. r-- »'
a V 2"” 1C
.‘ l

“mu 4-“?

0

4.

“wk

P1

‘Wfi

aw”;

want-J
h o n

r.“

,....__‘

5w.

'

. . 'l‘.’
u 4 ' I
v (Ll .. x ,p
= 'S‘M‘I'
1L ,,- a
’.'. ‘41.“?

r"

y-l’

Ayn—g"

.... \.
,,..}__

13:3“:1‘. .

vi"
1:"
~...

_ f.
$35“

.

‘t
.

g .
:ji-
41

Jr

 

THESIS

                                                                             

llhlllllllllllllllll

‘1 ‘ 1 . 31293 01564 9894

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

THE IMPACT OF AN ORGANIZATIONAL DOWNSIZING
EFFORT ON SURVIVORS' COMMUNICATION NETWORK
RELATIONSHIPS AND ATTITUDES

1

presented by

Alex Michael Susskind

has been accepted towards fulfillment
of the requirements for

Ph.D. degreein Commun‘i‘cation‘

 

 

t/W MIL

 

 

Major professor

7130/7é

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

PLACE Iii RETURN BOX to remove thie checkout from your record.
To AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

1:3".“5'5 “‘3 fl '
a.‘ 7’ -‘ v..-»

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Ie An Nflnnetive Action/Ewe! Opportunity lnetituion
Wanner

 

THE IMPACT OF AN ORGANIZATIONAL DOWNSIZING EFFORT ON
SURVIVORS’ COMMUNICATION NETWORK RELATIONSHIPS AND
ATI'ITUDES

By

Alex Michael Susskind

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Communication

I 996

ABSTRACT

THE IMPACT OF AN ORGANIZATIONAL DOWNSIZING EFFORT ON
SURVIVORS’ COMMUNICATION NETWORK RELATIONSHIPS AND
ATTITUDES
By

Alex Michael Susskind

A possible reason for downsizing’s meager success in the workplace is related to
the lack of attention paid to layoff survivors’ communication patterns and relationships.
This investigation examined changes among communication network relationships and
associated perceptions and attitudes in a hotel company’s corporate headquarters to
provide additional insight into layoff survivors’ responses to downsizing. Through the
use of communication network measurement techniques and survey measurements at
three points in time, this investigation examined the impact of structural holes on layoff
survivors’ perceptions of information deficiency, stress, job satisfaction, and
organizational commitment both prior to and following the downsizing effort. As a result
of downsizing, employees gained or lost a variety of network contacts. Layoff survivors
who experienced increases in structural holes perceived higher levels information
deficiency as newly formed gaps in their communication network restricted the flow of
information and resources. Increases in information deficiency were related to increases
in survivors’ levels of stress, which negatively impacted their affect toward their job in

the measurement period directly following the downsizing. Static and longitudinal

models of survivors’ structural hole experiences and attitudinal responses were presented
and tested. Results indicated modest support for the static models and revealed
differences among subgroups within the sample, while the longitudinal model was not

supported.

Copyright by
ALEX MICHAEL SUSSKIND
1996

Dedicated to Eric Lieberman and Bert Landauer in loving memory

ACKNOWLEDGMENTS

There are many people who have assisted me in my journey toward this doctoral
degree. Each acquaintance close or distant has had an impact on this process. I wish to
thank each and every one of them for their guidance, support, patience, insight, and
tolerance. Specifically, I would like to offer recognition to the following individuals for
their contributions to this process.

I would like to thank my parents Herbert and E. Suzanne Susskind for their
unconditional love, encouragement and support which provided me with the needed
motivation to pursue my ever-fluctuating professional interests over the years. I would also
like to thank my grandmother Ms. Lore Lieberman for her faith and belief in my true
potential and her continual support of my educational endeavors.

Much gratitude is extended to Dr. Bonnie F arber for introducing me to true
scholarship. As my mentor during my undergraduate years at Purdue University and early
graduate years at Michigan State University, Dr. F arber encouraged me to pursue the
interests which got me where I am today.

I would also like to thank The School of Hospitality Business and The Department
of Communication, its faculty, staff and students for providing me with technical, financial

and moral support during my entire graduate education.

vi

I would like to thank Dr. Carl P. Borchgrevink for his faith in my abilities as both a
teacher and researcher. His patience, understanding, and advice were crucial to my
educational experience (thank you for steering me away from the Sociology Department).

I wish to thank the members of my doctoral dissertation committee (alphabetically),
Dr. Franklin J. Boster, Dr. Georgia T. Chao, Dr. Ronald P. Cichy, Dr. J. David Johnson
and Dr. Vernon D. Miller. Each member brought unique contributions to the degree
earning process, and helped guide me through its completion. Dr. Boster selflessly shared
’ his methodological and statistical expertise, and reminded me that the acquisition of
knowledge is an imperfect, progressive process. Dr. Chao willingly shared her knowledge
of the field of management, psychology and downsizing in particular, and brought a fresh
insight into my work. Dr Cichy provided knowledge and guidance in the area of hospitality
management, and provided me with sound advice on how to survive this process, in
- addition to his role as my employer. Dr. Johnson shared his vast knowledge of network
analysis and related organizational communication issues with me and is primarily
responsible for helping me convert this dissertation topic into its current form.

Dr. Vernon D. Miller deserves particular mention for his role as chair of my
guidance committee (and as such will have his own paragraph). Dr. Miller provided me
with valuable teaching, research and consulting opportunities and acted as my mentor
through my entire doctoral program. Dr. Miller’s friendship, patience, tolerance,
understanding, and time invested in my education were crucial to its completion. I am glad
to have had the opportunity to study with Dr. Miller and as a result feel that I am well

prepared to embark upon my academic career.

vii

I would also like to thank Van Morrison and the late Miles Davis for providing me
with soothing and motivating tunes during times of stress, duress, and uncertainty. A special
thanks to Mr. Big Monkey for being BIG and being a monkey.

Finally, I would like to thank Annemarie Hodges for her love, friendship and
support during this process. Through my relationship with Ms. Hodges I learned much

about life, living and the pursuit of inner peace; for that, I am etemally grateful to her.

viii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................... xi
LIST OF FIGURES ........................................................................................................... .xii
CHAPTER 1
Introduction ........................................................................................................................ 1
Organizational Change and DownstzrngS
The Nature of Organizational Change ................................................................... .5
Downsizing ............................................................................................................. .6
Organization-Environment ........................................................................ .8
Dismissed Employees ................................................................................ 10
Layoff Survivors ...................................................................................... .11
Communication Networks ..................................................................................... 14
Structural Holes ...................................................................................... .15
Structural Hole Measurement ............................................................. 18
Research Question 1 ..................................................................... 21
Research Question 2 ..................................................................... 21
Perceptions and Attitudes ofDownsizing Survrvorle
Information Deficiency .................................................................................... .23
Stress ................................................................................................................. 23
Job Satisfaction and Organizational Commitment ............................................ 25
CHAPTER 2
Methods ..................................................................................................................... .30
Procedure ....................................................................................................... .30
Participants ...................................................................................................... 30
Nature of Downsizing .................................................................. 31
Structural Hole Measurement ......................................................... 31
Network Relationships ........................................................................ 31
The Structural Hole Index .................................................................. .32
Summated Measures ................................................................... 33
Statistical Analyses ........................................................................................ .37
CHAPTER 3
Results ....................................................................................................................... 38

Tests ofthe Research Questions... 38
Tests ofthe Static Model45

ix

The Hypothesized Model” ... 46
Tests ofthe Revised Static ModelSO

Tests of the Longitudinal Model ........................................................................ 56
CHAPTER 4
Discussion... .. . . .. ......................................................................................................... .60
Overview of the Study ................................................................... 6O
Findings and Implications ............................................................... 62
Limitations ....................................................................................................... .70
Future Research ............................................................................................... .74
References .......................................................................................... 78

LIST OF TABLES

Table 1 ...................................................................................................................... 20
Table 2 ...................................................................................................................... 34
Table 3 ...................................................................................................................... 39
Table 4 ...................................................................................................................... 40
Table 5 ...................................................................................................................... 42
Table 6 ...................................................................................................................... 46
Table 7 ...................................................................................................................... 47

xi

LIST OF FIGURES

Figure 1 .................................................................................................................... .4
Figure 2 .................................................................................................................... 28
Figure 3 .................................................................................................................... 48
Figure 4 .................................................................................................................... 51
Figure 5 .................................................................................................................... 54
Figure 6 .................................................................................................................... 57
Figure 7 .................................................................................................................... 58

xii

Chapter 1
Introduction

Organizational downsizing is a disruptive process which impacts the
communication patterns, perceptions and attitudes of surviving employees. Organizations
downsize their workforce in keeping with market and financial shifts in an attempt to
remain competitive (Cascio, 1993). In 1993 a record number of job cuts were announced.
This trend continued through 1994 when approximately 460,000 workers were laid off from
major US. companies, and it is expected to persist (Gottlieb & Conkling, 1995; Uchitelle &
Kleinfield, 1996).

Layoff survivors face many changes in their work environment, including: (a) loss
of ties to information sources, (b) loss of direct and indirect links to individuals with power
and influence, and (c) a loss of connection to and social interaction with dismissed
employees. These losses create voids in an organization’s communication network known
as structural holes. According to Burt (1992a, p. 2), structural holes create “entrepreneurial
opportunities for information access, timing, referrals and control.” On one hand, structural
holes force employees to seek new linkages which can provide unique sources of and/or
new opportunities for the brokerage of information. On the other hand, structural holes
create disruptions when members lose valued network connections. As a result of structural
holes, layoff survivors experience stress and information deficiency which ultimately
impact employee commitment and job satisfaction.

A number of investigations examine the effects of organizational downsizing on

surviving employees (see reviews by Brockner, 1988; Kozlowski, Chao, Smith, & Hedlund,

1993). Research by Brockner and colleagues (Brockner, Wiesenfeld, Reed, Grover, &
Martin, 1993; Davy, Kinicki, & Scheck, 1991; Brockner, 1988) indicates that layoff
survivors’ reactions are mediated by a number of factors, including coworkers’ perception
of justice for laid-off employees and their prior organizational commitment. While these
studies provide valuable insight into the effects of downsizing on individuals’ attitudes, few
existing studies examine the connection between changes in communication network
structure and the behavioral and attitudinal effects on surviving employees over time. A
lack of attention to downsizing’s long term effects on organizational network configurations
is particularly surprising since: (a) observable physical reconfigurations are likely to have
considerable impact on work processes and employee attitudes; (b) a considerable number
of often cited downsizing studies are laboratory versus field based (e.g., Brockner, Davy, &
Carter, 1985; Brockner, Grover, Reed, Dewitt, & O'Malley, 1987); (c) it is unclear how
organizational members cope with lost linkages and the need to form new linkages; (d)
resulting role changes may have a considerable impact on power and influence structures;
(e) longitudinal studies of task related networks in organizations appear essential to
understanding organizational efforts at efficiency and development (Kozlowski et a1.,
1993); and (f) few, if any, investigations examine the influence of organizational
downsizing on the emergence of the ever-fluctuating configuration of employee
communication networks.

Just as survivors’ reaction to layoffs are impacted by numerous factors, this study
proposes that changes in employees’ communication networks as a result of downsizing

mediate the relationship between the downsizing event and layoff survivors’ attitudes and

perceptions. As illustrated in Figure l, structural holes in a commrmication network
influence organizational members’ perceptions of information deficiency which then
impacts their experience of stress. Stress subsequently impacts their degree of job
satisfaction which in turn affects their commitment to the organization. Structural holes
represent a restriction of access to information contacts, and create disruptions in the flow of
communication and work processes. In essence, this study proposes that downsizing does
not affect surviving employees in the same way. Instead, a downsizing’s impact on
employee attitudes and perceptions depends on their relationship to structural holes.
Employees who experience structural holes in their networks should report greater changes
in their attitudes and perceptions than employees not experiencing structural holes due, in
part, to the perceived loss of established coworker support and work facilitation information
(Ray & Miller, 1991).

While Burt (1992a, 1992b) primarily focuses on economic and sociological
applications of structural holes and their use to bring advantage to network members and
organizations, this investigation extends structural hole research into the domain of
organizational downsizing and its impact on individual behavior, creating a link between
macro— and micro-level elements. To this end, I overview the relationship between
organizational change, downsizing, and communication networks. In particular, I describe
the relationships between structural holes and organizational networks and how structural
holes impact on employees' perceptions and attitudes. Chapter two reports the methods used
to measure structural holes and employee attitudes and perceptions in a hotel organization’s

corporate office during an organizational downsizing. In the third and fourth chapters,

E52 Ems—“U 23m uofimofionzm ._ 83E

 

 

 

 

 

 

 

EuEEEoU cozoamcam 3:29:00 5.2: 2o:
Ecoceficmwco pen I 3.25 couanuE Beau—Em

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

respectively, I present the results of the study and a discussion of their implications and
directions for future research.
Organizational Change and Downsizing

Employees’ responses to organizational change reflect the degree of acceptance of
planned organizational change (Lewis & Seibold, 1996; Miller, Johnson & Grau, 1994) and
contribute to its successful implementation and acceptance in organizations (Farmer, 1990).
The study of organizational change is central to an understanding of evolving organizational
structures and employee behaviors and attitudes (Berger & Cummings, 1979; Goodman,
Bazerman & Conlon, 1980; Levy & Merry, 1986; Porras & Robertson, 1992). To this end,
investigations examine employee responses to organizational development (e. g., Beer,
1980; Cummings & Huse, 1989), organizational transformation (e.g., Levy & Merry, 1986;
Tushman & Romanelli, 1985), and organizational diffusion of innovation (e.g., Kanter,
1988; Lewis & Seibold, 1993, 1996; Rogers, 1983).
The Ngure of Organizational Change

In general, researchers classify organizational change as planned or unplanned and
as a first-order or second-order event (Levy, 1986; Levy & Merry, 1986; Porras &
Robertson, 1992). Planned change is a deliberate, progressive decision made to improve
organizational functioning (Levy & Merry, 1986; Porras & Robertson, 1992). In contrast,
unplanned change is adaptive and evolutionary and is the result of the organization’s
reaction to extra-organizational or environmental stimuli (Porras & Robertson, 1992;
Tushman & Romanelli, 1985). Whether change is planned or unplanned, it may result in

first-order or second—order changes. F irst-order changes are minor adjustments which do not

alter the basic “core” of the organization and occur as the system grows and develops (von
Bertalanffy, 1968). Second order changes are more drastic modifications to the
organization’s mission, structure, or processes and may be accompanied by potentially
disruptive jolts to a relatively stable set of internal organizational configurations (Levy,
1986; Levy & Merry, 1986).

According to Levy and Merry (1986), second order change alters two or more of the
following organizational components: (a) the organization’s paradigms - the underlying
philosophy or “metarules” (p. 276) which shape organizational behavior; (b) its mission and
purpose - the explicit goals and policies governing organizational actions; (c) its culture - an
organization’s beliefs, values, and norms; and/or (d) functional processes - which include
organizational structure, communication, and decision making procedures (Levy & Merry,
1986). To date, research investigating second-order change examines the quality of work
life (Crunmings & Huse, 1989), employee participation in decision making (Beer &
Walton, 1987), and organizational restructuring in the form of mergers, acquisitions, and
downsizing (Kozlowski, et al., 1993). As such, second order change represents a
fundamental change to employee orientation and function and parallels the descriptions of
many organizational downsizing events.

Downsizing

Organizational downsizing is defined as a purposive action (planned) initiated to
reduce the size of the workforce in an attempt to increase the efficiency or effectiveness of
the organization (Freeman & Cameron, 1993; Kozlowski, et al., 1993). Downsizing can

also be a planned, second-order change when the reduction in the number of employees

forces a readjustment of work roles and relationships. Regardless of the proportion of
employees “let go,” significant adjustments in work behavior are usually required on behalf
of the remaining or “surviving” organizational members.

For the most part, downsizing efforts are implemented after a noticeable
organizational decline (Cameron, Kim & Whetten, 1987; Sutton & D'Aunno, 1989,
Whetten, 1980) in an effort to increase profitability and firm performance (McKinley,
Sanchez & Schick, 1995). However, organizational downsizing does not necessarily result
from attempts to correct decreases in firm performance (Cameron, Freeman & Mishra,
1991; Freeman & Cameron, 1993; Whetten, Keiser, & Urban, 1995). For instance, firms
downsize in an attempt to implement less hierarchical and more versatile organizational
structures (Cameron, et al., 1991; Cascio, 1993; Whetten & Cameron, 1994; Whetten, et al.,
1995)

In investigations of organizational downsizing events, researchers have considered
the organization-environment conditions precipitating a downsizing (Cameron, Kim, &
Whetten, 1987; Greenhalgh, et al., 1987), the strategies and procedures used to accomplish
downsizing (Bruton, Keels & Shook, 1996; Cameron, in press-a, in press-b; Dewitt 1993;
Whetten & Cameron, 1994; Whetten, et al., 1995), the impact of dismissal on former
employees (Buss & Redbum, 1987; Cobb & Kasl, 1977; Podgursky & Swaim, 1987; Warr
& Jackson, 1984), and the reactions of the surviving personnel (Brockner, 1988; Brockner,
Davy, & Carter, 1985; Brockner & Greenberg, 1990; Brockner, Grover, Reed, DeWitt &

O'Malley, 1987; Gutknecht & Keys, 1993; Larkey, 1993; McKinley, et al., 1995).

W Organization-environment investigations (Dewitt, 1993;
Tushman & Romanelli, 1985; Zammuto & Cameron, 1985) examine downsizing as a
function of either: (a) an internal reorganization that requires the shifting or adjustment of
human resources based on perceived financial or strategic needs (Whetten, et al., 1995); or
(b) a merger or acquisition to regain solvency through the divestiture of their assets or
through the acquisition of additional assets believed to be profitable (Bntton, et al., 1996;
Gutknecht & Keys, 1993). A common outcome of both internal reorganization and mergers
and acquisitions is a reduction in workforce (Gutknecht & Keys, 1993). An internal
reorganization normally reduces personnel as processes and positions are reconfigured.
Mergers eliminate personnel who have become redundant and unneeded in the merged
environment.

Dewitt (1993) highlights the importance of using a “multiple-contingency
downsizing strategy model” when examining organization-environment relationships. Key
elements in the model are the organization’s: (a) domain - needed changes to products,
services, or markets to enhance performance; (b) structure - needed changes in
decentralization or formalization; (c) reorientation strategy - simplifying or reducing
organizational activities by dropping a product line or service; and (d) retrenchment
strategy - shifting business focus by developing a new or revised product line. By reporting
changes to these elements, researchers are able to provide a more precise description of the
nature of downsizing and develop a greater understanding of effective strategies under

specific external environmental conditions.

Adapting the Structural Equilibritun Model (T ushman and Romanelli, 1985),
Freeman and Cameron (1993) propose that the successful implementation of downsizing be
examined as either a convergence or a reorientation transformation. Convergence is
characterized by gradual, progressive organizational changes such as redirecting an existing
company mission or introducing minor modifications to products and services. In contrast.
reorientation is achieved by changes to the management strategy, structure. and basic
organizational practices. Both approaches examine the relationship between downsizing.
the organizational structure and coordination processes. Yet, convergence downsizing
transformations center on the maintenance of organizational stability while a reorientation
stresses the shift of internal processes in order to adapt to a new external environment.

As downsizing holds popularity as a strategy for improving organizational
competitiveness (Gottlieb & Conkling, 1995; Uchitelle & Kleinfield, 1996), several
scholars provide critical appraisal of downsizing proponents’ claims of improved
organizational performance. Whetten and Cameron (1994) review several studies of
downsized organizations. They suggest that the “paradoxical” lack of organizational
improvement in financial performance (in terms of overhead and salary) and/or productivity
are due to organizations indiscriminately reducing the workforce rather than targeting
specific units for improvement. Based on skepticism regarding downsizing’s effectiveness
as a cost-cutting strategy, in a review of 210 organizations announcing layoffs, McKinley,
et a1. (1995) report downsizing being promoted, rationalized, and institutionalized through
social pressures in three ways: (a) constraining - the acceptance of a vision regarding how

their business image and actions should appear to others (e.g., lean and mean as opposed to

10

bureaucratic); (b) cloning - where companies jump on the “downsizing band-wagon”
because they are uncertain about the pursuit of other cost-cutting options and follow the
actions of leading firms in their industry; and (c) leaming - where cost—cutting behavior is
instilled and rewarded in organizational members. Whetten, et a1. (1995) argue that human
resource units in organizations must become more strategic in their activities in order to
compensate for the increase in personnel-related demands associated with a reduced and
restructured workforce. Others suggest that the magnitude of the downsizing and the
operating characteristics of the firm should be carefully considered prior to developing and
implementing downsizing strategies (Bruton, etal., 1996; Cameron, in press-a, in press-b ).
Dismissed Employees. The termination of employees through downsizing is often
accomplished in one of three ways (Fisher, Schoenfeldt & Shaw, 1990). The most severe
approach to layoffs from a human relations viewpoint entails the distribution of dismissal
notices to each terminated employee with the distribution severance benefits only as
mandated by law or union contract. Second, an organization terminates employees and then
voluntarily provides outplacement counseling and guidance. Outplacement counseling
typically focuses on job seeking skills, and the identification job/career alternatives. A third
approach to downsizing offers early retirement to targeted employees. The organization
offers financial incentives to workers nearing retirement age to encourage or induce
attrition. In this case, the termination of employment is somewhat voluntary as the worker
exchanges financial benefits for their term of employment. A single approach to reducing a
workforce is not clearly better than the others (Kozlowski, et al., 1993). The effectiveness
of the approach used is dependent upon market conditions, worker characteristics, and the

solvency of the organization.

11

Downsizing usually elicits numerous negative responses from involtmtarily
dismissed employees. Financial strain occm's for terminated workers when they are unable
to quickly find stable alternative employment (Buss & Redbum, 1987; Warr & Jackson,
1984) or if they must accept lower paying or part-time jobs which do not match their
previous earnings (Kozlowski, et al., 1993). Yet, the impact of job loss on involuntarily
dismissed employees far exceeds the problem of income stability. Job loss creates
psychological and physiological reactions which lead to stress and strain for the terminated
worker (Cobb & Kasl, 1977; Kozlowski, et al.). Some reported reactions to job loss are high
blood pressure and headaches (Cobb & Kasl, 1977), lead to poorer psychological health
(i.e., depression, anxiety, tension, anger) (Cobb & Kasl, 1977; Dooley & Catalano, 1988).
and create negative attributions or attitudes such as external locus of control, poor life
satisfaction, and low self-esteem (Burke, 1984; Liem & Liem, 1988). As unemployment
persists for an involuntarily dismissed worker, the intensity and inability to cope with the
consequences of downsizing increase (Kinicki, 1985; Kinicki & Latack, 1990).

Layoff Survivors. An organizational downsizing may also influence layoff
survivors' work behaviors and attitudes (Brockner, 1988). On a beneficial note, a reduced
workforce can bring a new sense of efficiency, effectiveness, or relief to an organization and
its members (Brockner, 1988; Gutknecht & Keys, 1993). For example, if a less productive
employee who constantly requires assistance to reach average performance is terminated,
the surviving workers may experience a sense of relief and increased work performance.
Additionally, as layoffs are announced, the surviving workers may be relieved once they

know they have not been laid off (Brockner, 1988). However, layoffs can produce

12

uncertainty, guilt or anger among survivors in the reconfigured work environment
(Gutknecht & Keys, 1993; O’Neill & Lenn, 1995). Brockner, Wiesenfeld, Reed, Grover
and Martin (1993) report that employees’ attitudes toward a downsizing effort depends on
(a) how fairly the survivors feel the layoffs were handled and (b) their perceptions of their
coworkers’ reactions to the layoffs. If survivors sense that the downsizing was handled in an
unfair manner, they are likely to define their new work context as unfair, to harbor feelings
of distrust and uncertainty toward the organization (Brockner et al., 1993 ), and to be less
accepting of the changes to their work environment (Goodman, Bazennan & Conlon,

1 980).

Survivors of a laboratory-simulated downsizing perceiving the procedure used to
terminate workers to be unfair also have negative responses to downsizing (Brockner, et al.,
1985). Other laboratory and case study research suggests that layoff survivors’ proximity to
terminated employees are negatively related to their perceptions of equity, the extent that
they can control the layoff process, and job security, but positively related to feelings of
guilt (Brockner, et al., 1985; Brockner, et al., 1987; Gutknecht & Keys, 1993). In addition,
survivors’ perceptions of their control of the layoff process, job security, and the fairness of
the downsizing are positively related to employees’ job satisfaction and organizational
commitment, but negatively related to intent to withdraw from the organization (Davy,
Kinicki & Scheck, 1991).

Survivors’ positive or negative perceptions may be reinforced by impressions they
receive from their fellow survivors (Ray & Miller, 1991; Salancik & Pfeffer, 1978). As

employees are dismissed from a work environment, the remaining employees may

13

positively or negatively characterize the downsizing depending on their gain or loss of
connections to valuable information and resources (Burt, 1992a). Certainly, the news media
emphasize downsizings’ harshest outcomes - the dismissal of valuable colleagues, increased
workloads, and the perceived lack of strategy employed by the executives (Kleinfield, 1996;
Uchitelle & Kleinfield, 1996). Regardless of their reaction to the downsizing, layoff
survivors Often need to acquire a different set of skills to perform their jobs in the post-
downsizing environment. Surviving employees may undergo training to facilitate the
addition of duties or responsibilities (job enlargement) or a modification to their existing
position (job enrichment) (Noe, 1986).

In sum, downsizing is a prevalent organizational change practice, may lead to
considerable changes in an organization’s mission and structure, and is commonly framed
by surviving employees in a negative light (Brockner, et al., 1993; Gutknecht & Keys,
1993). While researchers have devoted considerable attention to organizational downsizing
and its effects, Kozlowski, et a1. (1993) argue that our understanding of downsizing-related
phenomena would be improved by longitudinal field research investigations (versus
laboratory experiments with student participants). In addition, Kozlowski et a1. point out
that micro-macro integrative approaches are uncommon in this research domain and the
nature of downsizing highlights the need for more specific attention to the survivors’
attitudes and perceptions of the multi-level process. For instance, survivors’ reactions to a
downsizing may be considerably influenced by changes in communication network patterns
and the loss or gain of new task network members. Communication networks are the

primary mechanism through which employees send and receive work-related information.

14

Unexpected changes to patterns of communication relationships may disrupt the
information flow and negatively shape their perceptions and attitudes. However,
improvements in information flow through new network linkages may positively influence
employee reactions.

Communication Networks

Communication networks can be viewed as relatively stable configurations of
relationships in an organizational context (Johnson, 1993). Organizations formally divide
their members into departments or functional group structures, and a variety of
communication links exist within and between formal work groups. Members maintain task
and social communication with other members in their network within the same hierarchical
levels, different hierarchical levels, and across varying departmental and geographic
locations. Each member of an organization contributes to a communication network as: (a)
a member of the groups, clusters or cliques that comprise the network, (b) the individuals
who link the clusters or groups together as a liaison or bridge, or (c) those individuals who
are not a part of a group or do not link groups together (non-participants) (Monge &
Eisenberg, 1987). The nature of message exchanges between members determines their
status as a participant or a non-participant as well as their formal and informal roles in the
network.

In any communication network analysis, the linkages between all possible pairs of
members are of interest to researchers and indicate the presence or absence of relationships
between network members. Network linkages have four main properties: (a) strength or
intensity - representing the amount of information and affect that flows through the linkage,

(b) symmetry - representing the degree to which both parties enter into similar relations

15

with one another, (c) reciprocity - the extent to which two linked members report the same
relationship to one another, and (d) multiplexity - the degree to which the same players are
involved in different networks in the organization (Johnson, 1993). Together, these
properties provide the basis for understanding an employee’s role in the network system.
The strength, symmetry, reciprocity, and multiplexity of any employee’s network role are
unlikely to be an exact match with any other member’s role. Some employees have more
unique information contacts and coordinate their work efforts with a greater set of members
than others. Downsizing potentially places a particular strain on the availability of
members’ information and resources. Changes or disruptions to the access of information
and resources in a network may create structural holes.

Structural Holes. According to Burt (1992a, pp. 1-2), structural holes are
“disconnections or nonequivalencies between players in an arena.” As such, a structural
hole represents a separation between non-redundant contacts in a network (Burt 19923) and
may be measured on a continuum representing the extent that structural holes are prevalent
in an organization network (Susskind, 1996). Non-redundant contacts represent a unique
source of information and resources not available from other network contacts. An
employee’s separation from non-redundant ties represents a loss of unique information or
resources. In contrast, redundant ties are linkages through which a network member gains
similar information or resources from two or more people. Burt (1992a) suggests that
network members have a greater proportion of redundant ties than non-redundant ties in
their immediate work group and non-redundant ties occur more frequently outside a

member’s immediate work group.

16

Structural holes exist in a functioning organimtional network as part of its normal
state, but can increase as a result of voluntary turnover of personnel and transfers. The
proportion of structural holes in a network is likely to increase as organization-wide
downsizing occurs. However, downsizing may also remove structural holes from a network.
Employees who were once denied access to vital personnel may find themselves with
access to additional information or resources unavailable in the pre-downsizing network. It
is possible that an organizational downsizing can simultaneously close and open structural
holes in a network and create advantages for some members and disadvantages for others.
Burt (19923) primarily focuses on the competitive advantages that structural holes create
and suggests that members open and close structural holes by their own actions to maximize
their personal position in a network. In the case of organizational downsizing, structural
holes open and close as a function of the reduction in network personnel. Downsizing
removes much of the choice involved in manipulating structural holes for personal gain.

Burt (1992a ) identifies three kinds of “capital” (p. 8) or opportunities which may
arise out of the formation of structural holes: social, human, and financial. Social
opportunities concern the formation of new social relationships with others, and are
available to all network members. Human opportunities are the individual qualities a
network member shares with others. Each network member possesses individual talents
which may be valued by other network members. Financial opportunities represent tangible
resources which are needed and essential to work functioning. Financial resources are not
necessarily available to all network members as some network members have greater access

to funding, equipment, and supplies. In short, employees can benefit from structural holes

17

with increased access to social, hmnan and financial resources flour a network. Conversely,
structural holes can reduce an employee’s position power and access to resources through
the losses of social, human, and financial capital. Losses may be difficult to recover in the
new network and can isolate employees in the post-downsizing network.

Network research focusing on structural holes to date is limited. In a longitudinal
study, Burt (1992a) examines structural holes in producer-supplier-consumer markets
under stable market conditions. Specifically, he examines the relationships between
structural holes and producers’ ability to negotiate prices with both suppliers and
consumers. The results suggest that producers’ profit margins increase as they experience
fewer structural holes as evidenced by a decrease in structural autonomy. However, as
suppliers and consumers experience fewer structural holes, their autonomy and profit
both decrease (Burt, 1992a) indicating a different impact of structural holes for different
network participants.

In a second investigation, Burt (1992a) contrasts the promotion rate of managers
based on their adjacency to structural holes. Using a measure of constraint to indicate the
presence of structural holes, Burt assesses the “closeness” with which each manager
perceives other hierarchically similar network members. Comparing closeness to the
number of days since each manager’s last promotion, Burt reports that managers’
experience of structural holes is related to a rapid promotion. However, the promotion
effects of structural holes on promotion rates are less pronounced for sample groupings

such as women and older workers.

18

In a similar vein, Ibarra (1995) uses indicators of structural autonomy (e.g.,
density and number of contacts) to investigate the impact of minority interaction in
informal managerial networks. Ibarra (1995) suggests that managers encounter varying
effects from structural holes due to structural constraint present in their network
relationships. As access to information or resources appears limited, structural constraint
inhibits minorities’ network relationships (Ibarra, 1995).

Structural Hole Measurement. Structural holes can be identified in a network
analysis through measures of structural autonomy and prominence (Meyer, 1996;
Susskind, 1996). Structural autonomy indicators identify the structure of network
relationships which lead to the gain or the loss of access to sources of capital (Burt,
1992a). These indicators primarily examine horizontal relationships in a network and gauge
the breadth of a member’s network contacts (Burt, 1991). Prominence considers a network
member’s hierarchical influence and the extent to which they are considered central or
valuable in a network (Brass & Burkhardt, 1992; Burt, 1991).

Previous research on structural holes focuses primarily on measures of structural
autonomy. In Burt’s research (1992a), a combination of eight individual autonomy
indicators are used in varying combinations, with little explanation as to why some
indicators are used and others are not. Susskind (1996) suggests that (a) the use of
autonomy indicators in research to date is sample-specific and lacks a theoretical
explanation for the use or exclusion of each indicator in the measurement of structural holes
and (b) a rationale for the use of the indicators is wanting. Prominence examines the value

of relationships to others members in a network and explains network interaction as

19

members compete for resources and information. Investigations considering str'uctm'al
autonomy and prominence measures together in the study of weak ties (Meyer, 1996) and
structural constraint (Ibarra, 1995) suggest that the combination of autonomy and
prominence measures are useful in describing organizational communication behavior.

In an initial test of 6 structural autonomy measures used by Burt (1992a) and 2
theoretically pertinent prominence measures (Burt, 1991) as a combined index, Susskind
(1996) performed a confirmatory factor analysis of Burt’s (1991, 1992a) structural
autonomy and prominence indicators at three time periods in an organizational network. His
results suggested that a unified structural hole index comprised of two structural autonomy
measures (non-redundant contacts and constraint) and two prominence measures (extensive
ties and exclusive ties) demonstrated sound psychometric properties to measure structural
holes on a continuum. Table 1 describes the indicators which, when summed, create an
index with scores ranging from “many holes” to “few holes.”

As suggested by the factor analysis and subsequent tests of the structural
autonomy and prominence indicators as predictors of chaos and openness to change,
these autonomy and prominence indicators provide a complimentary measurement of
structural holes (Susskind, 1996). While autonomy examines how unique information and
resources are limited by network relationships (Burt, 1992a), autonomy does not
specifically consider hierarchical influence in network as a means of access to resources.
In contrast, prominence explicitly considers how hierarchical influence inherent in
network positions. The combination of autonomy and prominence in an index emphasizes

a network’s impact on the individual by allowing network properties to be associated with

20

0.838: 05 :_ 0850
0.5082 0. 053 $8888 <

0.838:
05 :_ 08586 .850 805 8050—00 5085:: .8 80.30 05 m:
8:808 : 85—3 05 58:8 05 $8880: 5023 .8: 00:05:88 <

8:. 0>_m:_0xm

 

85:08 05 8:: 58:85
25082 .8058:: 0—838: .850

.5038: 05 E
«89:88 850 :85 80520: 03880 808:8: 83000..
89:85 a £033 0: 8&0: 05 88055 85.3 :8: 00:85:08 <

8E. 0>_m:8xm

 

89:86 a .3 808M898

.8388: 30.5 0 :0 80—0800: 5 8500855

 

02:8: 80:85:: 85:00 2:. M8588 a «:88 8:3 05 88055 £023 80: 38088: :< 5:88:00
8850 05 8858:: x838:
. 8000: .8 30:8. 5 0.838: 850 88:5 8:850 80808: 8 852555 .8 8088 08880

25082 0.85:2: a .8 308850 2:.

83:: .5080:0::-:0: £8882 £033 :8: 3:888: :<

8058:0582

 

88m 88802

808m 9
£58520”—

88%:

ED: who—V:—

80: 08::— 0_0E 8:30:05 05 .8 85588Q <

_ 030,—.

21

individual—level perceptions and attitudes in organizations. The measurement of structural
holes’ influence on employees is particularly important in understanding the impact of
downsizing. Given the description of information flow and the loss of resources when a
work group member is dismissed, it is anticipated that structural holes will occur most
prominently in work groups with the greatest number of layoffs. While all organizational
networks have structural holes to some extent (Burt, 1992b), a downsizing is likely to create
a greater proportion of structural holes in the recently reconstructed communication
network.

At the present time, the impact of structural holes on individuals’ perceptions and
attitudes is largely conjectural. In particular, little is known regarding the stability of
structural holes among employees over time and how they may change as a result of an
organizational downsizing. Consequently, this investigation begins with the following
questions:

RQ 1: How prevalent are structural holes in an organizational network?
RQ 2: How do employee structural hole experiences change as a consequence of an

organizational downsizing?

Perceptions and Attitudes of Downsizing Survivors
Structural holes are a consequence of employees’ links to employees throughout an
organization, span the confinement of work group membership, and are, therefore, a macro-
level phenomenon (Miles, 1980). However, rather than having a generalized impact, each

member’s structural hole experience is likely to be unique and reflect the individual’s role

22

and the member’s task network. Of interest in this investigation is the relationship between
structural holes and employee attitudes and perceptions and an integration of micro- and
macro-levels of analysis (c.f. Rousseau, 1988). Calls for the integration of micro and macro
organizational processes are not new (Jablin, 1987; Krackhardt & Porter, 1985, 1986; Staw
& Sutton, 1993; Weick, 1993), especially with regard to organizational networks (lndik.
1965).

To date, few macro-micro integrative studies using network analysis exist.
Krackhardt and Porter (1985) identify the relationships among workers’ fiiendship
networks as an influence on employees’ attributions of dissatisfaction and organizational
commitment. They report that employees attribute dissatisfaction as a cause of turnover for
their peers, but do not attribute peers’ commitment to the organization as leading to
turnover. Hartman and Johnson (1989) investigate the communication network structure
property of contagion to explain individual reactions and report structural equivalence to be
a better indicator of organizational commitment than cohesion. This study extends the work
of Krackhardt and Porter (1985) and Hartman and Johnson (1989) by examining the
linkages between formal organizational communication network relationships, individual
reactions, and an organizational reconfiguration. Specifically, this investigation seeks to
identify the impact of the presence of structural holes on individuals’ levels of information
deficiency, stress, job satisfaction, and organizational commitment as a result of an

organizational downsizing.

23

Information Deficiency

The access to information is necessary to perform work-related functions and is
critical to making sense of organizational developments (Weick, 1979, 1993). The
perception of receiving inadequate amounts of information affects employees’ willingness
to participate in organizational changes (Miller, Grau & Johnson, 1994) and downsizing
survivors’ certainty about their career futures (Johnson, Bemhagen, Miller & Allen, 1996).
As a consequence of downsizing, emerging structural holes can disrupt the flow of
information among survivors and, subsequently, impact their perception of a deficiency of
work-related information.

Perceptions of information deficiency are likely to vary among layoff survivors. In
general, employees who are well connected in a network are likely to be more aware of the
formal and informal decision making processes and current developments within the
organization (Ibarra, 1993). Yet, a downsizing is likely to decrease the perceived amount of
information received by network members. Members who are disadvantaged in their
network connections and affected by structural holes are likely to have greater difficulty
obtaining information about the organizational changes and will be more aware of losses of
information sources and resources. Further, the loss of information associated with
structural hole experiences are likely to have both short and long term consequences.

S_tre_s_s_

Stress is believed to be a natural consequence of downsizing (Kalrn & Byosiere,

1992). Organizational stress can result in physiological harm (e.g., cardiovascular and

gastrointestinal symptoms) (Fried, Rowland & Ferris, 1984), psychological strain (e.g.,

24

burnout, job dissatisfaction and depression) (Cooper & Payne, 1988) and inappropriate
behavioral responses (e.g., absenteeism, drug and alcohol abuse, and accidents) (Davidson
& Cooper, 1986; Cobb & Kasl, 1977). Layoff survivors experience stress in the downsized
environment as a result of increased work requirements, the loss of valuable links to
informational resources, an increased sense of uncertainty, and/or a fear that they may be
laid off themselves. The potential impact of stress on employees experiencing an
organizational downsizing is exemplified in an investigation of the threat of job loss on
employee attitudes in manufacturing plants that closed and did not close. Cobb and Kasl
(1977) report the threat of job loss to be psychologically damaging to employees, in the
form of increased stress and anxiety, in plants both experiencing and not experiencing
layoffs. However, employees in plants which had experienced layoffs reported greater
psychological damage than those in plants not experiencing layoffs.

While increased levels of stress may be an inevitable result from downsizing (Cobb
& Kasl, 1977), the uniformity of the experience and the duration of increased stress is less
clear. As posited earlier, the emergence of a structural hole is likely to be localized to
certain network linkages and unlikely to affect all employees equally. While the threat of
job loss (Cobb & Kasl, 1977) or the perceptions of the fairness of a downsizing (Brockner
& Greenberg, 1990; Davy, et al., 1991) may increase surviving employees’ levels of stress,
with the experience of structural holes the loss of information and resources will have a
disproportionate influence on employees’ level of stress. In addition, employees’ levels of

stress are likely to continue until new relationships are formed and proven reliable in

25

providing information. In contrast, those employees who experience fewer structural holes
will perceive lower information deficiency and, consequently, lower levels of stress.
Job Satisfaction and Organizational Commitment

As depicted in Figure 1, information deficiency and, consequently, increased stress
result from organizational downsizing. Several studies link employee stress with job
dissatisfaction (Ivancevich, Matteson & Preston, 1982; Jackson, 1983; & Howell, Bellenger
& Wilcox, 1987) and organizational commitment (V redenburgh & Trinkaus, 1983).
Further, stress and job satisfaction are strongly correlated and in some cases act as
moderators of measures of organizational commitment (Mathieu & Zajac, 1990; Tett &
Meyer, 1993), implying that stress impacts affective responses and perceptions of workers
in relation to their jobs.

Job satisfaction is an attitudinal measure which relates past events and rewards to
current feelings about a job (Lawler, 1986). Changes in job satisfaction are particularly
relevant in regard to organizational downsizing due to its strong empirical relationship to
organizational outcomes such as turnover intention and organizational commitment (T ett &
Meyer, 1993; Davy, etal., 1991), and job performance (Meyer, Paunonen, Gellatly, Goffin
& Jackson, 1989; Iaffaldano & Muchinsky, 1985). It is anticipated that when dealing with
increased levels of stress, employees find it more difficult to comfortably perform normal
work functions. Difficulties in job performance related to downsizing are likely to lead to a
decrease in job satisfaction.

Organizational commitment consists of: (a) a strong belief in and acceptance of the

organization's goals and values; (b) a willingness to exert considerable effort on behalf of

26

the organization; and (c) a strong desire to maintain membership (employment) in the
organization (Mowday, Steers & Porter, 1979). Research indicates that layoff survivors are
less committed to the organization when they perceive their jobs to be less intrinsically
interesting, they perceive organizational changes to be unfair, they have solid attachment to
layoff victims, and their coworkers harbor a negative attitude (Brockner et al., 1993).
Organizational commitment may also mediate the relationship between job satisfaction and
employees’ behavioral intent to withdrawal (Davy, et al., 1991). Evidently, when workers
are satisfied with their employment relationship, they perceive it as an advantage to remain
employed (Lawler, 1986; Locke & Latham, 1990). Other research investigations support
job satisfaction’s positive impact on organizational commitment (Davy et al., 1991; Meyer
et al., 1989).

In sum, while numerous studies have established the overall link between
downsizing survivors’ job satisfaction and commitment, to date little is known regarding
the impact of employees’ network position on their perceptions of information adequacy,
stress or their job satisfaction and commitment. Consequently as depicted in Figure 1, the
general static model hypothesizes that a greater proportion of structural holes in a member’s
network is positively related to the perception of information deficiency. The perception of
information deficiency is then positively related to higher levels of stress. Subsequently,
stress is negatively associated with job satisfaction, which is positively related to
organizational commitment.

As a result of a downsizing event, layoff survivors will continue to redefine

communication relationships, establish new linkages, and acquire and provide information

27

to new or different sources. Structural holes emerging (or disappearing) from network
reconfigurations can affect layoff survivors' levels of stress, information deficiency, job
satisfaction, and organizational commitment in the downsizing aftermath. While
downsizing events continue to be used as a means for improving organizational survival
rates, efficiency, and profitability, a limited number of studies examine the effects of a
downsizing on surviving employees. Given potential changes to network structures and the
emergence (or disappearance) of structural holes, it is necessary to identify changes in
organizational networks and employee attitudes and perceptions over time.

In addition to the static model, it is hypothesized that longitudinal relationships
will exist among the variables. As presented in Figure 2, each Of the preceding variables
will exhibit autocorrelation in its subsequent time period. In other words, structural holes
prior to the downsizing should predict structural holes shortly following (e.g., 2 months)
which should then predict structural holes at a more distant time from the downsizing
(e.g., 5 months). In the absence of information about the extent to which structural holes
open or close around downsizing survivors, it is predicted that individuals beset with a
higher preponderance of structural holes will likely find themselves with a similar
network circumstance following the downsizing. Likewise, employees experiencing few
structural holes are more likely than not to find themselves surrounded by few structural
holes post-downsizing. In general, following a downsizing employees are likely to
continue their normal work functions with possible changes in the magnitude of their
tasks but with somewhat similar sets of task connections. The contention that employees’

network configurations are likely to remain relatively stable despite a downsizing finds

Prior to the
Downsizing

Two Months

After the
Downsizing

28

Five Months
After the
Downsizing

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

, + ,
Structural ———-e Structural
Holes Holes
+
Information Information
. + .
Deficrency —-—-v Deficrency
+
| Stress l + | Stress
——e
+
Job Job
Satrsfactron Satisfaction
+
- Organizational + Organizational
Commitment ——’ Commitment

 

 

 

 

 

 

__2‘___,

\
+

——e

 

 

Structural
Holes

 

 

 

Information
Deficiency

 

 

 

Stress

 

 

 

Job
Satisfaction

 

 

 

Organizational
Commitment

 

 

Figure 2. The Hypothesized Longitudinal Model

 

 

 

 

 

29

support from research indicating that (a) network positions such as liaisons and bridges
are associated with employee communication abilities (Johnson, 1993) and (b) network
positions remain fairly stable over time (Monge & Eisenberg, 1987). While it is
anticipated that a downsizing will create more structural holes for some employees and
reduce some for others, these changes are predicted not to override an employee’s overall
structural hole circumstance. It is also predicted that autocorrelation will be present
among information deficiency, stress, job satisfaction, and organizational commitment.
Given the enduring effect of an organizational change, whether it is manifested in
a convergence or reorientation intervention (Cameron & Freeman, 1993; Tushman &
Romanelli, 1985), it is also predicted that the impact of structural hole experiences on
employees’ perceptions and attitudes will continue over time. As indicated in Figure 2.
the cross-lagged relationships among the variables act as a causal force in the subsequent
time period. Specifically, structural hole experiences prior to the downsizing will
positively impact information deficiency shortly following the downsizing and
information deficiency prior to the downsizing will lead to increased stress shortly after
the downsizing. In turn, the experience of stress prior to the downsizing will negatively
influence employee job satisfaction following the downsizing, and job satisfaction prior
to the downsizing will lead to organizational commitment in the interim following the
downsizing. Two months following the downsizing, structural hole experiences will
predict information deficiency five months after the downsizing just as information
deficiency will lead to stress, stress will be negatively related to job satisfaction, and job

satisfaction will lead to organizational commitment five months after the downsizing.

Chapter 2
Methods

Procedure

Employees’ network relationships and perceptions and attitudes were measured 60
days prior to the downsizing event (T1), 60 days after the downsizing (T2), and 150 days
after the downsizing (I‘ 3). Subjects completed demographic information at T1 only. Due to
the nature of reporting communication network relationships, anonymity was not possible.
However, each participant was assured strict confidentiality in their responses and
guaranteed that their name or the name of the organization would not appear on any report
generated from the data.
Participants

One hundred and thirty employees at a large international hospitality company's
corporate headquarters surveyed at T1 are described as 56 % male and 44 % female and
between the ages of 20 and 57 (M = 32.42). Regarding the tenure of the participants at T1,
approximately 12 % (N=17) had been employed one year or less, 23 % =30) between 1
to 2 years, 36 % (N=46) between 2 to 5 years, and 29 % (N=37) for 5 or more years. The
demographic characteristics of the sample remained relatively stable at the next two time
periods. The surviving 98 employees surveyed at T2 and T3 were 58 % male and 42 %
female and between the ages of 20 and 57 (M = 32.92). Among surviving participants, 11
% (N=12) had been employed for one year or less, 21 % (N=20) between 1 to 2 years, 37%

=35) between 2 to 5 years, and 31 % (N=29) from 5 years or more. At all three data

30

31

collection points, subjects represented accounting, finance, marketing, rooms, engineering,
administration, and personnel units.
Nature of the Downsizing

In keeping with DeWitt’s (1993) recommendations, the downsizing of the
company’s corporate headquarters focused on a structrual reorientation and domain
retrenchment. Specifically, the organization attempted to remove redundancy from the
corporate headquarters by decentralizing a ntunber of corporate functions to the operational
unit level while simultaneously modifying the offering of and evaluation of primary
services provided to guests.
Structural Hole Measurement

Network Relationships. The survey initially asked participants to specify “all the
people you talk to within the corporate office on a regular basis during the course of a
normal work week.” Prior to Completing the survey, they were given specific oral
instructions to report formal work relationships related to the performance of their job.
Formal work relationships were targeted instead of formal and informal relationships in an
effort to satisfy the organization’s requirements and reduce the survey’s complexity.
Participants were provided with an alphabetized directory of all employees listed by
functional unit and were asked to report their formal work relationships only by indicating
which organizational members they had spoken with during the course of a normal work
week. This self-report data gathering technique represents one of several techniques
available to assess participants network relationships (Monge & Contractor, 1987). Due to

the extreme density within the network sample, only participants’ report of direct

32

linkages were used in the analyses. Each participant had an extensive number of indirect
linkages evidenced by the small number of step linkages separating all network members
(each member in the network could be reached within four step linkages). This extreme
network density confounded the interpretation of the network data. Consequently, only
direct linkages were used in the identification of structural holes in this sample.

While the original intent was to use both direct and indirect linkages in the
network analysis, the use of only direct linkages made the network data easier to
interpret. With indirect linkages, it is not possible to determine the content transferred
through the relationship. Indirect linkages commonly involve gatekeeping. blocking,
filtering, and innuendo which are less present in communication sent through direct
linkages (Huber & Daft, 1987; Stohl & Redding, 1987). Further, the organizational
change may exaggerate the number of indirect linkages due to speculation and the
uncertainty of future outcomes. See Susskind (1996) for further details on the use and
implications of only direct linkages in a network analysis.

The Structural Hole Index. The Structure 4.2 network analysis program (Burt,
1991) determined the communication patterns in the communication network at each data
collection period. The responses of the surviving 97 participants were used in the analyses
at all three time periods. The Structure 4.2 output summarized the distribution of autonomy
and prominence across individual relationships in the network by considering the proximity
and equivalence of each network member’s direct relationships.

Prior analyses revealed that non-redundant ties, constraint, extensive relations, and

exclusive relations indicators provided an appropriate measure of autonomy and

33

prominence elements (Susskind, 1996). In a test of the index, autonomy (i.e., non-
redundant ties and constraint) and prominence (i.e., extensive ties and exclusive ties)
network output were converted to a similar numerical format and standardized. This
procedure entailed multiplying constraint, extensive ties, and exclusive ties (which appeared
in decimal form) by a constant to yield an integer which was then converted to a z-score.
Non-redundant ties, extensive relations, and exclusive relations were reverse coded since
they were hypothesized to be negatively related to structural holes (Susskind, 1996). As a
consequence, a positive structural hole score indicated the presence of structural hole
experiences while a negative score indicated the absence of structural hole experiences.

In keeping with the criteria identified by Hunter and Gerbing (1982), the index
was judged to be unidirnensional as its indicators loaded primarily on the specified factor,
did not exceed their specified confidence interval at the p < .05 level, and sum of squared
errors for tests of internal consistency for the index was non-significant at the p > .05
level (x2 (6) = .30, x2 (6) = .41, and x2 (6) = .31, at T1, T2, and T3, respectively).
Furthermore, the index demonstrated criterion validity as evidenced by the correlations
between the index and its hypothesized outcomes across three time periods. Table 2
reports the final factor loadings and the index’s reliability for each time period.
Summated Measurements

In the survey participants also responded to established scales to assess their
perceived levels of stress, information deficiency, job satisfaction, and organizational
commitment. Participants were asked to indicate their level of agreement with each scale

item question on a five choice metric (strongly agree, agree, neutral, disagree, and strongly

34

0.880.805 80:25:80 88:88:90

05 :8: 80.. 080880. ..+.. 0:: 80:88:. 8:00.880. :0. 0:. 50:. «E0: 080880.. ...... 898:: 05 8.. 8:00 80.88: 0.03 85 8.0.. 88:0: :35: .082

 

 

 

 

 

 

ma. 8. .o. 2:? M88820
8. on. 2.. + $8880 :5 80:: 0.80 :80. _
2.. on. 2.. + 0:88.00 :5 .0: 8:83 :00: 0. 8:8 5 808.98 820 0:». :8 888 8000: 2:03 _
mm. 5. 8. E: + 8: b: :0 0.82:. 0:50: 0 83 38:88 :5 8.. :83 0. 9:28:
on. 8. so. + .8888 m2. .8 .8: :8 _ 8:. 8:8 :0. 0. 0:08 :8 _
on. 3. 8. + 8: 0.83 0. 888800 05:80: :8 8 .8: 05 a “:5 0:. 8..
8. Na. 2.. a .:0_. 0:. ::3 005:8... b0> 08 non £5 :0 0.80: .82
on. N». m». . .:0.n £5 :_ 0: _ x83 .8 0:2 05 5:3 8828 580:8 :8 _
2.. 8. a... 8. . 8.. a... 05:33.. 0.5... 8532. _
8. :2... .88.. .85.
no. 3. 2.. :::_< M80585
3. mm. 3. 1: 0: 0. 03: :8 _ :03 30: ::3 88:25 0: 0. 0.8: _ 0.83 .8 8:08 2:.
ca. 0:. 3. 0:. 88:88 0:0 8 0: 30: 0:: 0:. 8 8.55 808880 08805:: .3. 8:3 0.50:: :8 _
3. no. 2.. 08 .8 .0098 0800: 8:3 30:0. ..8: .03. _
co. 2.. 2.. :83 8 088:3 8. .8 88:8: 952—800 0:. @500. 8:80 _ 00:. .00.. _
805m
3. 3. S. :::_< 9:03:20
cw. 3. 2.. 0:. 53 8588.8 :w:0:0 08:... .0: 0: 88:80 :5 8 0.0: :0 m:_0w 08:3 305. 0:3 0.80: 0:.r
3. no. 2.. .08: 8:08 :0 m:_0m 98:3 80:: 9:58: 0:. 0:0. .0>0 0:0 0: 00.: .00: _
no. 3. 2.. Ev $58800 £5 8 :0 88» 98:3 80:: 02000. _ 8588:: 05 .23 8:58 28:08:. :8 _
8:088: .8858. .5. 5.
3. v9. 2... 2:? 0:00:85
am. on. an. S: 8:. 0302005
0:. no. 8. 3: 8:. 0:08:8—
3. ea. 0:. 88.880
2.. 2.. 2.. Q: 88:0U 88:82-82
.80..— 0_0: 8.5.025

m 08:. N 08E ~08;

 

 

N 080,—.

35

disagree). To assess the dimensionality of the scales, confirmatory factor analysis tests of
internal consistency and parallelism (Hunter & Gerbing, 1982) were also applied to the
scales for each data collection period (T1, T2, and T3). Based on the results of the
confirmatory factor analyses, scale items at each time period were retained or dropped in
keeping with the recommendation that (a) less than 5% of items exceed their confidence
interval and (b) chi-square tests on the sum of squared errors for tests of internal consistency
and parallelism be non-significant. The final factor structure retained at T1 was applied and
tested for both '12 and T3 factor structures. Items dropped at any one time period were
excluded from the other time periods, and the factor analyses were repeated for the other
time periods. The final retained scale item factor loadings and scale reliabilities for T1, T2,
and T3 appear in Table 2.

Information deficiency was measured using a four-item instrument from Miller, et
al. (1994) and assessed respondents’ perceptions of receiving sufficient amounts of job-
related information. Based on the results of confirmatory factor analyses, one item was
excluded from the analyses. The sum of squared errors for the tests of internal consistency
were non-significant at a p > .05 level ()6 (3) = .03, x2 (3) = .01, and x2 (3) = .08 at T1,
T2, and T3, respectively).

Employees’ stress (or job-related tension) was assessed using a nine-item scale by
Kahn, Wolfe, Quinn, and Snoek (1964). The items from this scale assessed the participants’
perceived job and supervisory demands. Four of the nine items were retained for use in all

three time periods following tests of internal consistency and parallelism. The sum of

36

squared errors for tests of internal consistency were non-significant at p > .05 level (x2
(6) = .18, x2 (6) = 4.56, and x2 (6) = .22 at T1, T2, and T3, respectively).

The five-item, general job satisfaction scale (Hackman & Oldham, 1975) was used
to assess participants’ general level of satisfaction with their job and perception of their
coworkers satisfaction. Organizational commitment was measured using the lS-item
Organizational Commitment Questionnaire (Mowday, Steers, & Porter, 1979). Factor
analytic tests revealed that although the scales were internally consistent, the lack of
parallelism showed the scales to measure one underlying construct. After combining these
scales eight items were judged to measure employees’ affective commitment to their job in
a manner similar to other affective commitment instruments (e.g., Allen & Meyer, 1990;
Meyer, Allen & Gellatly, 1990; Mowday, et al., 1979). The final eight-item scale was
composed of three job satisfaction items and five organizational commitment items. The
sum of squared errors for the tests of internal consistency were non-significant at the p >
.05 level ()8 (28) = 3.70, x2 (28) = 12.19, and x2 (28) = 9.84 at T1, T2, and T3,
respectively). Tests of the static model (Figure l) and the longitudinal model (Figure 2)
were modified in keeping with the factor analytic results.

Tests of parallelism indicated that the structural hole index, information deficiency.
stress, and the revised affective commitment scale items loaded primarily on the theoretical
factor of choice and produced non-significant sum of squared errors at p > .05 level (x2

( 128) = 63.34, x2 (128) = 89.80, and x2 ( 128) = 36.82 at T1, T2, and T3, respectively).

37

Statistical Analyses

In testing the first research question, the means and standard deviations for
structural holes at each time period were calculated based on scores from the structural
hole index. The distribution of structural holes were assessed by comparing individual
aggregate structural hole scores with their network position as determined through
Structure (Burt, 1991). For the second research question, pre- and post-downsizing
structural hole index scores were compared through t-tests of standardized scores.

The static path models for T1 , T2, and T3 (Figure 1) and the longitudinal path
model (Figure 2) were tested using Hamilton and Hunter’s (1995) PACKAGE path
analysis program. Path models were assessed for fit based on the recommendation that
within each time period: (a) less than 5% of items exceed their confidence interval; (b)
global chi-square tests on the sum of squared error for model be non-significant; (c) each
link be tested for significance by calculating a confidence interval around the observed

correlations; and (d) the static model remains consistent in all three time periods.

Chapter 3

Results

Tests of the Research Questions

The first research question sought to discover the extent to which structural holes
were prevalent in the organization’s corporate headquarters. Measures of the overall
mean, standard deviation, and range for the standardized structural hole index scores
across the three time periods are reported in Table 3. In keeping with the effect of
standardization, overall index scores maintain a mean of zero and a standard deviation of
one‘. A positive structural hole score indicates an above average presence of structural
holes for employees, while a negative structural hole score indicates a below average
presence. An examination of the dispersion of overall structural hole index scores (e.g.,
standard deviation, range) revealed variability in structural hole experiences among
employees.

In an effort to better understand the extent of structural hole experiences among
employees in the headquarters, the structural hole means, standard deviations, and range
for each department were examined. As presented in Table 4, the descriptive statistics
show considerable differences in structural hole experiences among units. For instance, at

T1 prior to the downsizing, the mean of the structural hole index scores ranged from

 

1 The standard deviations of the structural hole index in this case should not be zero due to the way in

which the standardized scores were combined. Each structural hole indicator (in this case 4) was
standardized independently. The structural hole index was then calculated by combining each of the four
indicators per subject and dividing the sum by four. As they were combined into the structural hole index
they were not restandardized. The means of the structural hole index items remained at zero (within
rounding error) because they are additive. In contrast, the standard deviations are not additive. Therefore,
the standard deviations reported for the index are the deviations of the combined index, not the individual
Standardized items (which have a standard deviation of one within rounding error).

38

39

 

 

Table 3
Mean, Standard Deviation, and Range of Overall Structural Hole Index Scores
T. w' T. T2 T3
Mean .004 -.002 ‘ -.006 -.005
Standard Deviation .65 .60 .44 .42
Maximum 1.42 1.42 1.15 1.08
Minimum -1.72 -l.50 -1.12 -1.11
N 130 97 97 97

 

' TI m, reports structural hole scores for the entire sample prior to the downsizing

40

 

 

 

 

 

2 2 _m a a a 2 Z
8.. 2..- 3: mm: 8.- 3.- _ I. 52552
mm. mu. 3. 3. 84 3. mm. 5:852
on. «m. E. 9.. G. 3. mm. 5:350 2355
3. 2.- 2.- 2. mm. 2. cm; :82
09:; 2::-
2 2 .N o w a 2 Z
Er 8.. 3.- 2..- mn: 3.. N2. 55:52
3. en. hm. ma. m _ ._ mm. mm. Sagas—2
me. mm. em. 9.. S. cm. 3. 8:250 ecu—65$
:. 8.- 3.- am. an. m _. cm: 53.2
03,—. 25%
E m _ _m a a ¢ 2 Z
t... _- o2- omc 3.- w:- Rg- an. _- 5255.2
N: co. 3. mm; mm; 9.. mm. :55me
3. mm. em. 3. 2.. 3. me. 8:250 235%
ac. 8.- 8.. co. 3. me. am; :32
0:0 25...

.053an 532.2582 octoosocm «Boom acacia: 855.... 95.582

 

E5 3:285."— zm “”8on .82: 3o: ESQ—hm =Eo>O no own-9m EB £26300 FEE—Sm .522

v 03a...

41

M = .60 for Rooms to M = -.29 for Accounting. Regarding employees’ structural hole
experiences within unit, their structural hole experiences ranged from 1.42 to -1.37 in
Personnel to .22 to -1.39 in Accounting. Although not as disparate, similar variability
exists in unit structural hole scores at two months (T2) and five months (T3) following
the downsizing.

Individual employee structural hole index scores are reported in Table 5. The
variability in structural hole scores suggested that employees experience unique levels of
structural holes in their network, and their experience of structural holes is embedded
within their work group which differs from other work groups in its coordination needs at
the corporate headquarters.

The second research question inquired into the impact of downsizing on
employees’ structural hole experiences. As such, t-tests were conducted to assess the
significance of change among participants. In a comparison of the overall means of
structural hole index scores of T1 and T2 means and T2 and T3 means, no significant
differences were present in the mean scores. When examining structural hole index scores
at the unit level, the results of t-tests revealed significant differences for 2 of the 7 groups
between T1 and T2. Accounting unit mean structural hole scores were significantly
different at T1 (M = -.29) and T2 (M = -.20) (t (18) = -3.28, p = .004). The structural hole
mean score for the Rooms unit for T] (M = .60) was also significantly different (t (8) =
2.26, p = .039) at T2 (M =.28). No significant differences were found in any of the work

groups between T2 and T3.

42

Table 5
dividual tructural Hole Index For 'ci ts At Time e Time Tw d Time
Time 1 Time 2 Time 3
Structural Structural Structural
1D Unit Hole Index Hole Index Hole Index
1 Accounting 0.09 0.17 0.00
2 Accounting 0.22 0.33 0.23
3 Accounting —0.26 -0. 16 -0.04
4 Accounting .012 -0.03 0.01
5 Accounting -0.23 -0.12 -0.19
6 Accounting -0.01 -0.10 -0.10
7 Accounting .025 -022 -0.16
8 Accounting -0.42 -0.34 -0.28
9 Accounting -0.28 -0.14 -0.17
10 Accounting -0.12 -0.02 -0.02
11 Accounting 0.06 . .
12 Accounting .014 -0.01 -0.01
13 Accounting -0.18 -0.16 -0.20
14 Accounting -1.39 -1.12 -1.11
15 Accounting -0.l4 . .
16 Accounting -0.31 -0.31 .031
17 Accounting -0.25 -0.08 -0.20
18 Accounting -0.12 -0.07 -0.13
19 Accounting -0.49 -0.43 -0.37
20 Accounting 0.20 0.05 0. 12
21 Accounting -0.32 . .
22 Accounting -1.38 -0.96 -0.88
23 Accounting -1 .73 . .
24 Finance 0.40 0.34 0.33
25 Finance 0.44
26 Finance 0.40 . .
27 Finance 0.38 0.36 0.36
28 Finance 0.41 0.21 0.47
29 Finance 0.68 . .
30 Finance 0.45 0.53 0.51
31 Finance -0.94 -0.88 -0.85
32 Finance 0.37 0.36 0.19
33 Finance 0.41 0.43 0.35
34 Finance 0.12 . .
35 Finance 0.40 0.16 0.32
36 Finance -0.05 0.43 0.25
37 Finance 427 -0.57 -0.49
38 Marketing 1.07 1.01 0.96
39 Marketing 1.22 1.15 1.08
40 Marketing 0.67 . .
41 Marketing -0.32 . .
42 Marketing 0.82 0.70 0.67
43 Marketing 0.52 0.76 0.73
44 Marketing 0.80 -009 -0.06
45 Marketing 0.98

Table 5 continued
1D Unit
46 Marketing

' 47 Marketing
48 Marketing
49 Marketing
50 Marketing
51 Rooms
52 Rooms
53 Rooms
54 Rooms
55 Rooms
56 Rooms
57 Rooms
58 Rooms
59 Rooms
60 Rooms
61 Rooms
62 Rooms
63 Engineering
64 Engineering
65 Engineering
66 Engineering
67 Engineering
68 Engineering.
69 Engineering
70 Engineering
71 Engineering
‘72 Engineering
73 Engineering
74 Engineering
75 Engineering
76 Engineering
77 Engineering
78 Engineering
79 Engineering
80 Engineering
81 Engineering
82 Engineering
83 Engineering
84 Engineering
85 Engineering
86 Engineering
87 Engineering
88 Engineering
89 Engineering
90 Admin.
91 Admin.
92 Admin.
93 Admin.

Time 1
Structural

Hole Index
0.01
-0.30
0.33
-1.16
0.78
0.46
0.41
1.18
1.00
0.17
0.91
1.43
1.35
-0.80
0.59
0.71
0.51
-0.05
-0.45
-0.22
-0.23
-0.41
-0.09
~0.53
-0.49
0.97
-0.13
-0.27
-0.53
0.03
0.04
-0.38
-0.19
0.07
-0.92
-0.12
-0.19
0.04
-0.47
-0.30
-0.03
-0.59
-0.38
-0.80
0.06
0.25
-0.05
~0.30

Time 2

Structural
Hole Index

-0.13
0.41
~0.75

0.3 l
0.02
0.92
0.77
0.14
-0.04

0.57
-0.45

0.29

0.23
-0.53
-0.10
-0.13
-0.04

0.13
-0.24
-0.54
-0.97

0.09

0.10

0.01
0.06
-0.11

-0.20
-0.20
0.37

-0.63

0. l4
-0.69

-0.43
0. l3
0.15

-0.07
0.08

43

Time 3
Structural
Hole Index

-0.09
-0.02
-0.63

0.23
-0.02
0.89
0.95
0. l 5
-0.02

0.59
-0.35

0.40

0.23
-0.41
-0. l6
-0. l 6
«0.07

0.02
-0.20
-0.45
-0.92

0.07

0.07

-0.02
-0.10
-0. 12

-0.l4
-0.14

0.26
-0.54

0.14
-0.79

-0.43
0.12
0.05

-0.02
0.09

Table 5 continued

ID

94
95

97
98

100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
124
125
126
127
128
129
130
131

Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Admin.
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel
Personnel

Time 1
Structural
Hole Index

-1.00
-0.02
0.23
0.10
0.36
0.23
0.15
0.60
0.46
0.32
0.15
0.03
0.34
0.19
-l .50
-I .08
0.31
0.52
-0.1 1
0.1 1
0.16
0.03
0.12
-0. 12
-0.81
0.65
0.43
-0.37
-0.23
1 .05
0.09
1 .42
-0.43
-0.25
~0.07
-0. 17
-1 .37

Time 2
Structural
I-Iole Index

-0.1 1
0.12
0.21

-0.02

0.24
~0.68

0.18
0.04

-0.03
-0.80
.054
-0.48

0.23
0.58
0.15
~0.04
0.14

0.65
0.10

«0.1 I
0.43
0.66

-0.1 1
-0.63

0.13
-0.71

44

Time 3
Structural
Hole Index

~0.07
0.12
0.25
-0.01

0.24
-0.70

0.06
0.06

o0.10
-0.78
-0.48
-0.43

0.26
0.31
0.19

0.03
0.13

0.61
0.13

-0.10
0.43

0.83
-0.21
~0.03

-0.05
-0.60

45

The changes in personnel per group between T1 and T2 in terms of number of
employees lost, the number of bridge linkages lost, and the type of personnel lost (e.g.,
clerical, managerial) are reported in Table 6. A bridge linkage was defined as a
connection between two network members who hold membership in different groups and
link their two groups together (Richards & Rice, 1985). Each work group sustained some
loss through the downsizing, ranging from 4 to 6 employees. The Accounting, Finance,
Engineering and Personnel units increased in structural hole experiences while
Marketing, Rooms and Administration experienced a decrease in structural holes.
Between T2 and T3, Rooms and Personnel experienced small increases in structural
holes, Finance, Marketing and Administration experienced small decreases in structural
hole scores, while Accounting and Engineering experienced no change in structural hole
scores.

Test of the Static Model

The modified hypothesized model of the relationships between structural holes,
information deficiency, stress, job affect were tested with the TI , T2, and T3 data. The
means and standard deviations of, and the correlations among structural hole index,
information deficiency, stress and job affect are reported in Table 7. Path coefficients and
their standard error for the static model are reported in Figure 3. As detailed below, initial
path analyses revealed that the hypothesized model at all three time periods was overall
not a good fit to the data. The model was subsequently revised and tested for each time

period.

.oo._5 0:... Ya 03. on... 5252. 9.8... 0.2. .EBusz a. 335:. cm 3.86:. 20% 33.8“. < a
.95 on... can one 2:: 5053 28m 22.. .8335. a. 388:. ca 8.86:. 203 25.8.. < .

 

46

b8.“ 98 48.5.0
.EoEomewE mo. 8. . \- . mm .2886;

 

been 28 48.3.0

 

 

 

 

 

 

.EoEowmcaE 5.- 8.. m m . om co..§m.c.E.o<
has can 48.3.0
.EoEomSaE co. No. v .m R meteocficm
.msm .65 78.8.6 8. mm: o. m N. 253.
as” Ea .863 8.- 8.- w m 2 masons:
can 8.- :. _ . o 3 855”.
can Ba .85: 8. 8. a e a. 3:582
1.18.. .oacoflom ..an - PC 88m .2 ... - E 28m awe. 03... 08.... .m 2.0 08.... um EoEtmnoQ
..o 2.»... 5.6:. 0.0... .EBosbm 38.65 0.3. .mhausbm magma... 898.qu mooonEm
:. 693:0 :82 a. 098:0 562 03:5 .0 .38: Z ..o 23:52

 

03... 2:...- 3 0:0 68.... ES... 955 ram mow—8:0 $58.83.
0 63m...-

4'7

.88 88.. En 8055... v0.6:— 52: 0.2. 88086 .8

9:88 2: .82? 9:88. 9 0.6 awnz 80:03:83 .8 0033.3 0.8 0.983 80.3: 05 a. mos—g c5 .8. v a a 285:».m 08 on. 05% 823.088 “082

 

 

 

 

 

 

 

 

 

 

 

 

 

- 8.8. 8.8. 8.8 :8. 8.8. :8 8.8 28 8.8 8.8. 8.8 8.8 2... .85 82:2.
8.8.. - 8.8 8.8. 8.8 2.8 8.8. one. :8. 8.8 2.8 2.8. 8.8 8.. 3258.2:
8.8. $8 - 8.8. :8. 28 8.8 8.8. 8.8. 28. :8 88. 8.8 8.. @5685 8.353388:
98 :8. 8.8. - :8. 2.8. 2.8 8.8 2.8 8.8. 8.8. 38 :8 88 Eusgomfiauéw 8. a.
z? 8.8 28. 8.8. - one. 8.8. 8.8. 8.8. 8.8 8.8 8.8 8.. 2 608.. 823 a.
8.8. :8 :8 :8. 8.8. - $8 2.8. 2.8. :8 :8 :8. 2.. :8 328 fl 6
2.8 8.8. 8.8 :8 :8. 88 . :8 :8 8.8. 8.8. 2.8. 8. 8.... 3528832533 6.
8.8 8.8. 8.8. 8.8 2.8. :8. 8.8 - 2.8 8.8. 8.8. 8.8 8.8 8.8 585%: Easém 8. 6
:8 2.8. 8.8. 2.8 28. 2.8. 2.8 :8 - 8.8. 8.8. 2.8 88 88 .85. no. a. 5
8.8 8.8 :8. 8.8. 8.8 2.8 88. 8.8. 8.8. - :8 :8. no 2 $.ng 5
8.8. 8.8 8.8 8.8. 8.8 2.8 8.8. 28. 8.8. :8 - :8. 8.8 8.~ bfinfioéfifioEfi 5
8.8 88. 8.8. 8.8 8.8 8.8. :8. 8.8 2.8 2.8. 8.8. - $8 8.8 x3503: 85252.8 5
a: a: 8: 6 a. E .8 6 5 5 E 8 E :8: E2338

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m 05:. 23 .~ 2.8. ._ 2.8. 8.. 8880a 285m 23 .882 882280 2.3
8 05.8

 

 

(a)

 

Structural
Hole Index

 

 

(b)

(c)

 

 

Structural
Hole Index

 

 

 

 

Structural
Hole Index

 

 

n39‘

910)]

48

 

 

 

 

 

 

 

 

 

 

I f . .54"‘
n ormation
L09h

Deficiency ‘
lnforrnation '48.

.09
Deficiency '1—1"
I f . .6l$.$
n ormatlon

i071]
Deficiency

 

 

 

 

Stress

 

 

 

 

Stress

 

 

 

 

Stress

 

 

 

 

 

 

 

 

 

£73.)" Aff r d
07 CC! owar
Job
”38. Affect Toward

£09h
Job
133?. Aff T d
04 ect owar
Job

 

 

 

Note: “*” = p <.05, “***” = p < .001; the standard errors are in parentheses.

Figure 3. Tests of the Modified Hypothesized Model at (a) Time 1, (b) Time 2. and (c)

Time 3

 

 

 

49

The Hmthesized Model. The modified hypothesized path model tested at T1 identified
significant relationships among the variables. However, the test of the model produced a
significant chi square score (x2 (3) = 7.92, p = .048, SSE = .21) indicating that additional
significant relationships among the variables in the model were not identified. It is
important to note that the path coefficient from structural hole index to information
deficiency was negative, the opposite of the hypothesized relationship.

Tests of the path models at T2 also revealed a large sum of squared errors and a
significant chi square score (x2 (3) = 30.63, p = .00, SSE = .67). indicating that the model
was not a good fit with the data. In this case, the path coefficient between the structural
hole index and information deficiency was positive as predicted, but nonsignificant (r =
.18, p > .05).

Analyses of the path model at T3 again resulted in a significant chi square score
()6 (3) = 8.01, p = .046, SSE = .16) despite significant relationships between the
variables. Similar to the path coefficient at T1, the path coefficient from structural hole
index to information deficiency was negative and counter to the hypothesized
relationship.

An examination of the variables in the model at T1, T2, and T3 suggested that the
correlations between employee stress and job affect were weaker than those between
information deficiency and job affect. Specifically, stress was consistently correlated
with information deficiency (r corrected for attenuation = .54, .48, and .61 at T1, T2. and
T3, respectively), but stress was less correlated with job affect (r corrected for attenuation

= -.73, -.38, and -.83 at T1, T2. and T3, respectively). In comparison, information

50

deficiency was more strongly correlated with job affect (r corrected for attenuation = -.84,
-.94, and -.68 at T1, T2, and T3, respectively).

In keeping with the findings of Research Question 2 that indicated considerable
differences in structural hole experiences among departments, the hypothesized path
model was tested at T2 by separating the sample into two groups: (a) units which
experienced increases in structural holes (Accounting, Finance, Engineering, and
Personnel); and (b) units which experienced decreases in structural holes (Marketing.
Rooms, and Administration). A similar procedure was followed for separating groups at
T3. At T3, employee responses in Rooms and Personnel were aggregated to form the
structural hole increase group, and Finance, Marketing, and Administration scores were
combined to form the decrease group. As mean structural hole scores did not change in
the Accounting and Engineering units, subjects’ responses from these units were not
tested in a path model at T3. Results of the path analyses for units with increased and
decreased structural hole experiences at T2 and T3 revealed the model to be a poor fit to
the data as significant chi-squares resulted from each test of the model.

Tests of the Revised Static Model; Based on the analyses of the
relationships between information deficiency, stress, and job affect. a revised model with
a direct link between information deficiency and job affect was tested at each time period
(see Figure 4). With the stress variable removed, analyses indicated the model was a good
fit to the data at T1, T2, and T3. A test of the path model at T1 produced a nonsignificant
chi square score (x2 (1) = 1.88, p = .17, SSE = .04). While the path coefficient between

the structural hole index and information deficiency was negative (r = -.39, p < .05) and

51

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”39' Information ‘84” Job Affect
Structural
(a) Hole Index J—L’. 10 Deficiency ‘1_)'"07
Structural ‘18 Information -94“ Job Affect
b) (.11)I (.02)I
( Hole Index Deficiency
Structural -27‘ Information -68" Job Affect
(c) (.10) i (.06)
Hole Index Deficiency

 

 

 

 

 

 

 

 

 

 

Note: “*” = p<.05, “*"'” = p<.01; the standard errors are in parentheses.

Figure 4. Tests of the Modified Static Model at (a) Time 1, (b) Time 2.
and (c) Time 3.

52

the hypothesized relationship countered the relationship between information deficiency
and job affect (r = -.84, p < .001) was keeping with theory (Miller, Johnson & Grau,
1994) and supported the reasoning that the experience of information deficiency reduces
affect toward one’s job.

Results of the tests of the revised static model at T2 revealed a nonsignificant chi
square (x2 (l) = .40, p = .53, SSE = .008). While the path coefficient from the structural
hole index to information deficiency was nonsignificant (r = .18, p > .05). information
deficiency was significantly predictive of job affect (r = -.94, p < .0001 ). This negative
relationship suggested that those experiencing information deficiency were less content
with their employment.

A path analysis of the revised model at T3 resulted in a nonsignificant chi square
score (x2 (1) = 2.19, p = .14, SSE =.05). As in the model at T1, the relationship between
structural hole index and information deficiency was negative (r = -.27, p < .05), and the
path coefficient between information deficiency and job affect was negative as well (r =
-.68, p < .05). It should be noted that the relationships between the variables at T3 were
weaker than those reported at T1.

In an attempt to further explain the data, post hoc regression analyses were
conducted to determine if organizational characteristics (i.e., age. tenure. level of
education, sex) produced any discernible linear patterns among the variables. Regression
analyses did reveal any alternative explanations beyond the findings reported above.
Additionally, scatter plots were computed in order to detect nonlinear trends among stress

and the other variables in each time period. As with the regression analyses, the scatter

53

plots indicated that the relationships of stress with the other variables failed to exhibit a
consistent pattern over time.

In keeping with the exploratory nature of earlier analyses revealing considerable
variability in structural hole experiences among units, separate post-hoc path analyses
were conducted for units which (a) structural hole experiences increased and (b) structural
hole experiences decreased (see Figure 5). Units which showed increases in aggregate
group structural hole index scores were separated from those units showing a decrease.
From T1 to T2 increases in structural holes were experienced in the Accounting, Finance.
Engineering, and Personnel units (N = 66) while Marketing, Rooms, and Administration
experienced decreases in structural holes (N = 31).

Path analyses of the revised model at T2 for the units reporting increased
structural hole activity revealed a nonsignificant chi square (x2 (1) = .32, p = .57, SSE =
.01), suggesting the model is consistent with the data for this subgroup. The path
coefficients from the structural hole index to information deficiency and from
information deficiency to job affect were both significant at p < .05 level (r = .30, and
r = -.89).

Path analyses for the revised model applied to participants in units with decreased
structural hole experiences at T2 indicated that the model was consistent with data (x2 (1)
= .14, p = .71, SSE = .008). The path coefficient from the structural hole index to
information deficiency was nonsignificant (r = .20, p > .05), while the path coefficient

from information deficiency to job affect was significant (r = -.94, at p < .05).

54

 

(a) Units with Increased Structural I-Iole Experiences at Time 2

 

 

 

.30‘ -89"

Structural Information Job Affect
(.13)I (04h

Hole Index Deficiency

 

 

 

 

 

 

 

 

 

Units with Decreased Structural Hole Experiences at Time 2

 

 

 

.20 -094. ‘

Structural Information Job Affect
(.18)l (04%

Hole Index Deficiency

 

 

 

 

 

 

 

 

 

 

(b) Units with Increased Structural Hole Experiences at Time 3

 

 

 

mlS m60‘

Structural Information Job Affect
(.14)I (09)

Hole Index Deficiency

 

 

 

 

 

 

 

 

 

Units with Decreased Structural Hole Experiences at Time 3

 

 

 

534- 883“

Structural Information Job Affect
(17). (071'

Hole Index Deficiency

 

 

 

 

 

 

 

 

 

 

Note: “"‘” = p<.05, “"” = p<.001; the standard errors are in parentheses.

Figure 5. Tests of the Revised Static Model For Units With Increased and Decreased
Structural Hole Experiences at (a) Time 2, and (b) Time 3.

55

A path analysis of the revised model at T3 applied to participants in units with
increased structural hole experiences (e.g., Rooms, Personnel, N = 25) indicated that the
model was consistent with data (12 (1) = .28, p = .60, SSE = .01). As with the T1 model
for this subgroup, the path coefficient from the structural hole index to information
deficiency was nonsignificant (r = -. 15, p > .05), while the path coefficient from
information deficiency to job affect was significant (r = —.59, at p < .05).

Path analyses for the T3 revised model applied to participants in units with
decreased structural hole experiences (e.g., Finance, Marketing, Administration, N = 32)
revealed a nonsignificant chi square (x2 (l) = 1.21, p = .27, SSE = .07). The path
coefficients from the structural hole index to information deficiency and from
information deficiency to job affect were both significant at p < .05 (r = -.34 and r =
-.83). The Accounting and Engineering units experienced no change in overall unit
structural hole scores at T3 and were excluded from the analysis.

In sum, an examination of the full revised model and subgroup analyses of the
revised model was consistent with the data at T1 , T2, and T3. At both T1 and T3, the
relationship between the structural hole index and information deficiency was negative.
However, subgroup analyses revealed that significant negative relationships were present
only in units experiencing a decrease in structural holes experiences. At T2, the
relationship between structural holes and information deficiency was positive, but only
significant for the units which had an increase in structural hole experiences. Although
significance is noted among the structural hole index and information deficiency in many

of the path models above, these results should be interpreted with caution. The large

56

confidence intervals around the coefficients suggest that a larger sample is needed to
insure that each subgroup’s score is statistically different from the others.
Tests of the Longitudinal Model

In keeping with the results of tests of the modified static models above,
employees’ levels of stress was removed from the cross-lagged model. Results of the path
analysis of this modified longitudinal model (see Figure 6) revealed that the model did
not produce a good fit to the data as a chi square test was significant ()8 (23) = 94.15, p =
.00, SSE = 2.22). Subsequent tests of alternative models were unable to produce an
acceptable fit of a cross-lagged model to the data. Additional post-hoe tests of cross-
lagged model were performed by separating the participants into their subgroups based on
increases and decreases of structural hole experiences. In a test of the longitudinal model
presented as Figure 7, the specified model fit neither the increase nor decrease subgroups
at T2 and T3 (x2 (23) = 43.91, p = .005, SSE = 1.71; x2 (23) = 41.96, p = .009, SSE =
3.14, respectively).

Analysis of the relationships among the variables indicated four reasons for the
model’s inability to fit the data. First, outside of the links between structural holes there
were few significant paths between the variables at T1 , T2, and T3. For instance, the
structural hole index at T1 was significantly related (r =.88, p < .001) to the structural
hole index at T2, and the T2 index was significantly related (r =.99, p < .001) to the T3
index. Only the cross-lagged path coefficient between the structural hole index at T2 and
information deficiency at T3 was significant (r = -.26, p < .05). Thus, while hypothesized

significant relationships were evidenced among the structural holes, only structural hole

57

 

 

 

 

 

 

 

 

 

 
 
 
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TIME 1 TIME 2 TIME 3
Structural M 7 Structural 19:122., Structural
Holes Holes Holes

-.I9 (.12) -.26‘(.l 1)
Information -. l 3 (.13) lnforrnation .12 (.1 I) Information
Deficiency Deficiency ——> Deficiency
-.15 (.31) .07 (.40)
Job Affect Job Affect Job Affect
-.23 (.31) -.05 (.40)
__, .____.

 

 

 

 

 

 

 

 

 

 

Note: “*” = p<.05, “**” = p< .001; the standard errors are in parentheses.

Figure 6. Test of the Modified Longitudinal Model

58

 

(a) Units with Increased Structural Hole Experiences

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

  

 

 

 

TIME 1 TIME 2 TIME
.0
Structural M Structural M Structural
Holes Holes Holes
.05 (.15)
Information -.07 (.16) Information Information
Deficiency Deficiency ———9 Deficiency
-.49 ( .72) . I 7 (.36)
Job Afi‘ect Job Affect Job Affect
-.63 (.72) .04 (.36)
: ~——+
(b) Units with Decreased Structural Hole Experiences
TIME I TIME 2 TIME 3
’ .84“I .07 1.04” .01
Structural —(--)-' Structural __(__1 Structural
I-loles Holes Holes
-.55* (.24) - 34‘(.18)
7 Information -.49 (.26) Informal“ .21 (.19) Information
Deficiency Deficiency ———-9 Deficrency
.31 (.24) -.IO (.72)
Job Affect Job Affect Job Affect
.14 (.26) -.07 (.71)
‘———9

 

 

 

 

 

 

 

 

 

Note: “*” = p<.05, “*""’ = p< .001; the standard errors are in parentheses.

Figure 7. Test of the Modified Longitudinal Model For Units With (a) Increased

and (b) Decreased Structural Hole experiences.

 

 

 

 

 

 

59

experiences at T2 predicted information deficiency at T3. Second, the cross-lagged model
was unable to provide a good fit to the data because the least squares approach used in
PATH (Hamilton & Hunter, 1995) was unable to secure a fit with multiple R’s
approximating 1.00 among the structural hole indexes and attitude variables with little or
no correlation.

Third, due to the strong static relationships observed in each time period between
information deficiency and job affect, large amounts of error were produced from the
static paths which were not specified in the cross-lagged model. Fourth, the lack of
autocorrelation among the variables indicates that participants significantly changed their
responses to the scale items between data collections. An examination of the means for
variables under consideration revealed considerable shifts in employees’ perceptions and
attitudes (see Table 6). For example, survivor’s perceptions of information deficiency at
T1 (M = 2.02) increased at T2 (M = 3.53, t (92) = -9.87, p = .0001), but decreased at T3
(M = 1.86, t (95) = 10.75, p = .0001 ). Similar changes were present in employees’ report
ofstress (M = 2.50 at T1 to M = 3.14 at T2, t (85) = -3.87, p = .0001; M = 1.85 at T3, t
(92) = 7.76, p = .0001) andjob affect (M = 3.75 at T1 to M = 2.50 at T2, t (88) = 9.68, p
= .0001; M = 4.15 at T3, t (91) = -1 1.32, p = .0001). The lack ofdiscemible pattern
among cross-lagged variables and shifts among mean scores may indicate that gamma
change occurred in these data as a result of the organizational downsizing
(Golembiewski, Billingsley & Yeager, 1976; Randolph, 1982; Thompson & Hunt, 1996).
With gamma change, subjects perceive scale items differently over time through a

cognitive redefinition of the construct (Thompson & Hunt, 1996).

Chapter 4
Discussion
Overview of the Study

Organizational downsizing is a common practice in business today (Rosenthal,
1996; Uchitelle, 1996), and most downsizing efforts are initiated to restore declining
businesses (Cascio, 1993). Despite downsizing’s overwhelming use in today’s workplace.
few downsizing efforts yield the projected outcomes of increased productivity and
profitability (Cameron & Whetten, 1994). One reason for this failure to attain projected
productivity and profitability may be related to the lack of attention to layoff survivors’
communication patterns and relationships. Examining how networks change in a
downsized organization may provide valuable insight into employee responses to the
upheaval and new environment.

Differences in organizational members’ attitudes toward organizational change
depends in part on how the change impacts their work routine. Among the changes,
employees gain or lose unique network contacts during an organizational downsizing.
The loss of unique information and resources creates structural holes and highlights the
degree of disruption to survivors’ communication networks. While structural holes are
theorized to remain constant in organizational networks (Burt, l992a, 1992b),
organizational change alters the location and impact of structural holes that are embedded
in members’ communication networks. Downsizing survivors who experience increases
in structural holes are hypothesized to perceive that they are receiving less information.

Newly formed gaps in their communication network restrict the flow of information and

60

61

resources. Subsequently, increases in information deficiency should lead to increases in
their level of stress which will negatively impact their affect towards their job.
Additionally, it is hypothesized that in a cross-lagged effect each variable will predict
itself and its static consequent in the subsequent time period. In other words, structural
hole experiences will predict both structural holes and information deficiency at a later
time period. Information deficiency will lead to information deficiency and stress at a
later time and so forth.

In examining the impact of structural holes on employees, this research
investigation examines these hypothesized responses of layofi‘ survivors of an
organizational downsizing over three points in time. Specifically, this examination
explores: (a) the presence of structural holes in an international hotel company’s
corporate headquarters; (b) how structural hole experiences changed following a
downsizing; and (c) the impact of structural hole experiences on layoff survivors’
perceptions and attitudes of their work environment prior to and following an
organizational downsizing. This longitudinal study of employees at corporate
headquarters also considers a number of issues which have hampered downsizing
research. For instance, according to Kozlowski et al. (1993) downsizing research rarely
examines organizational change and employee attitudes over time, relies too heavily on
simulations and lab studies, and does not consider how communication patterns change in
an organization. In addition, this study examines group- and organizational-level
influences on downsizing survivors’ attitudes, and thus considers the impact of macro

variables on survivors’ responses (Rousseau, 1985; Staw & Sutton, 1993). Finally, this

62

investigation extends research on structural holes. Prior structural hole research examines
non-fluctuating network conditions (e.g., Burt, 1992a) and has yet to consider how
structural holes vary under conditions of change.

In order to test the proposed relationships, communication network data and
attitudinal survey measurements (information deficiency, stress, job satisfaction, and
organizational commitment) were collected from a company’s headquarters during the
process of undergoing an organizational downsizing. Pre-downsizing measurements were
taken sixty days prior to the downsizing (T1), and post-downsizing measures were taken
at 60 days (T2) and 150 days (T3) following the downsizing. Communication network
data were converted, standardized, and tested to form a measure of structural holes. The
configuration and impact of structural holes on layoff survivors were examined in each
time period in concert with specific tests of causal relationships within and between each
time period.

Findings and Implications

Five notable findings emerge from this investigation. First, the computation of the
structural hole index reveals considerable variability in network members' structural hole
experiences. For instance, at the initial measurement prior to the downsizing, employees
structural hole index scores range from - 1.50 to 1.52. Ranges for their scores at two and
five months following the downsizing evidence similar dispersion with high structural
hole scores at 1.15 and 1.08, respectively, and low scores at -1.12 and - 1.11,
respectively. Unit level analyses also reveal variability among departments’ structural

hole index scores. At Tl , employees in Marketing, Rooms, and Personnel units

63

experience more structural holes on average than employees in the Accounting, Finance,
Engineering, and Administration units. Further, individual structural hole experiences are
predictive of their future structural hole encounters (rTm = .83 , r1m = .97, and rTm =
.84).

Overall these data suggest that structural hole experiences vary among employees
and each employee's experience tends to remain constant. Employees with high structural
hole scores continue to have more structural hole experiences following downsizing and
remain in disadvantaged network positions. Employees with low structural hole scores
continue to experience fewer structural holes, are well-connected, and are in
comparatively advantageous positions. However, where employees are in the network is
critical. Evidence suggests that employee network position and unit assignment
determine their experience of structural holes. Some units experience more structural
holes than others. Apparently, each work unit sustains varying levels of intra- and inter-
group interaction. For example, the Administration unit is interactive by design as its
function requires input and interaction from the other functional units. The Accounting
unit is more self-contained and requires less inter-group communication. Prior research
tends to omit the work unit context in the report of the structural hole. Burt (1992) and
Ibarra (1995) examine managers' proximity to structural holes without using department
affiliation to assess systemic influence. In this case, aggregating structural hole index
scores by departments provides insights into the variety of network experiences within

units and throughout the organization.

64

Second, in spite of the tendency of structural hole experiences to remain constant,
the distribution of structural holes among individuals and work groups changed with the
downsizing at the corporate headquarters. While the overall mean structural hole index
scores showed little variation (as the standardization of scores produces an overall mean
of zero), at the unit level significant shifts in structural hole experiences were noted. The
Finance unit experienced the greatest post-downsizing increase in structural holes at T2.
In contrast, the Marketing department exhibited decreases in their structural hole
experiences from T1 to T2 and continued to experience decreases into T3, suggesting that
the unit became more insular over time. Accounting and Rooms units also experienced
decreases in structural holes at two months following the downsizing. In short, the unit
level comparisons suggest that individual and work unit were differentially affected by
the downsizing.

Current descriptions depict downsizing as having uniformly negative repercussions
on survivors (Cameron, in press-a; Gutknecht & Keys, 1993; Kleinfield, 1996; Whetten,
et al., 1995). Yet, unit level analyses indicate that employees in some units benefit from a
downsizing. Those network members experiencing a loss in structural holes gain unique
contacts and prominence in a network. Layoff survivors reporting increases in
structural holes either lost unique contacts or became less prominent in their network.
The increase in structural holes suggests that their post-downsizing position in the
organization has become redundant to (or just the same as) others in the network, or they
have become more disconnected in the organization through the loss of unique contacts in

the network. Yet, a number of questions remain about the effects of downsizing on

65

structural hole configurations. While this study indicates that some employees benefited
from the downsizing, it remains unclear if employees in this organization experienced
more or less structural holes than members of different, downsized organizations. For
those who benefited fiom the downsizing, it is unclear to what extent they chose to close
structural holes or the network closed the hole for them. In addition, due to standardizing
index scores, this investigation is unable to determine if the total number of structural
holes in this organization increased or decreased at each point of measurement. Clearly,
future investigations of structural holes should address these concerns.

Third, path analyses revealed that the relationship between structural hole
experiences and perceptions of information deficiency reversed across time. Employees’
structural hole index scores were negatively related to information deficiency at T1 and
T3, the opposite of the predicted relationship. However, at T2 structural hole experiences
were positively related to information deficiency. Other findings from the path analyses
at T1, T2, and T3 indicated that: (a) information deficiency was a more consistent
predictor of affect toward the organization than stress, as an inspection of the stress-affect
and information deficiency-affect relationships revealed considerable variability in the
post-downsizing environment at T2; (b) the static model produced a better fit to the data
when stress was removed from the model; and (c) at each stage of measurement,
information deficiency was negatively related to job affect.

At T1 and T3, structural holes do not lead to information deficiency. It is possible
that employees embedded in established or "stable" networks (characterized by

incremental change) become accustomed to the benefits and limitations of each of their

66

available network contacts. In this case information deficiency is mediated by the
supervisor’s communication behaviors (Miller, et al., 1994). At T1, the pro-downsizing
network had not experienced any major shifts during the fiscal year leading up to the
downsizing. Thus, the only period characterized by “network instability” is the period
directly following the downsizing at T2. With the dismissal of 25 percent of employees,
the network changed its configuration and may have created uncertainty among survivors.
At T2, the experience of structural holes following a downsizing leads to higher
information deficiency and lowers affect toward the job.

Fourth, path analyses of the static model at the unit level revealed stronger
relationships between structural holes and information deficiency in comparison to the
aggregated complete sample. For example, at T2 path analyses of the units experiencing
an overall increase in structural holes demonstrated a significant positive relationship
between the structural hole index and information deficiency (r = .30, p < .05) while the
test of the model with the fiill sample indicated that this relationship was nonsignificant
(r = .18, p > .05). At T3, units reporting a decrease in structural hole experiences
demonstrated a stronger negative relationship with information deficiency (r = -.41, p
<.05) compared to the full sample (r = -.27, p < .05).

Naturally, stronger relationships emerge when separating subjects into high or low
structural hole units. However, these sub-analyses suggest that managers should address
the information concerns of employees in high structural hole units who, after a
downsizing, may be more apt to lose job satisfaction and subsequently voluntarily

turnover (Davy et al., 1991; Tett & Meyer, 1993). At the very least, employees in high

67

structural hole units are more likely to resist the organizational change following a
downsizing. Other applications include examining unit structural hole scores as a means
of explaining why some units are more upset than others and managing change processes
during periods of strategic reorientation (Tushman & Romanelli, 1985; Dewitt, 1993).

Fifth, as noted earlier, experiences of structural holes were predictive of structural
hole experiences in subsequent time periods, Yet, tests of other path relationships
indicated that only structural holes at T2 were significantly predictive of information
deficiency at T3. While respondents' descriptions of their network were consistent over
time and structural hole experiences at T2 decreased information deficiency at T3,
employees' report of information deficiency and job affect were not significantly related
over time. At odds fi'om findings of the static models were the lack of relationships in: (a)
information deficiency between T1 and T2 and between T2 and T3; (b) job affect across
time; and (c) information deficiency with job affect between T1 and T2 as well as T2 and
T3.

These patterns suggest that layoff survivors experience profound shifts in their
perceptions and attitudes over time. These shifts may be a result of the loss or gain of
relationships with coworkers, supervisors, or their jobs themselves. When considering
the lack of discernible linear patterns in the longitudinal analyses, several researchers
point out the connection between second order change and gamma change (Bartunek,
1993, Levy & Merry, 1986; Thompson & Hunt, 1996). Gamma change occurs when
subjects shift their perceptions of the construct and the meaning assigned to it (Randolph,

1982). Uncertainty about work relationships and their career future may lead layoff

68

survivors to gamma shifts in their perceptions and attitudes, may impact the manner in
which the questionnaire items are cognitively defined, and create a difference in their
reported responses from one time period to the next. In effect participants respond
differentially as the survey items develop a different meaning across time. Based on an
approach to identifying gamma change through the examination of the factor structures
over time, in a preliminary sense, these data, exhibit similar characteristicsto gamma
change identified by Bedeian, Armenakis, and Randolph (1988). The factor loadings for
the individual items within each scale displayed noticeable shifts across each time period,
indicating that the responses to the items changed over time (Bedeian, et al., 198 8).

It is interesting to note that gamma change was only evident among the
psychological measurements. Stable longitudinal relationships emerged from analyses of
the structural network variables measured by the structural hole index. While the index
was based on respondents' identification of their work—related contacts, these measures
provided comparatively objective responses. In contrast, psychological-based measures
were subject to gamma change despite evidence of stable factor structures across each
time period.

There are a variety of ways structural holes could have been measured, such as
counting gains or losses of individual links. The statistical method which resulted in a
structural hole index provided a standardized and valid approach to measuring structural
holes. While in some cases the use of the structural hole index revealed few overall
changes in employees’ structural hole circumstance, the shifts are very important.

Namely, increases in structural holes lead to higher levels of stress and information

69

deficiency and lower levels of job affect. Conversely, decreases in structural holes result
in higher levels of job affect and decreases in information deficiency and stress.
Consequently, while the structural hole index provides a conservative measure of
employee network gains and losses, the index shows the impact of downsizing on
employees and has the potential to reveal network gains and losses in other organizational
settings.

In sum, this research investigation contributes to the field of organizational
communication and research on downsizing through its longitudinal approach to
examining communication network structure and its impact on layoff survivors.
Specifically, this investigation extends Burt’s (1992a) concept of structural holes in
networks by coupling validated autonomy and prominence indicators, and it focuses on
the longitudinal impact of structural holes on surviving worker’s experiences of
information deficiency, stress, and affect toward their job. Furthermore, this investigation
bridges the gap between micro- and macro-level processes by considering the
individual’s reaction to changes in network structure. While the investigation of network
structure and worker attitudes is not new, few investigations, examine structural and
perceptual variables under conditions of organizational change. By investigating the
commmrication relationships of workers in terms of structural holes, future studies may
be able to gain an increased understanding of layoff survivors’ reactions to an

organizational downsizing.

70

Limitations

In any organization, both formal and informal network relationships drive the
behavior of network participants (Monge & Eisenberg, 1987; Johnson, 1993). In this
study, respondents were asked to report only their formal work relationships. Yet, as
formal relationships are embedded in the organizational hierarchy they may under-
represent participants’ true network interaction. Workers may exchange useful work-
related communication with others who are not a part of their formal organizational
contacts through “the grapevine” or informal social conversations with members of other
departments (Hellweg, 1987; Johnson, 1993; Stohl & Redding, I987). The data in this
study may also under represent employees network connections as they may be reluctant
to include informal contacts due to fears that their superiors may deem their informal
contacts inappropriate during this major organizational change. The importance of formal
and informal relationships should be made clear in future investigations so that
communication relationships specifically leading to the formation of structural holes and
subsequent changes in perceptions and attitudes can be identified.

The findings and interpretation of this study may also be limited by the use of a
corporate headquarters as a sample in comparison to examining the network of a large
national organization. For instance, the extreme density of this network sample may be an
outcome of investigating the corporate headquarters. Even though individual units within
the headquarters varied in function and duties, it is possible that the corporate office is
uncharacteristically integrated (compared to the rest of the organization) and may not

accurately represent an entire organizational workforce’s reaction to downsizing.

71

Ftu'thermore, a hospitality-based organization may respond differently to downsizing in
comparison to a manufacturing-based organization. While employees in hospitality
organizations may be expected to communicate with employees at the same level and
upper levels in their unit and of other units in order to accommodate the customer,
manufacturing organizations may have formalized and regulated channels for vertical and
horizontal communication. Thus, while hotel employees may have numerous, regular
contacts throughout the organization (especially at the headquarters) and may rebound
quickly to the loss of a few contacts, the loss of a few contacts may be very problematic
in organizations where formal, task interactions within and across units are limited.
Related issues of concern to any network analysis are office configuration and
worker immediacy. Previous research in office design associates privacy in a work
environment with improved work-related satisfaction (Oldham, 1988). However, most
work environments involve considerable interaction with coworkers. In this investigation.
the physical proximity and office configuration of the research participants in their work
environment are not examined. As employees lose contact with a portion of their
coworkers through downsizing, those who sharing a common work space may be affected
more intensely than those coworkers who were physically distant (Hatch, 1987). It is
possible that workers become accustomed to other members in their immediate work
environment, and a level of comfort and security may form. Future research should
examine the impact that physical proximity has on layoff survivors’ reactions to an

organizational downsizing and the formation and experience of structural holes.

72

Network analysis relies on binary socio-metric choice data. Binary network data
does not describe the “quality” of the linkages or relationships reported by network
participants. Members may lose unique sources of information or contacts in the network.
but may not consider them important or valuable. Prominence measures included in the
index adjust for this limitation due to their focus on relationship strength, as opposed to
uniqueness (Burt, 1991). However, the structural hole index primarily identifies
uniqueness of relationships and does not ascertain the quality of the information
exchanged through the unique network connections. Future research should consider
collecting additional information concerning the quality and importance of the reported
linkages from network members to determine which severed linkages are deemed the
most valuable by the participants (Burt, l992a, l992b; Ibarra, 1995).

Due to the extreme density of the network sample, only direct, non-mediated,
linkages of the participants were examined in the network analyses. Each participant had
what can be considered to be an extremely high number of indirect linkages as evidenced
by the small number of step linkages separating network members (M = 4). The plethora
of indirect linkages rendered the communication network data uninterpretable until the
indirect linkages were removed. As a consequence of examining the participants direct
linkages only, Burt’s (1991) prominence measures of choice status, power, and percent
power reflected could not be accurately calculated and were subsequently excluded from
the index.

The analysis of the network was further limited by the exclusion of several

measures of structural autonomy (Burt, 1991, 1992a). The number of contacts were

73

excluded because of their lack of differentiation between redundant and non-redundant
contacts. The measure of proportional density was also excluded from the index because
the binary network data collected here restricted the range of the responses and produced
the same results as for network density. Finally, the measure of oligopoly could not be
calculated without additional survey data not available from the participants at the times
of collection and was excluded from the index. While the confirmatory factor analyses of
the structural hole indicators were based on the remaining network indicators of non-
redundant ties, constraint, network density, contact efficiency, hierarchy, extensive
relations, and exclusive relations, the inclusion of other indicators such as oligopoly.
proportional density, and the number contacts could provide a more precise measure of
structural hole experiences.

Finally, in regard to study design, a number of scholars (Johnson, 1985; Van de
Ven & Poole, 1995; Williams & Podsakoff, 1989) advocate collecting longitudinal data
at intervals appropriate to the issues under study. It is possible that different relationships
between structural holes, information deficiency, stress, and job affect may have existed
at times not observed in this investigation. For example, the sampling frame selected in
this study measured employees’ relationships and reactions in terms of two and five
months post-downsizing. Yet, at two weeks following the downsizing it is possible that
confusion and uncertainty reigns in the immediate wake of the reconfiguration and
evokes an outcry from survivors. While it is possible that a two week sampling frame
may have produced results similar to those in the present study, smaller sampling

intervals prior to and following the downsizing might capture the extremity of network

74

members’ reactions of information deficiency, stress, and job affect. Thus, the rationale
for the sampling frame should be more clearly specified in future research (Williams &
Podsakoff, 1989). Although each field investigation has its tmique needs and limits
(Monge, 1995), the sampling rationale should be carefully addressed in the research design
and grounded in theory.
Future Research

This longitudinal study examines the relationship among structural holes,
information deficiency, stress, and job affect. While tests of these relationships represent
an initial, exploratory investigation, a number of factors may be influencing the variables
under study. For instance, perceptions of support from coworkers, supervisors and the
organization as a whole may shape employee reactions under pressured conditions (Ray
& Miller, 1991; Salancik & Pfeffer, 1978). In times of uncertainty, workers may rely on
support from their coworkers and supervisors to regain a sense of control over their work
domain and aid in their adjustment to their changing work environment (Davy, et al.,
1991). Future research should determine the role of coworker and supervisory support in
the post-downsizing adjustment process. Relatedly, social information from coworkers
contributes to the formation of their attitudes about their work environment (Salancik &
Pfeffer, 1978). The messages received from coworkers during and after the downsizing
may create positive or negative attitudes (Hopfl & Linstead, 1993). Thus future research
should discuss what coworkers discuss and how they characterize events during this

transformation.

75

While the data collected in this study described the diversity of contacts among
employees’ in their network, it does not address the frequency of communication among
contacts. Network members losing frequent contacts (or strong ties) may have
experienced a greater disruption in information flow as fi'equent contacts can be situated
in their immediate work group (Granovetter, 1982; Krackhardt, 1992). Under conditions
of change and uncertainty, employees rely on strong ties for support and comfort
(Krackhardt, 1992). Yet, the loss of less frequent contacts (or weak ties) may also be
limiting. Less frequent contacts often provide information and resources not available in
their immediate work groups (Granovetter, 1973, 1982). It is also unclear how the loss of
perceived high quality contacts shapes network members’ perceptions and attitudes
following downsizing. Future research should consider the impact which contact
frequency and perceived linkage quality has on layoff survivors’ adjustment in a post-
downsized network.

In this investigation, the measurement of stress, job satisfaction, and
organizational commitment did not remain constant over time. Not only did their mean
scores vary over time, but these constructs did not demonstrate predictive validity over
time. For instance information deficiency at T1 did not significantly correlate with
information deficiency at T2. These findings are particularly notable given that they
occurred two months (T2) following the downsizing rather than two days or two weeks
after the organizational event when emotions may have been particularly raw. Since
confirmatory factor analysis indicated that the item measures of these constructs to be

internally consistent and distinct from other constructs in the same time period, future

76

research should examine the impact of potentially catastrophic events on employees.
Explorations of these constructs are particularly important given some of these
constructs’ rich history as having reliable and valid measures (Brooke, Russel, & Price.
1988; Locke & Latham, 1990; Williams & Hazer, 1986) and given the potential for
similar upheavals in other organizations (Rosenthal, 1996; Uchitelle, 1996). Thus. these
variables should be further studied over time in conjunction with organizational change to
determine if the reported results are sample-specific or represent a predictable pattern
which should be associated with layoff survivors and organizational downsizing.

To further clarify the relationships among the variables examined in this
investigation several suggestions are presented. First, the observed static and longitudinal
relationships between information deficiency and stress suggest that a larger underlying
construct may be present in the form of second order unidimensional construct (Hunter &
Gerbing, 1982). As measured, both stress and information deficiency tap into similar
responses of a respondent’s sense of organizational control and uncertainty. Both stress
and information deficiency showed stable relationships with one another over time,
suggesting that some conceptual overlap exists. Based on both of the variables’ important
role in the causal models proposed (and practical importance in dealing with
organizational change), it would be valuable to determine if a more general construct of
organizational uncertainty could better explain layoff survivors’ reactions to the
experience of structural holes following downsizing. Second, through further
simplification of the variables measured, a different causal model may better explain the

data. For example, rather than positing structural holes as the exogenous variable in the

77

causal chain, organizational uncertainty could be the proximate cause of changes to
survivors’ structural hole experiences and job affect in turn. Future investigations of
downsizing and communication relationships should consider alternative causal models

which accurately reflect the participants’ organizational experiences and reactions to

change.

78

References
Allen, N.J., & Meyer, J .P. (1990). The measurement and antecedents of affective,
continuance, and normative commitment to the organization. Journal of Occumtional

Psychology, Q, 1-18.

Bartunek, J .M. (1993). The multiple cognitions and conflicts associated with second
order organizational change. In K. Mumighan (Ed.), Social psychology in organizations
(pp. 322-349). Englewood Cliffs, NJ: Prentice Hall.

Bedeian, A.G., Armenakis, A.A., & Randolph, W.A. (1988). The significance of
congruence coefficients: A comment and statistical test. Journal of Management, 14, 559-
566.

Beer, M., & Walton, R. (1987). Organizational change and development. A_nnfll
Review of Psychology, 3_8_, 339-367.

Berger, C.J., & Cummings, LL. (1979). Organizational structure, attitudes, and
behaviors. Research In Organizational Behavior, 1, 169-208.

Brass, D.J., & Burkhardt, ME. (1992). Centrality and power in organizations. In N.
Nohria and R.G. Eccles (Eds), Networks and organizations, structure, form and action, (pp.
191-215). Boston, MA: Harvard Business School Press.

Brockner, J. (1988). The effects of work layoffs on survivors: Research, theory and
practice. Research in Organizational Behavior, 10, 213-255.

Brockner, J ., Davy, J ., & Carter, C. (1985). Layoffs, self-esteem, and survivor guilt:
Motivational, affective and attitudinal consequences. Organizational Behavior and Human

Decision Processes 36, 229-244.

 

79

Brockner, J ., & Greenberg J. (1990). The impact of layoffs on survivors: An
organizational justice perspective. In J. Carroll (Ed.), Applied social pychology and
organizational settings (pp. 45-75). I-Iilldale, NJ: Erlbaum.

Brockner, J ., Grover, 8., Reed, T., Dewitt, R., & O'Malley, M. (1987). Survivors’
reactions to layoffs: We get by with a little help from our fiiends. Administrative Science
Quarterly, g, 526-541.

Brockner, J., Wiesenfeld, B.M., Reed, T., Grover, S., & Martin, C. (1993).
Interactive effect of job content and context on the reactions of layoff survivors. Journal of
Personalig and Social Psychology, 64, 187-197.

Brooke, P.P., Russel, D.W., & Price, J .L. (1988). Discriminant validation of
measures of job satisfaction, job involvement, and organizational commitment. Journal of

Applied Pachology, 7_3_, 139-145.

Bmton, G.D., Keels, J.K., & Shook, CL. (1996). Downsizing the firm: Answering
the strategic questions. Academy of Management Executive, M2), 38-46.

Burke, R.J. (1984). The closing at Canadian Admiral: Correlates of individual well
being sixteen months after shutdown. Psychological Remus, _5_5,_91-98.

Burke, RJ. (1986). Reemployment on a poorer job after a plant closing,
Psychological Remus, _5_§, 559-570.

Burt, RS. (1991). Structure reference manual version 4.2. Center for the Social
Sciences, Columbia University: New York, NY.

Burt, R.S. (1992a). Structural holes, the social structure of commtition. Harvard

University Press: Cambridge, MA.

8 O
Burt, RS. (1992b). The social structure of competition. In N. Nohria & R.G. Eccles

(Eds), Networks and ogganizations, strugture, form and actiop, (pp. 57-91). Boston, MA:
Harvard Business School Press
Buss, T.F., & Redbum, F .S. (1987). Plant closings: Impacts and responses.

Economic Development Quarterly, _1_, 170-177.

Cameron, K.S. (in press-a). Strategic organizational downsizing: An Extreme

case—Pt. 1. Organizational Dymics.

Cameron, K.S. (in press-b). Strategic organizational downsizing: An Extreme
case—-Pt. 2. Organizational Mnamics.

Cameron, K.S., Freeman, S.J., & Mishra, AK. (1991). Best practices in white collar
downsizing: Managing contradictions. Academy of Mmgement Executive, 5(3), 57-73.

Cameron, K.S., Kim, M.U., & Whetten, DA. (1987). Organizational effects of

decline and turbulence. Administrative Science Merly, 3_2_, 222-240

Cascio, W.F. (1993). Downsizing: What do we know? What have we learned?
Academy of Mmement Executive, 1(1), 95-1 04.

Cobb, S., & Kasl, S.V. (1977). Termination: The consequences of job loss. US.
Department of Health Education and Welfare. Washington, DC: Government Printing
Office.

Cooper, C.L. & Payne, R. (1988). Causes, c0ping, and con_sequences of _stress_a_t

work. Chichester, England: Wiley.

 

Cummings, T.G., & Huse, E.F. (1989). Organizational development and change.

New York: West Publishing Co.

8 1
Davidson, M.J., & Cooper, CL. (1986). Executive women under pressure.
Occupational and life stress and the family [Special issue]. International Review of

Applied Psychology, 3_5_, 301-326.

Davy, J.A., Kinicki, A.J., & Scheck, CL. (1991). Developing and testing a model of

survivor responses to layoffs. Journal of Vocational Behavior 38, 302-317.

 

Dooley, D., & Catalano, R. (1988). Recent research on the psychological effects of

UE. Journal of Social Issues fl, 1-12.

 

Fisher, C.D., Schoenfeldt, L.F., & Shaw, J .B. (1990). Human resource management.
Boston, MA: Houghton Mifflin Company.

Freeman, S.J., & Cameron, KS. (1993). Organizational downsizing: A convergence
and reorientation fi‘arnework. Organization Science, 4, 10-29.

Fried, Y., Rowland, K.M., & Fenis, GR. (1984). The physiological measurement of
work stress: A critique. Personnel Psychology, 17, 583-615.

Golembiewski, R.T., Billingsley, K., & Yeager, S. (1976). Measuring change and
persistence in human affairs: Types of change generated by OD designs. Journal of

Applied Behavioral Science, _1_2, 133-157.

Goodman, P.S, Bazerman, M., & Conlon, E. (1980). Institutionalization of planned

change. Research in Organizational Behavior, 2, 215-246.
Gottlieb, M.R., & Conkling, L. (1995). Managing the workplace survivors:

Organizational downsizing and the commitment gap. Westport, CT: Greenwood.

Granovetter, MS. (1973). The strength of weak ties. American Journal of

Sociology, fl, 1360-1380.

82

Granovetter, MS. (1982). The strength of weak ties: A network theory revisited. In
P.V. Marsden and Nan Lin (Eds), Social structure and network analysis (pp. 105-130).
Beverly Hills, CA: Sage.

Greenhalgh, L. (1983). Organizational decline. In S.B. Bacharach (Ed.), Research in
the Sociology of Organizations, 2, 231-276.

Gutknecht, J.B., & Keys, J.B. (1993). Mergers, acquisitions and takeovers:
Maintaining morale of survivors and protecting employees. Academy of Management

Executive 7(2 1, 26-36.

 

Hackman, J .R., & Oldham, GR. (1975). Development of the job diagnostic survey.
Journal of Applied Psychology, Q0, 159-170.

Hackman, J .R., & Oldham, GR. (1980). Work redesigp. Reading, MA: Addison-
Wesley.

Harnilton, M.A., & Hunter, J .E. (1995). Pathe: A prom in basica. Unpublished
manuscript, University of Connecticut, Stoors.

Hartman, R.L., & Johnson, J .D. (1989). Social contagion and multiplexity:
Communication networks as predictors of commitment and role ambiguity. Human
Communication Research, 15, 523-548.

Hatch, M.J. (1987). Physical barriers, task characteristics, and interaction activity in
research and development firms. Administrative Science Quarterly, 12, 387-399.

Hellweg, SA. (1987). Organizational grapevines. In B. Dervin & M.J. Voight
(Eds), Proggess in communication sciences, Vol. 8 (pp. 213-230). Norwood, NJ: Ablex

Publishing Corp.

8 3
prfl, H., & Linstead, S. (1993). Passion and performance: Sufi‘ering and the
carrying of organizational roles. In S. Fineman (Ed.), Emotion in organizations. (pp. 76-93).
Newbury Park, CA: Sage.
Howell, R.D., Bellenger, D.N., & Wilcox, J .B. (1987). Self-esteem, role stress, and

job satisfaction among marketing managers. Journal of Business Research 1;, 71-84.

 

Huber, G.P., & Daft, R.L. (1987). The information environments of organizations.
In F.M. Jablin, L.L. Putnam, K.H. Roberts & L.W. Porter (Eds), Handbook of
Organizational Communication (pp. 130-164). Newbm'y Park: Sage Publications

Hunter, J. E., & Gerbing, D.W. (1982). Unidirnensional measurement, second order
factor analysis, and causal models. Research in Organizational Behavior, 4, 267-320.

Iaffaldano, M.T., & Muchinsky, PM. (1985). Job satisfaction and job performance:
A meta-analysis. Psychological Bulletin, 22, 251-273.

Ibarra, H. (1993). Network centrality, power, and innovation involvement:
Determinants of technical and administrative roles. The Academy of Management Journal,
36, 471-501.

Ibarra, H. (1995). Race opportunity and diversity of social circles in managerial
networks. Academy of Management Journal, 33, 673-703.

Indik, B.P. (1965). Organization size and member participation: Some empirical

tests of alternative explanations. Human Relations 1_8_, 339-350.

 

Ivancevich, J., Matteson, M., & Preston, C. (1982). Occupational stress, Type A

behavior, and physical well-being. Academy of Management Journal, 25, 373-391.

84

Jablin, F.M. (1987). Formal organization structure. In F.M. Jablin, L.L. Putnam,
K.H. Roberts & L.W. Porter (Eds), Handbook of organizational communication (pp. 389-
419). Newbury Park, CA: Sage Publications.

Jackson, SE. (1983). Participation in decision making as a strategy for reducing job
related strain. Journal of Applied Psychology, 68, 3-19.

Johnson J .D. (1993). Organizational communication structure. Norwood, NJ: Ablex
Publishing Corp.

Johnson, J.R., Bemhagen, M.J., Miller, V.D., & Allen, M. (1996). The role of
communication in managing reductions in work force. The Journal of Applied

Communication Research A1654 80.

 

Kahn, RL., Wolfe, D.M., Quinn, R.P., & Snoek, JD. (1964). Organizational stress:

studies in role conflict and arnbigpity. New York: Wiley.
Kahn, R.L., & Byosiere, P. (1992). Stress in organizations. In M.D. Dunnette &

L.M. Hough (Eds), The Handbook of industrial and organizational psychology, Vol. 3.
(2nd ed., p. 571-649). Palo Alto, CA: Consulting Psychologists Press.

Kinicki, A.J. (1985). Personal consequences of plant closings: A model and

preliminary test. Human Relations 3_8,_197-212.

 

Kinicki, A.J., & Latack, J .C. (1990). Explication of the construct of coping with

involuntary job loss. Journal of Vocational Behavior &, 339-360.

 

Kleinfield, NR. (1996, March 4). The company as family, no more. The New York

Times, pp. A1, A12-A14.

85

Kozlowski, S.W.J., Chao, G.T., Smith, E., & Hedlund, J. (1993). Organimtional
downsizing: Strategies, interventions, and research implications. In C.L. Copper & I..T.
Robertson (Eds), International review of industrial and organizational psychology, Vol. 8.
(pp. 264-322). New York: John Wiley and Sons Ltd.

Krackhardt, D. (1992). The strength of strong ties: The importance of philos in
organizations. In N. Nohria & R.G. Eccles (Eds), Networks and organizations, structure,
form and actioa, (pp. 216-239). Boston, MA: Harvard Business School Press.

Krackhardt, D., & Porter, L.W. (1985). When fi'iends leave: A structural analysis of
the relationship between turnover and stayers’ attitudes. Administrative Science Quarterly,
fl, 242-261.

Krackhardt, D., & Porter, L.W. (1986). The snowball effect: Turnover embedded in
communication networks. Journal of Applied Psychology, 7_1, 50-55.

Larkey, L.K. (1993). Perceptions of discrimination during downsizing.
Management Communication Quarterly, 1, 158-180.

Lawler, BE. (1986). High-involvement management. San Francisco, CA: Jossey-
Bass Publishers.

Levy, A. (1986). Second-order planned change: Definition and conceptualization.

Organizational Dypamics, £440-

Levy, A., & Merry, U. (1986). Organizational transformation: Approaches,

strategies, theories. New York: Praeger.
Lewis, L.K., & Seibold, DR. (1993). Innovation modification during

intraorganizational adoption. Academy of Management Review, 18, 322—3 54.

86

Lewis, L.K., & Seibold, DR. (1996). Communication during intraorganizational
innovation adoption: Predicting users' behavioral coping responses to innovations in
organizations. Communication Monographs, 83, 131-157.

Liem, R., & Liem, J .H. (1988). Psychological effects of unemployment on workers

and their families. Journal of Social Issues fl, 87-105.

 

Locke, E.A., (1976). Nature and causes of job satisfaction. In M.D. Dunnette (Ed.).
Handbook of industrial and organizational psychology (pp. 1297-1 350). Chicago, IL: Rand
Mcnally.

Locke, E.A., & Latham, GP. (1990). Work motivation and satisfaction: Light at the

end of the tunnel. Psychological Science, 1, 240-246.

Mathieu, J .E., & Zajac, D.M. (1990). A review and meta-analysis of the
antecedents, correlates, and consequences of organizational commitment. Psychological

Bulletin fl, 171-194.

 

McKinley, W., Sanchez, C.M., & Schick, AG. (1995). Organizational downsizing,
constraining, cloning, learning. The Acadgmy of Mmement Executive, 2Q), 32-44.

Meyer, J .P., Allen, N.J., & Gellatly, LR. (1990). Affective and continuance
commitment to the organization: Evaluation of measures and analysis of concurrent and
time-lagged relations. Journal of Applied Psychology, E, 710-720.

Meyer, J .P., Allen, N.J., & Smith, GA. (1993). Commitment to organizations:
Extension and test of a three-component conceptualization. Journal of Applied Psychology,

18, 538-551.

87

Meyer, J.P., Paunonen, S.V., Gellatly, I.R., Goffin, R.D., & Jackson, D.N. (1989).

Organizational commitment and job performance: It’s the nature of the commitment that

counts. Journal of Applied Psychology, 14, 152-156.

Meyer, ME. (1996) The effects of weak ties on mrceived organizational

innovativeness and innovation characteristics. Unpublished doctoral dissertation, Michigan
State University, East Lansing.

Miles, RH. (1980). Macro organizational behavior. Santa Monica, CA: Goodyear
Publishing.

Miller, V.D., Johnson, J.R., & Grau, J. (1994). Antecedents of willingness to

participate in a planned organizational change. Journal of Applied Communication

Research, Q, 59-80.

Monge, RR. (1995). Theoretical and analytical issues in studying organizational
processes. In G.P. Huber & A.H. Van de Ven (Eds), Longitudinal field research methods:
Studying processes of organizational change. (pp. 267-298). Thousand Oaks, CA: Sage.

Monge, P.R., & Contractor, NS. (1987). Communication networks: Measurement
techniques In C.H. Tardy (Ed.), A handbook for the stjgv of human communication (pp.
107-138). Norwood, NJ: Ablex Publishing.

Monge, P.R., & Eisenberg, EM. (1987). Emergent communication networks. In
F.M. Jablin, L.L. Putnam, K.H. Roberts & L.W. Porter (Eds), fiandboolgof organizational
communication (pp. 304-342). Newbury Park: Sage Publications.

Mowday R.T., Steers, R.M., & Porter, L.W. (1979). The measurement of

organizational commitment. Journal of Vocational Behavior 14 224-247.

,—,

88

Noe, RA. (1986). Trainees’ attributes and attitudes: Neglected influences on
training effectiveness. Acadgmy of Management Review, 81, 736-749.

Oldham, GR. (1988). Effects of changes in workspace partitions and spatial density
on employee reactions: A quasi-experiment. Journal of Applied Psychology, 13, 253-258.

O’Neill, H.M., & Lenn, DJ. (1995). Voices of survivors: Words that downsizing
CEOs should hear. Academy of Mmement Executive, 2(4), 23-33.

Podgursky, M., & Swaim, P. (1987). Job displacement and eamings loss: Evidence

 

from the displaced worker survey. Industrial and Labor Relations Review 1L, 17-29

Porras, J .I., & Robertson, P.J. (1992). Organizational development: Theory, practice
and research. In M.D. Dunnette & L.M. Hough (Eds), Handb*ook of industrial and
organizational psychology, Vol. 3. (2nd ed., pp. 719-822). Palo Alto, CA: Consulting
Psychologists Press.

Ray, E.B., & Miller, KL (1991). The influence of communication structure and
social support on job stress and burnout. Management Communication Quarterly, 3, 506-
527.

Randolph, W.A. (1982). Planned organizational change and its measurement.
Personnel Psychology, _3_5_, 1 17-139.

Richards, W.D., & Rice, RE. (1985). NEGOPY network analysis program. 8%

Networks 3, 215-223.

 

Rogers, EM. (1983). Diffusion of innovations. New York: Free Press.

Rousseau, D.M. (1985). Issues of level in organizational research: Multi-level and

cross-level perspectives. Research in Organizational Behavior, 1, 1-38.

89

Rosenthal, E. (1996, August, 19). Once in big demand, nurses are targets for

hospital cuts. The New York Times pp. B1, B4.

 

Salancik, G.R., & Pfeffer, J. (1978). A social information processing approach to
job attitudes and task design. Administrative Science Quarterly, 28, 224-253.

Staw, B.M., & Sutton, RI. (1993). Macro organizational psychology. In K.
Mumighan (Ed.), Social psychology in ogganizations (pp. 350-384). Englewood Cliffs. NJ:
Prentice Hall.

Stohl, C., & Redding, WC. (1987). Messages and message exchange processes. In
F .M. Jablin, L.L. Putnam, K.H. Roberts, & L.W. Porter (Eds), Handbook of organizational
communication (pp. 451-502). Newbury Park: Sage Publications.

Susskind, AM. (1996). Structural holes: An expansion of Burt’s arguments.
Unpublished manuscript, Michigan State University, East Lansing.

Sutton, R.I., & D'Aunno, TD. (1989). Decreasing organizational size: Untangling
the effects of money and people. Academy of Management Review, E, 194-212.

Tett, R.P., & Meyer, J .P. (1993). Job satisfaction, organizational commitment,
turnover intention, and turnover: Path analyses based on meta-analytic findings. Personnel
Psychology, fl, 259-291.

Thompson, R.C., & Hunt, J .G. (1996). Inside the black box of alpha, beta, and

gamma change: Using a cognitive-processing model to assess attitude structure. Academy

of Management Review, 81, 655-690.

90

Tushman, M.L., & Romanelli, E. (1985). Organizational evolution: A
metamorphosis model of convergence and reorientation. Re§garch in Qrggizational

Behavior 2, 171-222.

 

Uchitelle, L. (1996, August 23). Despite drop, layoffs remain high. The New York
Times, pp. A1, D2-28.
Uchitelle, L., & Klienfield, NR. (1996, March 3). On the battlefields of business,

millions of casualties. The New York Times pp. 1, 26-28.

 

Van de Ven, A.H., & Poole, MS. (1995). Methods for studying innovation
development in the Minnesota Innovation Research Program. In G.P. Huber & A.H. Van de

Ven (Eds), Longitudinal field research methods: Studying processes of organizational

gang; (pp. 155-185). Thousand Oaks, CA: Sage.

von Bertalanffy, L. (1968). Systems theog: Foundations, developments,
applications. New York: Braziller.

Vredenburgh, D.J., & Trinkaus, R.J. (1983) An analysis of role stress among

hospital nurses. Journal of Vocational Behavior, 2_2_, 82-95.

Warr, P., & Jackson, P. (1984). Men without jobs: Some correlates of age and
length of unemployment. Joumal of Occupational Psychology, 3], 77-85.

Weick, K.E. (1979). The social psychology of oganizations. New York: Random
House.

Weick, K.E. (1993). Sensemaking in organizations: Small structures with large

consequences. In K. Mumighan (Ed.), Social Psychology in Organizations (pp. 10-37)

Englewood Cliffs, NJ: Prentice Hall.

91

Whetten, D. A., & Cameron, KS. (1994). Organizational-level productivity
initiatives: The case of downsizing. In D.H. Hanis (Ed.), Organizational Ewes:
Understanding the productivig paradox (pp. 262-290). Washington, DC: National
Academy Press.

Whetten, D.A., Keiser, J .D., & Urban, T. (1995). Implications of organizational
downsizing for the human resource management firnction. In G.R. Ferris, S.D. Rosen, &
D.T. Barnum (Eds), Handbook of human resource mgagement. (pp. 282-296). Cambridge,
MA: Blackwell Publishers.

Williams, L.J., & Podsakoff, RM. (1989). Longitudinal field methods for
studying reciprocal relationships in organizational behavior research: Toward improved
causal analysis. Research in Organizational Behavior, _1_1_, 247-292.

Williams, L.J., & Hazer, J .T. (1986). Antecedents and consequences of
satisfaction and commitment in turnover models: A reanalysis using latent variable
structural equation methods. Journal of Applied Psychology, 78, 219-231.

Zammuto, R.F., & Cameron, KS. (1985). Environmental decline and organizational

response. Research in Organizational Behavior, L 223-262.

HICHIGAN STATE UNIV. LIBRARIES
lllllllllllllllllllllllllllllll"llllllllllllllllllllllllllHI
31293015649894