CEO
and
political
less likely to be
publicly held industrial firms
all
from
and Salancik, 1978). Their findings were
likely to
may be more
in
smaller publicly held firms, outsider
and those with more
boards as governing councils
political structure of
bring an
possible explanation for the disparate
and are more consistent with a managerialist model
where the the
likely to
the stock, to lead the corporation.
semiconductor firms
sample may more be more
selection. In both privately-held firms
model
One
here.
the different populations sample. Boeker and Goodstein
privitely held firms
of resource
measured
ownership had a positive main effect
outsider, possibly from the firm or organization that
the findings of Boeker and Goodstein (1992).
not
ownership holdings are more
selection. Firms with large outside
on outsider
is
from 1960-1990,
selection,
and the
applicable.
DISCUSSION AND CONCLUSIONS
The general
pattern of results provides strong empirical support for directors' reliance on formal
and informal routines
of
CEO
selection
and
for the
view of corporate boards as normative
arenas. Both the adoption of internal labor markets for
new CEOs and
the past reliance on
informal routines of insider and outsider selection had large, statistically significant effects on
the competing risks of insider versus outsider
easily derived from a
model
CEO
succession The general findings cannot be
of adaptive behavior by
30
board
members which
follow a logic of
Boards as Normative Arenas
consequences
decision making, nor from existing theories of boards as governing
in their
councils, which mobilize resources
support of contending interests. The results are more
in
consistent with the theory that boards are constituted as normative arenas, which follow a logic
of appropriateness
Routines of
and
rule-following,
CEO Selection. At one
as embodied
in
paper applies and extends both old ideas from
level, this
classic theories about organizational routines (March
and more recent quantitative empirical research on
Amburgey and
and
CEO
Miner, 1992) to a
selection.
new area
The paper extends
of persistence of routines
is
precedents, norms, and routines.
and Simon, 1958; Cyert and March, 1963)
their effects (Kelly
and Amburgey, 1991;
of application, the study of corporate
this line of
research by
explicitly testing
governance
how
the effects
mediated by economic performance and the structure of
ownership, and by distinguishing between formal and informal routines. The results do not
support the early March-Simon-Cyert formulation of failure-induced change
routines, but are
more consistent
with the view that responses to
in
organizational
economic adversity
rely
on
dominant organizational routines (Staw, Sandelands, and Button, 1981; Ocasio, 1995).
Organizations change
in
response
dominant patterns and rules
to
economic
of behavior.
that firms will persist with their strategies
economic performance
agency theory
is
These
and
good or whether
interpretations, the reliance
decreases, with inside ownership by board
The
effects of organizational
our understanding of
how
ways
that are consistent with
results support Milliken
is
poor.
The
results also
and Lant's (1991) view
whether recent
show
that,
contrary to
on organizational routines increases, rather than
members and
of formal
top executives.
and informal routines contributes
organizational inertia increases with firm age. Formalization
an
activity
of insider
may be
versus outsider
CEO
31
If
a routine
is
The
selection suggests that
a necessary condition to guarantee
as firms grow older (Selznick, 1957).
to
was
sufficient to explain the persistence of routines with age.
dominant routines
internal political support for
institutionalization
it
in
routines, independently of
age on persistence
found to be neither necessary nor
different effects for
adversity, but
at
its
odds with the
interests
Boards as Normative Arenas
of internal groups, as
in
the case of routines of outsider selections,
its
persistence
increase with age. This suggests that the institutionalization of action requires
is
unlikely to
political
support
historical precedent.
as well as
The formal
routines of insider
CEO
selection
were found
to strongly increase insider
succession, and these effects were relatively invariant to the level of economic performance,
firm age,
ownership structure, or whether succession occured under retirement or early
departures.
The formal
commitments
internal labor
to insider succession,
commitments, independent
markets
for
CEOs
imply both
and boards were much more
political
likely
circumstances facing the firm or the
of the
CEO
and normative
than not to honor these
CEO
succession
^
process.
.»
Boards as Normative Arenas. At another
level, this
paper develops a new theoretical
perspective on corporate governance, that of boards of directors as normative arenas.
The
theory integrates the work of March and Olsen (1989) on the institutionalization of action and
the logic of approphateness
of discipline
variants of
in
in political institutions
social structures.
one form
reinterpret existing theories of corporate
in
U.S. industrial terms
is
interest.
of corporate activities to protect
This normative accounting for board decision making
The theory
is
boards as normative arenas implies
that
at
odds with
board decisions
one
or
strongly supported by
also explains previous descnptions of
board behavior by participants and observers, which were
of
argue that
normative arenas. Board decision process constitute a form of peer review
the empirical results of this paper.
The view
I
better understood as following another form of
and evaluation, rather than an instrumental monitoring
another
governance as
of discipline identified by White, that of governing council.
corporate governance
discipline, that of
I
with that of White (1992) on alternative forms
norms
of
existing theohes.
reflect socially
constructed rules of behavior, which cannot be reduced to either the interests of the participants
nor to adaptively, rational behavior driven by a calculus of consequences. As described above,
32
Boards as Normative Arenas
CEO
the effects of performance and insider ownership on persistence of
suggest that agency interpretations and bounded
norms
selection
by themselves,
rationality cannot, at least
explain the reliance on and persistence of routines of corporate governance. While there
some evidence
is
that routines of insider selection are sustained by managerial interests, the
general pattern of results also suggests that managerial interests alone cannot explain the
routines of corporate governance. Note that there
evidence both
is
for the
existence of a
generalized norm of outsider succession under adversity, and for increased reliance on routines
of outsider succession for
CEOs
departing prior to retirement age. Neither effect
with a purely political explanation of
CEO
additional analysis, not reported here,
selection routines. Furthermore,
I
is
consistent
also conducted
of the mediating effects of the political structure of the
board as measured by a separate chairman, the proportion of outside directors, and board
on routine persistence.
I
found that none of these measures of managerial control over the
board of directors had a noticeable nor a
formal or informal
CEO
size,
statistically significant effect
on the
utilization of either
selection routines. This strongly suggests that the operation of
corporate governance routines cannot be reduced to an explanation based on the view of
boards as "pawns" of management (Lorsch, 1989).
The theory
of
boards as normative arenas highlights the commitments and
boards of directors with the mission, purpose, and social
(Selznick, 1957). Legal responsibilities of
reflected
in
norms and
does not imply
apolitical
and
same
identification of the corporation
and normative commitments by board members are
board members abiding by the
routine. At the
identification of
institutional
embodiment
of
purpose
in
time, this routine-bound character of board decision
that routines are always consistently applied nor that the process
conflict free.
The
effects of routines
the view that routines matter, and matter greatly
their application
is
inside ownership,
far
the firm's
in
in
is
making
purely
the transition rate models strongly support
board decisions, but they also show that
from inevitable The different moderating effects of age, performance,
and early
CEO
departures for insider and outsider selection routines, imply
33
Boards as Normative Arenas
that
norms and
politics,
game
corporate
in
both as vehicles for and constraints on managerial power. The reliance on norms and
routines of
in
precedents serve to establish the rules of the
historical
CEO
selection
and corporate governance also suggests
that political
maneuverings
organizations are unlikely to be directly expressed nor mobilized within the board of directors.
This
is
consistent with a view of managerial power and
power struggles rather than
fixed political coalitions
managerial entrenchment, and the
political
dynamics driven by emergent
and where overt
institutionalization of executive
effective (Ocasio, 1994). Political struggles
efforts at social influence,
power are not
and maneuverings occur not as
likely to
be
part of the formal
board process, but are driven by inside participants and hidden behind the bureaucratic ethos
and
rhetoric that
The
routinization of the
predomina*3s
CEO
in
U.S. corporations
'^Jackall,
^
1988).
selection process by boards as normative arenas
is
closely tied to
the routinization of charisma and depersonalization of executive power. This process embodies
and legitimizes the authority
and control by
of the corporation in
specific individuals
(Weber
bureaucracy, boards as normative arenas
procedures
to define appropriate
its
rules
and procedures rather than
1922], 1978). But unlike the pure form of
rely
on
historical
power
in
Weberian
precedents as well as formal
behavior and to guide their decisions.
Normative arenas and governing councils can be considered as ideal types, rather than as
exclusive descriptions of
understood as different
all
forms of board behavior. The two forms of discipline can be
identities that
external contingencies (White, 1992).
are occasions for a
number
boards of directors take when facing different internal or
March and Olsen (1976) argued
of different things including; (1) fulfilling
executing standard operating procedures;
(2) defining truth
and
that choice
processes
role-commitments and
virtue; (3) exercising
relationships of trust, antagonisms, power, or status, and, (4) expressing self-interests
group
interests.
The
governing councils.
first
In
two conditions are characteristic of normative arenas, the
board decision making,
like in
34
last
and
two
other decision processes, different
of
Boards as Normative Arenas
orientations
may
prevail
depending on the problems and solutions being considered, and on the
structure of participation (Cohen, March,
and Olsen, 1972; March and Olsen, 1976).
Further theoretical development and empirical research are required to explain under what
conditions
is
the model of boards as normative arenas applicable, and under what conditions do
governing councils, or other forms of
discipline,
if
any, apply (White, 1992). White's statement
that "identities are triggered by contingencies" (1992: 5)
that
may
and form
is
evocative of the emergent process
lead certain types of boards, under certain conditions to adopt one or another identity
of discipline. Unfortunately,
no testable propositions can be readily derived from
White's theoretical framework on what these contingencies
likely to
have.
In
Valley,
I
suggested that boards of
many
governing councils than
have
privately held or
is
is
true
in
governing councils
CEO
on
young semiconductor firms
may be more
in
Silicon
operate as
likely to
necessary on whether and how the separation of ownership and control may
to that of
boards of directors from
their traditional role
as
normative arenas. The discrepant findings could also suggest that
the view of boards as governing councils
Inertia
effects they are
general for U.S. industrial corporations. This suggests that
led to a transformation of the functions of
industries, or in
what
effects of proportion of outside directors
with substantial reliance on venture capital,
further research
or
presenting the difference between the findings of this study and those of
Boeker and Goodstein (1992) on the
selection,
may be
may be more
applicable for
new and emerging
cases where resource dependencies on external interests are
and the "Failure" of Corporate Governance.
I
began
this
high.
paper by invoking Jensen's
(1993) presidential address to the American Finance Association, and his lament regarding the
inertia in
corporate governance and the failure of corporate boards to foster adaptation
industrial corporations.
The view
rationale for the observed inertia
empirical evidence on
needs
to
CEO
of boards as normative
in
be reevaluated according
to different criteria.
35
U.S.
arenas provides a coherent theoretical
corporate governance, and one that
selection. But this view also
in
is
supported by
suggests that board effectiveness
Agency
theorists, led by
Jensen, have
Boards as Normative Arenas
evaluated U.S. corporate boards as mechanisms of instrumental control and have found them
to
be
directors
is
that the predominant, albeit latent, function of
that of normative control, rather than instrumental control: to
and accountability
their
paper suggests
deficient. This
in
guarantee
and procedures
of the firm's rules
and
that values accountability
reliability,
Freeman, 1984). The peer review function
in
and normative commitments
ways
that the reliability
terms of
may prove
boards
of
Another larger question remains, however,
changed
in
is,
in
in
with the bureaucratic
model
of organization,
routinization in corporate strategies
directors
have
failed or not,
and whether they need
of course, pretend to
institutions
answer
and functions
to the bureaucratic
this
to
evaluating board effectiveness.
that results
variability of
less desirable?
There
and with the benefits
is
inertia in
corporate
environmental change
increasing dissatisfaction
of standardization
this
and
model. Whether boards of
our evaluation of whether bureaucracies have
in this
paper.
governance
in
I
will
36
will not,
suggest, however, that the current
boards as normative arenas
bureaucracies and to their maintenance and
sustainability.
I
U.S. industnal corporations are closely tied
of organization, as the role of
to the legitimate authority of corporate
from the
Has the world
be supplanted by other organizational forms.
question
of corporate
model
be adaptive (Hannan and
and routines associated with
failed or not, is closely tied to
relatively stable, the
providing stability and enforcement of
governance become less attractive? Are the frequency and
is
world
under such circumstances, highly desirable.
and accountability
increasing such that organizational stability
to
their construction
their appropriateness. In a
and when environments are
persistence of routines of corporate governance
fiduciary duties
reliability
the rules of corporate behavior. Board responsibilities are defined, not by
independent evaluation of the consequences of corporate actions, but by
and reproduction
boards of
is critical
Boards as Normative Arenas
REFERENCES
Aldrich,
1986
Howard,
E.,
and
Ellen Auster
"Even dwarfs started
Barry Staw and L.L.
small: Liabilities of
Cummings
age and size and
(eds), Research
in
their strategic implications." In
Organizational Behavior,
8:
156-198. Greenwich, CT: JAI Press.
Amburgey, Terry
1992
activity.:
S.
Strategic
and Gardiner
Berle, Adolf, Jr.,
1967
and Anne
momentum: The
"Strategic
merger
L.,
Miner
effects of repetitive, positional
Management
C.
and contextual momentum on
Journal, 13: 335-348.
Means
The Modern Corporation and
Private Property.
New
York: Harcourt, Brace, an'^ World.
Boeker, Warren
1992
"Power and managerial dismissal: Scapegoating
at the top." Administrative
Science
Quarterly, 37:400-421.
Boeker, Warren, and Jerry Goodstein
1992
"Performance and successor choice: The moderating effects of governance and
Academy
ownership."
of
Management
Journal. 36: 172-186.
Bowen, William G.
1994
Inside the Boardroom:
Governance by Directors and Trustees. Ne York: John Wiley &
Sons.
Cannella, Albert A., and Michael Lubatkin
1993
"Succession as a sociopolitical process: Internal impediments
Academy
Carroll,
1984
Glenn
of
Management
to outsider selection."
Journal. 37:
R.
'Dynamics
of publisher
succession
in
newspaper organizations." Administrative Science
Quarterly, 29: 93-113.
37
Boards as Normative Arenas
Clark, Robert C.
1991
"Agency costs versus
fiduciary duties," In Pratt,
eds. Principals and Agents:
The Structure
John W. and Richard
J.
of Business: 55-79. Boston:
Zeckhauser,
Harvard
Business School Press.
Cohen, Michael
1972
D.,
James
G. March, and Johan P. Olsen
"A garbage can model of organizational choice." Administrative Science Quarterly, 17:
1-25.
and James G. March
Cyert, Richard M.,
1963
A
Behavioral Theory of the Firm. Englewood
Davis, GerTid
1991
F.
.v*^
Cliffs,
NJ: Prentice-Hall.
5^
-
"Agents without pnnciples? The spread of the poison
pill
through the intercorporate
network." Administrative Science Quarterly, 36: 583-613.
DiMaggio, Paul
1983
The
iron
D.,
and Walter W. Powell
cage
revisited: Institutional
organizational fields,"
Doeringer, Peter
1971
Internal
Fama, Eugene
1983
B.,
Amencan
and Michael
J.
isomorphism and
collective rationality in
Sociological Review. 48: 147-160.
Piore
Labor Markets and Manpower Analysis. Lexington, MA: Heath Lexington Books.
L.,
and Michael
C.
Jensen
"Separation of ownership and control
"
Journal of
Law and Economics,
26: 301-325.
Garfinkel, Harold
1967
Studies
in
ethnomethodology. Englewood
Gibbons, Robert, and Kevin
1990
'"Relative
J.
Cliffs,
NJ: Prentice-Hall
Murphy
performance evaluation
for
Chief Executive Officers." Industrial and Labor
Relations Review, 45: 30S-51S
38
Boards as Normative Arenas
Gouldner, Alvin W.
1954
Patterns of Industrial Bureaucracy.
New
York: Free Press.
Guo, Guang
1993
"Event-history analysis for left-truncated data."
Methodology 1993: 217-243. Cambridge, MA:
Hannan, Michael
1984
In
Peter V. Marsden
(ed.), Sociological
Basil Blackwell.
and John Freeman
T.,
"Structural inertia
and organizational change." Amehcan Sociological Review, 49:
149-164.
1989
Organizational Ecology. Cambridge, MA: Harvard University Press.
Helmich, Donald
1972
and Warren
L.,
B.
Brown
"Successor type and organizational change
Science Quarterly,
17:
in
the corporate enterprise." Administrative
371-381.
Hirsch, Paul
1982
"Network data versus personal accounts: The normative culture of interlocking
directorates."
Graduate School
of Business, University of Chicago,
unpublished
manuscript.
Jensen, Michael C.
1993
"
The modern
industhal revolution,
exit,
and the
failure of internal control
systems."
Journal of Finance, 48: 831-880.
Jackall, Robert
1988
Moral Mazes: The World of Corporate Managers.
New
York: Oxford Univ?,rsity Press.
Jensen, Michael C, and William H. Meckling
1976
"Theory of the
firm:
Managerial behavior, agency costs and ownership structure."
Journal of Financial Economics,
3:
305-360.
39
Boards as Normative Arenas
Kelly,
1991
Dawn, and Terry
L.
Amburgey
momentum; A dynamic model
"Organizational inertia and
of
Management
of strategic change."
Academy
Journal, 34: 591-612.
Lorsch, Jay W.
1989
Pawns
Mace, Myron
1971
or Potentates. Boston: Harvard Business School Press.
L.
Myth and
Directors:
Reality. Boston:
Harvard Graduate School of Business
Administration.
March, James G.
1994
A
Primer on Decision Making:
March, James
G.,
and Johan
1976
Ambiguity and Choice
1989
Rediscovering
March, James
1958
G.,
in
and Herbert
New
Organizations.
Decisions Happen.
New
York: Free Press.
^"
Olsen
Organizations. Bergen, Norway: Universitesforlaget.
Institutions:
The Organizational Basis
A.
of Politics.
New
York: Free Press.
Simon
York: Wiley.
Meyer, John W., and Brian
1977
P.
How
Rowan
"Institutionalized organizations:
Formal structure as myth and ceremony." American
Journal of Sociology, 83: 340-363.
Millii^en,
1991
Frances
"'The
J.
and Theresa
K.
Lant
impact of an organization's recent performance history on strategic persistence
and change."
Strategic
in
Paul Shrivastava,
Management,
7
Anne
Huff,
and Jane Dutton
(eds.).
Advances
129-156. Greenwich, CT: JAI Press.
Mizruchi, Mark
1982
The American Corporate Network, 2904-2974 Beverly
40
Hills,
CA; Sage.
in
Boards as Normative Arenas
Ocasio, William
1994
dynamics and the
"Political
circulation of power:
CEO
Succession
in
U.S. Industrial
Corporations, 1960-1990." Adnninistrative Science Quarterly, 39: 285-312.
1995
A
"The enactment of economic adversity:
change and
threat-hgidity." In L.L.
Organizational Behavior, 17:
(In
reconciliation of theories of failure-induced
Cummings and
Barry M. Staw (eds.), Research
in
Press). Greenwich, CT: JAI Press.
Palmer, Donald A.
1983
"Broken
ties:
Interlocking directorates
and intercorporate coordination." Administrative
Science Quarterly, 28: 40-55.
Pfeffer, Jeffrey,
1978
The
and Gerald
External Control of Organizations.
Puffer, Sheila M.,
1991
R. Salancik
and Joseph
B.
New
York: Harper and
Row.
Weintrop
"Corporate performance and
CEO
turnover:
The
role of
performance expectations."
Administrative Science Quarterly, 36: 1-19.
Rohwer, Gotz
1991
Analysis of transition data:
European University
A
practical introduction with
TDA. Unpublished manuscript,
Institute.
Selznick, Philip
1957
Leadership
in
Administration. Evanston,
IL:
Row, Peterson.
Staw, Barry M., Lloyd E. Sandelands, and Jane E. Dutton
1981
"Threat-rigidity effects in organizational behavior:
A
multilevel analysis." Administrative
Science Quarterly, 26: 501-52''.
Tosi,
1989
Henry
L.,
and Luis Gomez-Mejia
"The decoupling of
CEO
pay and performance: An agency theory perspective."
Administrative Science Quarterly, 34: 169-189
41
Boards as Normative Arenas
Tuma, Nancy
1984
and Michael
B.,
Hannan
Social Dynamics: Models and Methods. Orlando: FL:
Tushman, Michael
1985
T.
L.,
and Elaine Romanelli
"Organizational evolution:
In L.L.
Academic Press.
Cummings and
A metamorphosis model
of
convergence and reohentation."
Barry M, Staw (eds), Research
in
Organizational Behavior.
7:
171-222. Greenwich, CT: JAI Press.
Useem, Michael
1984
The Inner
Circle.
New
York: Oxford University Press.
Vancil, Richard M.
1987
Passing the Baton: Managing the Process of
CEO
Succession. Cambridge, MA:
Harvard Business School Press.
Wade, James, Charles
1990
A. O'Reilly,
"Golden parachutes:
and
CEOs and
Ike
Chandratat
the exercise of social influence." Administrative Science
Quarterly, 35: 587-603.
Walsh, James
1990
"On the
P.,
and James
K.
Seward
efficiency of internal
Management
and external corporate control mechanisms." Academy of
Review., 15: 421-458.
Weber, Max
1978
Economy and
Society Guenther Roth and Claus Wittich (eds.). Berkeley: University of
California Press.
White, Harrison 0.
1992
Identity
and Control: A
Structural
Theory
of Social Action. Princeton, N.J.: Princeton
University Press.
42
Boards as Normative Arenas
Zajac,
Edward
1990
"CEO
J.
selection, compensation,
empirical analysis." Strategic
and
firm
performance:
Management
A
theoretical interpretation
and
Journal, 11: 217-230.
Zucker, Lynn G.
1977
"The
role of institutionalization in cultural persistence."
42: 726-743.
43
American Sociological Review,
in
,-„u.
^r-^(0^tD^tD^
C
o
c
o
Q.
-a
-„iD04o2^^'-^(DOO^OJCMt-^V,01
ooo^o^o^oootoco^cn^ogo
^
o n
o o^ o
»- TOO
to f^
_
T-
_
r-
-^ CM -^
o
h-
SinQU-)r--^h-^cD'-'ooooo*^OT^odoiricMQmoooo
00 in
O
00
Table 3
Likelihood Estimates of the Competing Risks of Insider vs. Outsider CEO Selection,
Piec ewise Exponential Models Comparison of Full S a mple with Early CEO Departures*
Model 6a
Model 6
Model 2a
Model 2
Maximum
Variables
to 2 years
Insider
-7.708
-6.168""
(0.756)
7 to
-5.782""
2 years
1
(.730)
-5.590""
12 to 17 years
(.728)
1
-5.484*"*
7 years or over
(.608)
ROA
(.011)
.086**"
Age60
*
ROA
.079"
Age
Prop Outside
Directors
Inside
(.008)
(.012)
(.009)
.048
.150*
.130
In
&
O
(.086)
(.149)
.009
-4.960"
(.028)
In
&
(.035)
(031)
-.016
-.017
.141"
-.040
(.096)
(.084)
.057****
(.122)
(.017)
.602**
(.028)
.040"
(.020)
.539"
.621*
.014
(.014)
-.262"
.029
.657
(.228)
(.271)
(.425)
(.232)
(.362)
(.271)
(.432)
-0.642
-.993
-0.293
-.616
-0.745
-1.425
-0.505
-.980
(.736)
(1.135)
(.536)
(.901)
(.740)
(1.115)
-.012
.017
-.003
.035"
(.010)
(Oil)
(.014)
-.004
(.014)
(.012)
.863***
(Oil)
(.904)
.009
-.003
.029*"
(.009)
.907**"
.424
(.010)
(.006)
(.008)
(.009)
.003
.033*"
-.006
(Oil)
(012)
.906*"
(.243)
.749**
(.322)
(.304)
(.534)
.252
1.480***
(.403)
(.535)
.651*
1.226"
(.382)
(.618)
.375
.034****
.003
(.009)
(.009)
.837**"
.307
269
(.381)
(.244)
(.323)
(.296)
530
1.355
-.043
1.077
-.280
2.366"
(.471)
(.860)
(.454)
(.781)
1.955****
(.620)
(1.010)
(.584)
(.605)
2.388*"
.393
(.547)
(861)
1.885"
(.513)
.536
2.500*"
.315
(.947)
(.417)
(.788)
.068****
(.756)
(.021)
.101****
(.047)
(.031)
(.027)
.018
.027*
.022
(.016)
(.014)
(.020)
-.003
(.588)
(.009)
(Oil)
.039**
In .&
(.015)
.05;
in
***
-473.148
-736.628
38
85
48
159
57
p<.0001.
38
parenthesis.
46
.461
(.914)
3.397*"
(1.195)
.076""
.053
(.023)
.100*"
.024
-.067*"
.007
Events
2.746"*
.335
.021"
38
159
p< .01;
.156
(.295)
.016
-756.809
.037**"
(.389)
Out
Inside ownership
10;
-.019
(.013)
(.010)
.450
-.027*
(Oil)
-.216"
.028*"
Out.
Standard errors
(.017)
(.357)
Likelihood Ratio
'
(1.768)
-.089****
(.030)
.027
**p<
-.035"
(1.971)
-7.456****
(.015)
.092***
Inside ownership
of
(1.147)
-.085*"*
(.024)
Combined
-6.985**"
-6.690
-7.416**"
.153
Past Insiders
* Firm age
Past Insiders
-p<
(1.239)
(1.976)
-6.735""
(.152)
ROA
Df
Number
(1.367)
(.044)
Combined
*
(1.244)
(2.065)
-7.483****
(.087)
*ROA
*
(1.398)
-6.571*
-6.862"**
.146*
Past Outsiders
*
(1.302)
-7.994"**
(.015)
.037
.629"
(1.459)
-6.839*
(2.206)
-7.575""
.042
(.062)
(.017)
(1.368)
(1.192)
-10.204*
(.015)
(.016)
.097****
(.008)
(.020)
(1.626)
-7.352*
-.035"
Out sider
Insider
-8.350'
-9.600*
-4.818"
(.009)
-.199*
(010)
.579
-5.124"
-.015*
-.006
(.006)
-5.527"
-.023**'
(.106)
(.005)
Combined
(.660)
(1.766)
-7.537"
(.081)
.053***
Outside Ownershi -.001
Past Outsiders
(.766)
(1.091)
-.021*
-7.144"
(083)
.040"
Ownership -.003
Past Insiders
(1.932)
-7.379*
(1.165)
-7.297"
(.059)
.028***
(.528)
Formal ILM
(.773)
(1.184)
,58*
.563"
Chairman
(1.951)
-7.172*
(1.323)
-6.289****
.112*
of Hire
(0.798)
(1.204)
(007)
(.004)
Year
(2.038)
-7.142*
(1.356)
-6.064****
-.014"
.000
Size
(.862)
(1.257)
Outsider
Insider
-6.450"
(2.157)
-7.598*
(.024)
(.034)
Firm
-7.939"
(1.409)
-6.246****
(.029)
-.040*
(.015)
Age64
(1.327)
-6.683**"
-.024*"
-.023"
Outsider
-9.260*
Insider
-8.663"
(1.544)
(.823)
2 to 7 years
Outsider
-8.439'
(.035)
-.108*"
.010
(.024)
(.016)
(.035)
-.018
.035*
-.026
(.825)
(.020)
(.027)
-454.120
57
48
85
38
in
o
o
CO
O
LU
O
o
"D
CO
CO
i_
g
jo
CO
>
g
(D
lO
o
O
CO
c
g
o
CO
CO
^°
CO
Q.
CO
E
o
O
E
a:
(D
ZJ
Ll
o
—
DC
o
o
(D
E
i_
o
LL
^
CD
C
o
CO
CO
o
LU
O
3
CD
g
O
'co
*a)
o6
g
-c
CO
CO
Q.
g
'co
c
"5
O
g
CO
CO
CD
_c
g
D
CO
O
Q_
to
E
o
"co
CO
CO
Q_
_c
"co
CO
"cc
CO
C
O
CO
CL
47
C
O
o
m
c
g
(/)
(/)
<
0)
<]
o
o
u
I
Q
CO
CO
O
LU
O
g
o
"(/)
o
^
C/)
C/)
0)
o
o
g
C/)
(/)
0)
o
o
13
o
O
g
LJJ
L_
'(/)
c
o
Mil
(/)
llflRARieS
o
00
g
CO
0)
(/)
c
$
0)
o
o
O
g
CO
o
LU
O
g
o
CD
0)
•U
C
~
i_
CD
CO
CO
c
>^
X
X
X
X
X
X
X
'^
CO
o
g
X
X
o
J^
X
X
X
g
c
O
CD
CO
(/)
CD
O
o
a.
CO
o
CM
o
LU
O
g
i_
CD
CO
—
C/)
o o
l^
.
.
CO
03
Lo q:
CD
^—
=3
.2)
LL
o
r—
i3
C
O
2
176
:^iO
51
Date Due
OEC
3
!
1119
^1i4l!9S
Lib-26-67