. .
a . .... o
luau. .

z. .. . . ., 3....
1... 2 .. ,
.6... f um... ,

“57:0... ti
”3.x 1......- ‘
t-J:l »

(.5... :5 I

.3.

 

 

 

C A
24.7
a.

v.24"!

:82. 13...?

aa—
3.5.

. swam”...
Wflflmfihwz

. fl]... .31
-. . I
if A 5.3.5.1..

 

LIBRARY
Micnlgan State
University

 

 

This is to certify that the
dissertation entitled

FAMILY OWNED BUSINESSES: AN EXAMINATION OF
STRUCTURE, FAMILY DYNAMICS AND VALUES

presented by

BRIAN JOHN DISTELBERG

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Family and Child Ecology

 

 

lire.

Major Professor’s Signature
.24 LE: 17,007

Date

 

MSU IS an Affirmative Action/Equal Opportunity Employer

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KzlProleccsprelelRC/DateDuerindd

 

 

FAMILY OWNED BUSINESSES: AN EXAMINATION OF
STRUCTURE, FAMILY DYNAMICS AND VALUES

By

Brian John Distelberg

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Family and Child Iicology

2009

 

business system. Employees tended to believe that their FOB favored the family system
while family members tended to believe that the FOB favored the business system.
FOBS were able to unify this perception across the owners, family members and
nonfamily employees when they allowed information to flow through a permeable
boundary between the family and business systems. FOBs that did not allow information
to flow from the family to the business had dissenting opinions between family members
and employees and significantly lower levels of satisfaction through the FOB system.
Conclusions from this study point to the need to use in depth sampling procedures and
include family dynamics, value orientations. and family to business boundary

measurements when study FOBs.

ACKNOWLEDGEMENTS

Many people are responsible for my ability to complete this dissertation and my
degree. Some have been directly involved in the research process while others have
indirectly helped through their support and care. Each one of you has played an
important and immeasurable role in my life and this project.

First, to all of my committee members. I want to thank you for all the time you
have put into this study and my education. I am especially grateful for your willingness
to help me explore family businesses. even though we were all somewhat novices in this
area at the beginning. you all allowed me to take a risk in exploring this area. Many times
you were able to bring my interests out of broad and somewhat foggy conceptual areas
into scientific and testable questions. For that I am indebted to you.

T 0 Ellen Kossek. you have been a great inspiration to me. Your tireless efforts in
the field are amazing. Even more impressive is your willingness to take on and mentor
new students. You were my first bridge from the field of family studies to the field
organizational behavior. Without your help and mentoring I would not have taken on this
study.

To Barbara Ames, your leadership and guidance throughout my time at Michigan
State University will not be forgotten. I am grateful to you for all of your advice on the
process and steps involved in this dissertation. Thank you.

To Marsha Carolan, it is difficult to imagine a more empathic teacher and mentor.
From my entrance into the doctoral program to the conclusion of this dissertation you
have always expressed a genuine interest and offered the highest level of support. You

have been a great mentor to me. and also a loving friend to me and my family.

To Adrian Blow, my committee chair. I am indebted to you for all you have done
for me. Form the very beginning you pushed me to be more than what I thought I could
be. I am extremely grateful for your willingness, and encouragement to break out of the
standard pathway of a doctoral program and carve a new path which was unique. a great
adventure, and one that has forged my professional identity. Every step ofthe way you
have believed in me, and been there to pick up the pieces when things fell apart. I am
honored to call you my committee chair, mentor and friend.

To the Family Business Alliance. and especially Mary Novello, there is no way I
would have attempted this study without your help. Your willingness to provide
resources and connections played a valuable role in this study, the importance of which
you may never know. The time I spent with you has greatly informed not only my
research but who I am and my passion for working family businesses.

' To the family businesses that took part in this study. thank you for letting me
come into your worlds, businesses and even your families. I truly appreciate how
difficult and personal that was. I have learned so much from our relationships. You have
continuously amazed me with your desire to see me succeed and your willingness to
extend your own resources to see it happen, this is what truly sets family businesses apart
from other businesses. It is my hope that we will continue these relationships.

To my sister Karen Canniff and by brother .Ieff Distelbcrg. as well as your
families. you have been a tremendous support to me and our family during the last year.
Your interest and willingness to hear me vent the stress of a dissertation was extremely

helpful. I am truly blessed by you both.

To Don and Natalie, my parents. you have done so much for me over the years.
No one could ask for better parents. You are both role models to me. Thank you for
seeing potential in me long before anyone else. Mom your love and kindness made me
feel like I could do anything. Dad your character. and sacrifice for our family has shown
me what it means to be a father.
To my children, Alina and Gideon. you are wonderful. Thank you for
understanding when I had to take time away to write. And thank you for always
I‘erninding n‘le what is important in life.
To my wife Michelle, you do it all. and you always do it better than anyone else.
I Can’t imagine life without you. As we start a new adventure I have no worries because I
kllow that when I‘m with you we can do anything. I love you.
Final 1y to my friends and colleagues in the Family and Child Ecology
D Qpaftme ' s . ' 3 ~ -‘ . ,
nt, the time we have spent together was priceless. I am proud to say that I am a
n1 ember of tljis family.

Than k you!

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................... x
LIST OF FIGURES .......................................................................................................... xii
CHAPTER I: OVERVIEW ................................................................................................. 1
Introduction .............................................................................................................. 1
Theoretical Frame Limitations .................................................................... 2

Methodological Limitations ......................................................................... 3

Statement of the Problem ......................................................................................... 5
Purpose Statement ................................................................................................... 8
Specific Aims ......................................................................................................... 10
Theory Development ............................................................................................. 11
General System Theory ............................................................................. 11

Three Circle Model .................................................................................... l3

Developmental Model for Family Businesses (DMFB) ............................ 15

Adaptation and Cohesion ........................................................................... 16

Value Orientation ....................................................................................... 18

Conceptual Model ...................................................................................... 20

Research Questions, Hypotheses and Variable Definitions .................................. 23
Specific Questions and Hypotheses ........................................................... 23

Variable Definitions ................................................................................... 31

Q HAPTER II: LITERATURE REVIEW .......................................................................... 34
Introduction ............................................................................................................ 34
Defilling a Sample Population ............................................................................... 36
satisfaction in FOBs .............................................................................................. 38
Strucmre in F085 .................................................................................................. 4o
Overlap Between the Family and the Business Systems ........................... 41

Overlap of Family, Business and Ownership systems ............................... 43

Boundaries and Cohesion within FOB Systems ........................................ 45

values .................................................................................................................... 49

vii

Agreement on Values ................................................................................ 51

Adaptability ........................................................................................................... 52
Other Variables to Consider .................................................................................. 55
Firm Size .................................................................................................... 55

Gender ........................................................................................................ 57

Industry ...................................................................................................... 58
Geographic Location ................................................................................. 59

Age of firm ................................................................................................ 60
Summary ................................................................................................................ 61
CHAPTER III: METHODOLOGY ................................................................................... 63
Introduction ............................................................................................................ 63
Methods ................................................................................................................. 63
Sampling Procedures ................................................................................. 63

Data Collection .......................................................................................... 65

Data Imputation ......................................................................................... 67
Ownership Validation ................................................................................ 68

Study Participants: Individuals .................................................................. 68

Study Participants: Businesses .................................................................. 70

Measures .................................................................................................... 71
Dependent Variables .............................................................................................. 77
Independent Variables ........................................................................................... 82
Data Analysis Procedures ...................................................................................... 96
QHAPTER Iv: RESULTS ................................................................................................ 97
Phase 1: Step 1 ....................................................................................................... 97
Phase I: Step 2 ....................................................................................................... 99
Company Summaries ............................................................................... 103
Discussion of Research Questions ........................................................... 160

Summary of Phase 1: Step 2 .................................................................... 166

Phase 2: Step 1 ..................................................................................................... I67
Reliability for Phase 2: Step 1 Model ...................................................... 176

Phase 2: Step 2 ..................................................................................................... 179

viii

Reliability for Phase 2: Step 2 Model ...................................................... 186

Summary of Findings .......................................................................................... 191
CHAPTER V: DISCUSSION .......................................................................................... 193
Introduction .......................................................................................................... 193
Discussion of Results ........................................................................................... I94
Evaluating the Three Circle Model .......................................................... 194

Expanded Three Circle Model ................................................................. 196

Discussion of Methods: Limitations .................................................................... 209
Discussion of Methods: Strengths ....................................................................... 211
Implications for Family Owned Businesses ........................................................ 213
Implications for Future Research ......................................................................... 216
Implications for Systemic Clinical Interventions ................................................ 218
Concluding Remarks ........................................................................................... 221
APPENDICES ................................................................................................................. 223
APPENDIX A: Gate Keeper Interview Guide .................................................... 224
APPENDIX B: FACES III .................................................................................. 225
APPENDIX C: Family Member Survey .............................................................. 227
APPENDIX D: Participant Survey ...................................................................... 229
APPENDIX E: Informed Consent ....................................................................... 232
APPENDIX F: Additional Sociograms ............................................................... 237
I{flatterences - - .. ................................................................................................................... 260

LIST OF TABLES

Table 3.1: Participants by Subgroup ......................................................................... 69
Table 3.2: FOB Participant Demographics .............................................................. 71
Table 3.3: Value Continuum Items ........................................................................... 72
Table 3.4: FACE III .................................................................................................... 76
Table 3.5: Value Continuum Reliability .................................................................. 78
Table 3.7: Value Continuum Descriptive Statistics ................................................ 79
Table 3.8: Satisfaction Scale Items ........................................................................... 80
Table 3.10: Satisfaction Descriptive Statistics ........................................................ 81
Table 3.12: Cohesion Descriptive Statistics ............................................................ 83
Table 3.14: Adaptability Descriptive Statistics ....................................................... 85
Table 3.15: Family Access Descriptive Statistics ................................................... 89
Table 3.17: Ownership Access Descriptive Statistics ............................................ 91
Table 3.18: Ownership Access Histogram ............................................................... 91
Table 3.19: Employee Access Descriptive Statistics .............................................. 92
Table 3.21: Total Access Descriptive Statistics ...................................................... 94
Talale 4.1 Three Circle Model Fit with Total Communication Matrix as
Dependent ..................................................................................................................... 99
Talale 4.2.3 Company 1 Summary. .......................................................................... 108
Table 4.2.6: Company 2 Summary .......................................................................... 114
Table 4.2.9 Company 3 Summary Table ................................................................ 119

Table 4.2.12 Company 4 Summary Table .............................................................. 124

Table 4.2.15 Company 5 Summary ......................................................................... 129
Table 4.2.18 Company 6 Summary ......................................................................... 134
Table 4.2.21 Company 7 Summary ......................................................................... 139
Table 4.2.22 Company 8: Total Communication .................................................. 141
Table 4.2.23 Company 8: Family Communication ............................................... 142
Table 4.2.24 Company 8 Summary Table .............................................................. 144
Table 4.2.25 Company 9: Total Communication .................................................. 146
Table 4.2.26 Company 9: Family Communication ............................................... 147
Table 4.2.27 Company 9 Summary ......................................................................... 149
Table 4.2.28: Company 10: Total Communication ............................................... 151
Table 4.2.29 Company 10: Family Communication ............................................. 152
Table 4.2.30: Company 10 Summary Table .......................................................... 154
Table 4.2.31 Company 11: Total Communication ................................................ 156
Table 4.2.32 Company 11: Family Communication ............................................. 157
Table 4.2.33: Company 11: Summary .................................................................... 159
Table 4.3.1: First Model Summary ......................................................................... 176
Table 4.4.1: Summary of Second Model ................................................................ 186

xi

LIST OF FIGURES

Figure 1.1: Three Circle Model ........................................................................... 14
Figure 3.6: Value Continuum Histogram ............................................................ 79
Figure 3.9: Satisfaction Histogram ..................................................................... 81
Figure 3.11: Cohesion Histogram ....................................................................... 83
Figure 3.13: Adaptability Histogram .................................................................. 85
Figure 3.16: Family Access Histogram ............................................................... 90
Figure 3.20: Employee Access Histogram .......................................................... 93
Figure 3.22: Total Access Histogram .................................................................. 95
Figure 4.2.1: Company 1: Total Communication ............................................. 104
Figure 4.2.2: Company 1: Family Communication .......................................... 105
Figure 4.2.4: Company 2: Total Communication ............................................. 111
Figure 4.2.5: Company 2: Family Communication .......................................... 112
Figure 4.2.7: Company 3: Total Communication ............................................. 116
Figure 4.2.8: Company 3: Family Communication .......................................... 117
Figure 4.2.10 Company 4: Total Communication ............................................ 121
Figure 4.2.11 Company 4: Family Communication ......................................... 122
Figure 4.2.13 Company 5: Total Communication ............................................ 126
Figure 4.2.14 Company 5 Family Communication ........................................... 127
Figure 4.2.16 Company 6: Total Communication ............................................ 131
Figure 4.2.17 Company 6: Family Communication ......................................... I32

xii

Figure 4.2.19 Company 7: Total Communication ............................................ 136

Figure 4.2.20 Company 7: Family Communication ......................................... 137
Figure 4.2.34 Satisfaction and Value Orientation ............................................. 161
Figure 4.2.35 Cohesion and Satisfaction ........................................................... 164
Figure 4.3.2: Box Plot of Residuals by 11 FOBs .............................................. 177
Figure 4.3.3: Scatterplot of level 1 residuals against fitted values .................... 178
Figure 4.3.4: P-P plot of the level 1 residuals ................................................... 179
Figure 4.4.2: Box Plot of Residuals by each of the 11 FOBs ........................... 188 .
Figure 4.4.3: Scatterplot of level 1 residuals against fitted values .................... 189
Figure 4.4.4: P-P plot of level 1 residuals ......................................................... 190
Figure 6.1 Company 1: Employee Communication .......................................... 237
Figure 6.2: Company 1: Ownership Communication ....................................... 238
Figure 6.3: Company 2: Employee Communication ........................................ 239
Figure 6.4: Company 2: Owner Communication .............................................. 240
Figure 6.5: Company 3: Employee Communication ........................................ 241
Figure 6.6: Company 3: Owner Communication .............................................. 242
Figure 6.7: Company 4: Employee Communication ........................................ 243
Figure 6.8: Company 4: Owner Communication .............................................. 244
Figllre 6.9: Company 5: Employee Communication ........................................ 245
Figllre 6.10: Company 5: Owner Communication ............................................ 246
Figlue 6.11: Company 6: Employee Communication ...................................... 247
Figure 6.12: Company 6: Owner Communication ............................................ 248
Figure 6.13: Company 7: Employee Communication ...................................... 249

xiii

Figure 6.14:
Figure 6.15:
Figure 6.16:
Figure 6.17:
Figure 6.18:
Figure 6.19:
Figure 6.20:
Figure 6.21:

Figure 6.22:

Company 7: Owner Communication ............................................ 250

Company 8: Employee Communication ...................................... 251
Company 8: Owner Communication ............................................ 252
Company 9: Employee Communication ...................................... 253
Company 9: Owner Communication ............................................ 254
Company 10: Employee Communication .................................... 255
Company 10: Owner Communication .......................................... 256
Company 11: Employee Communication .................................... 257
Company 11: Owner Communication .......................................... 258

xiv

CHAPTER I: OVERVIEW
Introduction

Family owned businesses (FOBs) are a cornerstone to the world‘s economy.
Some estimates suggest that as many as 89% ofall businesses in North America are
family owned. These same sources attribute 64% of the US GDP to F OBS (Astrachan &
Sllanker, 2003). Additionally, depending on how one defines “family owned”, FOBs
€111 ploy somewhere between 27% and 62% ofall U.S. employees (Astrakhan & Shanker.
2 () 03; U.S. Census. 2007). Due to the prevalence of FOBs. many social. family, and
o I‘ganizational researchers and theorists began studying them in the 19805. The
c: QnClusion from these early studies showed that FOBs have numerous strengths which
1L1 '31P them outperform nonfamily businesses. but these businesses also struggle to
1‘ baintain the complex balance between the business and the family (Aronoff. Ward &
{\strachan, 2002; Ward, 1987).
A review ofthe literature on FOBs since the early 19803 shows that: 1) F085 are
t\ \lade up Of three interdependent systems (the familv. business and ownership systems)
(if (36FSICIL Dam/is. I-Iampton & Lansberg. 1997; Shanna, 2004; Stafford, Danes. Duncan &
\Ninters. 1 999; Taguiri & Davis, 1982); 2) Successful FOBs begin with successful
8 ani118 film ilies, and specifically, owning families that are flexible and unified (Davis &
& terns. 198 1 ; Galvin, Astrachan & Green, 2007; Zody. Sprenklt‘. MacDermid. &
& chrank, 2006); 3) F OBS vary in why they exist. with some existing to support short term

f 21 111111" 80318, and others supporting long term business growth (Dean. 1992; Distelberg &

Sorensen, 2009; Shanna, Chrisman & Chua. I997; Sharma & Nordqvist. 2008', Wong.
McReynolds, & Wong. 1992).

When these three issues are combined, as they always are in F OBs, they create a
very complex system, with as much variance across individual FOBS as there is between
non-family owned businesses and FOBS. This is especially true when the outcome
variables of interest include survival over time (Jorissen, Laveren. Martens, Reheul.

2 005), ownership structures employed (Anderson, Mansi & Reeb, 2003; Daily &
Dollinger, 1992: Sonfield & Lussier. 2005). business performance (Chrisman. Chua &
L i tZ, 2003), retention of nonfamily employees (Galvin et al.. 2007) and perceptions of
health or satisfaction (Amarapurka & Danes, 2005; Olson, Zuiker, Danes, Stafford. Heck
& Danes, 2003). Even though there has been much theorizing about the differences

I‘ cross FOBS as well as between FOBS and non-FOB businesses. research has been

thoroughly explain the exact nature ofthese differences and their influence on

“liable to

he key OLItcomes of performance, satisfaction, and longev1ty. This is mostly due to two

\.

fiery lmpOt‘tant limitations within the current research literature. the current theoretical
{;\ . . .
rame and the research methodologies employed to test these theories.

Theme/foul Frame Limitations

0V8! the last three decades many theories have evolved to explain how healthy

152::(385 balance the complexity of business and family. Although early studies and early
3 evelopments in FOB theory helped bring public and academic attention to an
k1 mderserved population, many of the theoretical attempts have been limited by their

Llrjderlymg assumption of “health”, and they all too often use patriarchal, Western

assumptions of family systems as their model of health (e.g. Dyer. 1986; 2006: Fleming
2000). These types of models overlook the great variability possible within family
systems (Bronfenbrenner, 1979; Bubolz & Sontag. 1993', Carter & McGoldrick. 1998.
Minuchin, 1974) and therefore across FOBs (Shanna. et al.. 1997). Also there is a
common assumption in the field that “health" is determined by FOB generational
transfers of ownership and business growth (Fleming, 2000; Glavin et al., 2007; Gersick
et al., 1 997). but not all FOBs value this transfer of ownership. and for that matter not all
P OBS have the same value tied to business growth (Dean. 1992; Distelberg & Sorensen,
2 O 09; Sharma, Chrisman & Chua. I997; Sharma & Nordqvist. 2008; Wong.
l\/I<3Reynolds, & Wong, 1992). Very few theories explore the growth of the family. the
rQle Ofdiversity (e.g. ethnic background. step families. or social economic status) or
\“‘ ariations in FOBs values. desired goals. and success over time. For the field to move

t“ . . .
”Ward With an incluswe theory of health these areas must be evaluated.

Methodological Limitations

The second limitation in the literature is the lack of statistical methodology
§ it blC f . , . -. . 4
U1 a 01‘ evaluating a complex system like a FOB. While the FOB field was founded
“11 General Systems Theory (GST) (Sharma. et al., 1997), much ofthe research on FOBS
1333 “53d re search methodologies that work against the assumptions of GST. For example,
r510“ “the research in FOB literature uses either univariate ANOVA (Analysis of
Variance) Or OLS (Ordinary Least Squares) methodologies. The problem with these

‘11 CthOdOIOgies is that they assume individual independence (Wasserman & Faust. 2004)

,1 .
1.31
l ‘,."‘

I .
Syd

 

4""
L.

.r.

I
3..I

‘l i;

.- J

 

 

 

-. .4
.. N
35..
'1
I. .
~14.“
-_
l“,
‘.
h

 

meaning that data are treated as if individual participants are not influenced or connected
to other participants within the same sample population.
Some researchers have ignored this independence assumption in their research
and sampled multiple representatives from the same FOB (e.g. Fetch & Zimmerman.

l 999). According to the assumptions of systems theories these sampled individuals are
not independent, but interdependent (Bertalanffy, 1969); therefore. these studies are
statistically flawed. Other researchers get around the independence assumption by

Sampling only one representative from each FOB (Chrisman. Chua & Litz. 2003; Zody et
a1 - .. 2006; Zuiker, 1998). While the later meets univariate assumptions. it is open to
SEil‘npling errors as the leaders ofa FOB may not have a holistic view oftheir FOB. In
I: hi3 Case, sampling an owner ofa FOB would not give reliable findings for FOB variables

a 3 a Whole, but rather only for owners ofFOBs. For example. an owner may see his/her

OB as privileging the growth and development of the busmess while that same owner’s
f\

{imily

On-famil y employees see the business as privileging the growth and development of the

In () rder to reliably measure variables within FOBs. a researcher would have to

" _;\
T; JFSI sample multiple representatives from within each FOB (and ideally the majority of

\

T Individuals within each FOB), and then that researcher would have to use statistical

NethOdOIOg ies that do not assume individual independence. These methodologies would
1 liCIUde approaches such as Hierarchical Linear Modeling, (I-ILM: Raudenbush, & Bryk.
:2 002), Sti'llctural Equation Modeling, (SEM; Raykov, & Marcoulides, 2006), Dyadic

D ata Anal ysis (Kenny, Kashy & Cook, 2006) or Social Network Analysis. (SNA;

wasserman & Faust. 1994)-

Statement of the Problem

As addressed in the introduction. FOBS are prevalent. Nearly 80-90% of
businesses in the US. are family owned. and these businesses together are the largest
source of employment and the largest contributor to the Gross Domestic Product in the
United States (Astrachan. 2003). While this population. and its influence on communities

and families has been mostly overlooked in business and family science fields, it is
ga 'ning attention with an increasing flow ofservice providers and scientific research, as
in icated by the creation of an academic journal (The Family Business Review). and
3C ademic organizations for networking and certification of service providers (Family
[‘Til‘m Institute, and Family Enterprise Research Council). Currently the FOB focus is
b 886d on numerous theories most of which were developed in the early 1980‘s and which
\\:. ]IOW promise, but they have had very little empirical support. This lack of empirical
§ “PPOTI iS primarily due to the young developmental stage of field. As a result. FOBS are
§uided by information that has not been thoroughly tested.
The se early theories acknowledged the importance of using systems based
K 3190095 10 Linderstand the complex world of FOBs. The first attempt. the Three Circle
Nodel, illustrated the importance of the interdependence between the family. business.
§nd ownership systems (Taguiri & Davis; 1982). Since that first theoretical step. many
t 11301258 ha‘ve built on to this model (Blancc-Mazagato. de Qucvcdo-I’ucnte & Castrillo.
2 007; Dav i S & Stems. 1996; Dyer. 2006: Gersick et al.. 1997'). While these theories have

1:) TOVIded halpful insights into the complex world of FOBs. they have only limited

empirical support, and most have not thoroughly explored the relationship between the

owning family system and the larger FOB system.
The lack of empirical support for theories within this field is particularly

concerning because the field continues to grow and relay more and more on these

foundational theories. such as the Three Circle Model. Furthermore. many promising

adaptations of the Three Circle Model have been proposed. and continue to gain support.

F or example. it has been shown that owning families vary in three ways: I ) they have

d i Ifering goals and values (Distelberg & Sorensen. 2009: Dyer. 2006: Galvin et al..

2 007), 2) the strength of the boundaries between the family and business systems vary

( Dyer, 2006; Levinson, I971; Zody. Sprenkle. MacDermid. & Schrank, 2006), and 3)

t 1193’ differ in levels of adaptability and cohesion (Davis & Sterns. 1981). In other fields

“Ch as Family Science. Psychology. and Organizational Behavior. it is understood that

x
i; liese issues (Value, Adaptability. Cohesion and Boundaries) co-vary (Ackoff, 1977;

i Eahmmi, 1 992; Eppink, 1978; Krijnen; 1979; Olson 2000; Overholt, 1997; Whitchurch &
:Onstantine. 1993), but we have yet to understand the validity of these integrations as

K hey have e ither limited or no empirical support.

It 1 S not difficult to see why understanding success and health in FOBS has been a

Viifficult task; By using a methodology that can explore each ofthese issues in

i \ elationship to the other, this study was able to provide a much clearer picture of the role

Qt family C1)r’namics, internal values. boundary creation. and satisfaction within FOBs.

In the first phase of this study, the Three Circle Model was tested through social

I). etwork methods, and specifically communication patterns were measured and tested

{1 gainst the model’s assumptions. In other words. it was expected that if there was any

.—-—’

validity to this model. the communication patterns within the sampled FOBs would
follow the subgroup assumptions of the model. More specifically, family communication
would be confined (to some degree) within the family system, and similarly employee
and ownership communication would be confined within the employee and ownership
Subgroups.
The next step in the study explored the validity of integrating concepts that the
li terature had previously purposed as important adaptabtions to the Three Circle Model.
While these integrations have been previously discussed in the literature, they currently
have little to no empirical support. For example, integrating adaptability and cohesion.
( both owning family dynamics). have been discussed theoretically by Davis and Sterns
( ’1 98 I; l 996) and three studies have found limited empirically support for family
Q1ynamics in FOB functioning (Lansberg & Astrachan, 1994; Lee, 2006; Zody,
I\480Der'1‘n id & Sprenkle, 2006). These sources suggest that these family dynamics
i Jifluence the health or success of FOBS. Also. integrating Value Orientations (or
\Whether an FOB values the business. family or both systems equally) was theoretically
Eur posed by Distelberg and Sorensen (2009) but not yet tested. Finally the FOB field has
Vonsistently linked to general systems theory (Sharma, 2004). but few studies have
§xplored 3)" stemic concepts such as communication patterns. system boundaries. and
Qlosed or Open systems. Each ofthese three areas can be directly linked to an underlying
t\<)undation in systems theory and therefore have overlaps and similarities. In this study.
t 116 eXPIOI‘at ion of the integrations to the Three Circle Model provided information about

t he effecfiveness of these three purposed integrations. The conclusion of the entire phase

() 1’16 (Three Circle Model exploration and integration exploration) yielded a new

integrated Three Circle Model. This model will add a great deal to the field due to the
depth of exploration, and the resulting depth of information gained on each integration.
The greatest benefit from this phase was that the field has not looked at these three areas

in relationship to each other. This study was able to measure the strength of each

individual integrated concept in relationship to the others.

Since this phase could be considered somewhat "qualitative" or descriptive in
nature (in that the social network and case study methods used in this study may be seen
as closely aligned with qualitative methods due to the level of depth in the social network
measurements used within each sampled FOB) the second phase provided a quantitative
e Valuation ofthis new integrated model. In other words. ifthe findings in the first phase
‘51 re Sapported with the'quantitative methodology in phase 2. the new integrated model

\S'Vill be viewed as a contribution to the field, providing insight into the role of family

3.Vnamics., family and organizational structure, and internal values.

Purpose Statement

The: primary purpose of this study was to build upon existing systems based

t 1160? ies 0 f F OBS. This study accomplished this by examining three broad areas directly

ilx

elated t0 the influence of owing families on FOB systems: I) the role of family
Qynamics, 2) the boundaries between the family and business systems and 3) differing

"Wels 0f Satisfaction among the family, owners, and nonfamily employee systems.

I.
General Sy Stems Theory, Organizational Theory, Family Theory and Family Business

I. heory all Suggest that these issues are interrelated. and therefore studying the

1 1’1 teractions between these issues will help create a useable theory that can be employed

to explain other complex issues within FOBS (e.g. succession difficulties. retention of
non-family employees. and variation in values and goals).

This study first tested the field’s primary model, the Three Circle Model. This
model has many benefits, but has not been thoroughly tested (Sharma & Nordqvist,
2008). It is thought that the subsystem boundaries in the model may not be accurate and
rn ay not provide enough explanation for the variance across F OBs. Structurally it is true

that individuals are either family. employees, owners or some combination of the three.
but it is unclear whether this structural categorization provides any insight into the
Functioning ofthe FOB. The validity ofthis model can be tested by measuring the actual
i nteractions between individuals within each business and then attempting to explain
t 11686 interactions by using the Three Circle Model. This study did show significant
1 imitations in the Three Circle Model’s ability to explain interactions, and therefore the
Qtudy explored the benefit of expanding the Three Circle Model. Measurements for
Ii\amily dynamics. the strength of the business-family boundary and the level of
\atisfaction across and within each business were added. The findings from this phase of

T.: he study are qualitative in nature. and generated testable hypotheses. which were

 

iii‘iEV’aluated VV‘ith quantitative methods in the second phase.

The second phase used Hierarchical Linear Modeling (I ILM) to test the
lePO‘heSBS of phase I. FILM allowed the researcher to test interactions within and across
§3Ch busine 33. Furthermore HLM allowed the assumption of independence to be relaxed

\e‘w’hiCh made it possible to test variations between individuals within the same FOB (not

i rddependel'lt due to a shared membership in the same FOB).

Specific Aims

Specific Aim 1 .' To evaluate the validity ofthe Three (.‘ircle xix/ode] 's a.s".s'umptions
and its ability to explain interactional patterns within FOB systems. This aim
tested the hypothesis that the Three Circle Model does not fully account for all
possible variations in FOB connnunication structures. Actual I’OB structures were
measured through Social Network Analysis (SNA) methodology and compared to

the assumed FOB subsystem structures in the Three Circle Model.

Specific Aim 2.“ Increase the Three ('irc/e Model 's validity through the inclusion
offamily dynamics, value orientations. and boundary creation. 'l'his aim tested
the hypothesis that FOB structure is affected by family system dynamics ( e. g.
value. cohesion and adaptation). Each FOB was evaluated structurally. and
variations across FOBS were compared. qualitatively. to measures of value
orientation, satisfaction. adaptability. and cohesion within the FOB system.
Support for this hypothesis provided valuable insight into the effects of variations

in family dynamics.

Specific Aim 3 .' Test the new expanded model for its ability to explain the level of
satisfaction within and across F 08s. This aim tested the hypotheses generated
from the qualitative exploration in Specific Aim 2 by fitting a multi-level model

with the findings in Specific Aim 2.

10

Specific Aim 4: Test the new expanded model . for its ahilitv to explain variations
within and across FOB value orientations. This aim tested the expanded Three
Circle Model’s ability to explain variations in individuals perception about

his/her FOB.

Theory Development

The strength of the field of FOB is the systemically rooted theories that have been
developed over the last 3 decades. Theory within this field began with General Systems
Theory (Sharma. 2004). and the most referenced theory to date. the Three Circle Model
came directly out of this foundation. While this is the starting point for the field. these
theories have not been tested. No study to date has evaluated system concepts such as
subsystem norms and roles, or subsystem boundaries (or open and close systems) in a
way that is consistent with General System Theory. Even research that is systemically
rooted tends to be limited by methodologies that do not follow systems assumptions. For
example, there has been research looking at a lifespan development integration
(Rutherford, Muse & Oswald, 2006). and research looking at adaptability and cohesion
(Lansberg & Astrachan. 1994). but these studies are limited by single rater viewpoints.
and univariate analyses and consequently these studies report tentative and limited

findings.

General System Theory

From the inception ofthe field of FOB, General Systems Theory (GST) concepts
and assumptions have been central. To this day theories regarding F OBS contain

explanations of communication patterns. system boundaries. flexibility and

11

interdependence, which are all rooted in the assumptions ofGST (Aronoff, Ward. &
Astrachan. 2002; Shamia. 2004).

From a GST perspective, communication and all interactions follow cybernetic
principles such as negative and positive feedback within closed and open systems
(Bertalanffy, I969). GST builds on to the assumptions of cybernetics and adds that
individuals are interdependent with their surrounding systems (Davis & Stems, 1981;
Gersick et al., 1997; Taguir & Davis. 1982). In other words. behaviors and values are not
solely the product of internal processes but are a response to systemic influences.

For FOBS, this idea relates to the predicament of individuals who are pulled
between two competing systems. the family and the business. From a GST foundation
many theories have been created to explain the unique Suprasystem (Whitchurch &
Constantine. 1993) or the larger FOB system that houses the interdependent family and
business systems. The most referenced attempt is the Three Circle Model (Taguiri &
Davis, 1982). This model was an early model that discussed the characteristics or roles
and rules, of individuals based on where they are located in relationship to the overlaps
between the family. business. and ownership systems. Later on. Gersick et al.. (1997)
determined that the interdependence of the systems in the Three Circle Model produced
interdependent developmental trajectories. with family development affected by business
development, and business development affected by the business development. Also,
Davis and Stems (1981) discussed the need for adaptation and cohesion within each
system to facilitate the interdependence ofthe three systems. Iiach one ofthese theories
has evaluated the relationships between systems, and based on their evaluations. these

theorists have suggested that each individual system is not independent, but is affected by

12

and affects other systems through a shared connection to a larger suprasystcm (in this
case the FOB suprasystcm). In other words, a FOB family system is different from a non-

FOB family system and non-FOB business system.

Three ( 'ircle Model

Theories between the 19605 and 19805 tended to view the FOB system as two
separate systems (family and business). each with separate goals. tasks and
developmental trajectories (Levinson, I97l ). In the l980s. the field began to recognize
FOBS as suprasystems (Whitchurch & Constantine. 1993) or nested systems that together
form the larger FOB system. Tagiuri and Davis (1982) presented one of the first models
to depict FOBS as a nested suprasystcm. These theorists argued that FOBS are made up
of three nested systems (family. business and ownership) which create the larger FOB
system. These theorists believed that the nesting (and subsequent overlapping of
systems) creates seven distinct systems within the larger FOB system. This model has

been termed the Three Circle Model (Gersick et al.. 1997) (See figure l.l ).

13

Figure 1.1: Three Circle Model

'5

Taguiri, R., & Davis. J.A., (1982). Bivalent attributes ofthefamily/inn.

Working paper, Harvard Business School, Cambridge Mass. Reprinted

1996, Family Business Review, 9(2): [99-208.

This Three Circle Model (T aguiri & Davis, 1982) was the first substantive
attempt to recognize not only the overlap between the family and business system. but
also the importance of the Ownership system. This new model gave the field a new
respect for the complexity of family businesses. and even more importantly. it brought a
desire to understand the different experiences and characteristics of each interdepent
system. For example Anderson and Reeb (2003) attempted to explore the importance of

family versus nonfamily managers or in other words. different levels of overlap between

the family and the ownership systems. They found that family owners who also were the

14

FOB managers outperformed nonfamily owners/managers (Anderson. Mansi & Reeb,
2003; Anderson & Reeb, 2003). Therefore an overlap between the family and ownership
systems was found to be beneficial.

While this model was a large theoretical step forward for the field at the time. it
has some limitations. The largest of which is the lack of discussion in four areas; 1)
development over time, 2) possible variations in the definition of “health” across FOBS,
3) variations in the strength of the boundary between systems, and 4) the role and

influence of the owning family system dynamics (Sharma & Nordqvist, 2008).

Developmental Model for Family Businesses (DittFB)

Gersick and colleagues (1997) saw the developmental limitations in the Three
Circle Model and expanded it to account for the development of FOBs over time.
Specifically Gersick et al (1997) theorized that each ofthe three systems in the Three
Circle Model had its own developmental trajectory (e. g. family development, business
development, and ownership development). Only one study in the history of FOB
literature has attempted to validate this model. Rutherford, Muse and Oswald (2006),
sampled over 900 FOBS in the US. and found that the DMFB can be used to typologize
F 085, but other variables such as ownership orientation for growth (business growth
versus family growth), and the level of tension (or cohesion) within the family system are
better indicators of differences between F OBs. Therefore it is possible to use the three
dimensional model purposed by Gersick et al. (1997), but this model does not give

enough information to separate out enough of the substantive differences between F 0B3.

15

Adaptation and Cohesion

Prior to the creation of the Three Circle Model, a few theorists had been exploring
the role of adaptability and cohesion within FOBS (Davis & Sterns. I981: I996). Davis
and Stern (1981; 1996) first outlined the importance ofthese concepts and argued that the
owning family and the FOB must be adaptable and exhibit a certain level of closeness to
survive. They defined adaptability through two concepts: legitimate structures and
emotional containment, with “emotional containment” being the ability to handle
business and family emotions within the appropriate system, and “legitimate structures”
being a division of roles and rules by an individual’s position in a given system. These
concepts contained pieces similar to the family systems definition of cohesion and
adaptability as well as pieces similar to Bowen‘s concept of differentiation (Kerr &
Bowen, 1988), which has been defined as the ability ofindividuals to be balanced
emotionally, to tolerate individual differences. Nichols and Schwartz (2004) described a
differentiated family system as a system that can deal with problems within subgroups,
without directly engaging the entire FOB system. While the concepts of adaptability and
cohesion within Davis and Stems (l 981; 1996) are helpful and add a contribution to the
field, family systems definitions of cohesion and adaptability provide a better dichotomy
of cohesion and adaptability, and these definitions have been empirically tested.

For example, Olson, Sprenkle. and Russell (1979a; 1979b) defined cohesion as
the emotional connection between family members. while adaptability is the family’s
ability to change in the face of external or internal stimuli. In both cases. a family can
exhibit too little or too much cohesion and adaptability. For cohesion, a family system

can be disconnected or cut off (low cohesion) or overly connected or enmeshed (high

16

cohesion). For adaptability, a family system can be rigid and not respond to needed
changes (low adaptability), and a family can be too adaptable. producing chaos due to no
foundation to the system (high adaptability).

In most empirical studies of cohesion and adaptability in F OBs. researchers have
used the Circumplex model (e.g. FACES II or III) (Olson et al., 1985). For example
Lansberg and Astrachan (1994) used the Circumplex Model to test Olson‘s ct al. ( 197%:
197%) concepts of adaptability and cohesion and found that in FOB systems,
adaptability and cohesion generally have a positive linear relationship with succession
planning and succession training. The limitation with this work is that the researchers
only sampled the owners and successors of FOBS. They did not sample other owning
family members (e.g using the Three Circle Model, individuals in subgroup 6 and 7 were
sampled but not subgroup l). Secondly, they assumed a linear relationship between
cohesion, adaptation, and success. Both of these assumptions are not in line with Olson‘s
recommendations for studying cohesion and adaptability within family systems (Olson.
2000). Olson has suggested that a proper exploration of adaptability and cohesion
involves sampling multiple members from the same owning family systems. Olson
(2000) has also suggested that adaptability and cohesion have a curvilinear relationship
with functionality, meaning that family systems on each end of the adaptability and
cohesion continuums exhibit problems in comparison to family systems located in the
middle of each continuum.

Zody, MacDermid. and Sprenkle (2006) conducted a similar study and found that
cohesion was negatively related to conflict throughout the FOB system. In this study the

researchers found that overly connected family systems had less conllict than overly

17

disconnected family systems. Both of these studies (Lansberg & Astrachan. I994: Zody
et al.. 2006) support a linear hypothesis with cohesion and adaptability rather than
Olson‘s (2000) suggested curvilinear hypothesis. In other words. there does not seem to
be a cut off point for adaptability or cohesion for FOBs in the existing research.
Therefore there does not seem to be a danger of being too adaptable or overly connected.
as researchers have noted in family systems research.

While there does seem to be evidence that the adaptability and cohesion ofthe
owning family plays a role in the health and functioning of the entire FOB system,
existing methodological limitations in the research limit our understanding of this role.
The largest limitation has been the tendency to sample only one representative from each
FOB. which historically has not been a good measure of adaptability and cohesion. This
point is confirmed by Thomas and Ozechowski ( 2000) who found that the individual self
reports on the cohesion and adaptation scales in FACES III are not as reliable as multi-
rater versions. Therefore, a better exploration of cohesion and adaptation within FOBs
would include multiple raters from the same FOB system. Finally. measures ofcohesion
in studies regarding FOBs have been somewhat unsatisfactory. with most showing
limited explanatory power for cohesion. especially when controlling for adaption

(Lansberg & Astrachan. I994; Lee. 2006).

Value Orientation
The FOB field has struggled to understand what constitutes a “healthy FOB
system" (Sharma et al.. 1997: Sharma & Nordqvist, 2008). The main reason for this

difficulty has been the field‘s tendency to privilege the business system goals prior to

18

evaluating the real desired goals within actual FOBs. Often researchers define success or
“health” for these systems through measures such as return on assets. growth in sales.
revenue, number of employees and survival rate (Dess, & Robinson, 1984; Kalleberg &
Leicht, 1991; Miner, 1997). In studies like these, specific business variables are created,
and if the FOB reaches an a priori threshold, then the FOB is considered successful or
“healthy”. The problem is that we have yet to understand what FOBs perceive as success
or “what are the meaningful developmental goals” (Castillo & Wakefield, 2007;
Distelberg & Sorensen, 2009; Sharma, et al., 1997).

Human Ecological Theory offers a solution to this problem. Human Ecology
defines health as a system‘s ability to obtain and transfer resources to meet goals that the
system values (Bubolz & Sontag. 1993). A recent exploration of goals. resources and
values suggested that FOBs define “health” through their internal values (valuing the
family and business systems equally or privileging one over the other). therefore holding
certain developmental goals higher than others and using available resources to meet
these goals (Distelberg & Sorensen. 2009). This theory proposes a continuum of values
for FOBs, with a business-first value orientation on one end and a family-first value
orientation on the other.

This systems perspective brings to light the importance of identifying values
within FOBs, as FOBs with different values define health differently. The inclusion of
value orientation is supported directly with a previous study on the DMFB (Rutherford, et
al.. 2006), which found that “ownership orientation“ (or whether the ownership valued
the growth of the business. or the growth of the family) accounted for more variance

across FOBs than the DMFB alone.

19

One final point ofinterest regarding values in F083 is that GST tells us that the
system will influence the values of the system members. It is acceptable to assume that
within a FOB there will be a great deal of agreement on values. For example if the
owners believe that the FOB exists to support the growth and development of the family
system then the employees should to some degree share this understanding. But this does
not mean that they like it. which leads us to assume that FOBs with a Value Orientation
that favors the growth and development of the family system will likely produce higher
levels of satisfaction within the family system, but lower levels of satisfaction within
nonfamily employees. Furthermore this unity in values assumes a functioning system

where no cut offs exists.

Conceptual il-Iodel

The current study integrates the concepts above (including the structural
assumptions ofthe Three Circle Model) in an effort to strengthen (expand) the Three
Circle Model. This study hypothesizes that the following will play a role in the health
and functioning of an FOB: I) The owning family‘s dynamics (adaptability and
cohesion). 2) The value orientation ofthe business. and 3) system boundaries. The quality
ofthis integration will be judged by fitting a model that incorporates these concepts. If
this expanded model can accurately explain variations in satisfaction and perceptions it
will be seen as a step forward in the FOB literature and will address important issues
within FOBs such as the top two most frequently indentified weaknesses: 1) failure in
generational transfers of ownership. and 2) retaining nonfamily employees (Galvin et al..

2007).

20

There are three major points to this expanded Three Circle Model that must be
explored. First, it is clear that FOBs contain three interdependent systems as presented in
the Three Circle Model (Taguiri & Davis, 1982). But what is not clear is whether the
three systems overlap in the same fashion for all F OBs. In other words. does the strength
of the boundary between the family and the business vary from one FOB to another? This
can be explored through the social network phase of this study. By using social network
tools one can measure the actual interactions within each FOB. These real interactions
will tell us how closely real F OBs follow the Three Circle Model structural assumptions.
For example. ifthe Three Circle Model is 100% accurate across all FOBS. we would
expect that the majority ofcommunication regarding the owning family to be limited to
the family subsystem (and the overlapping family systems). and little to no
communication regarding the owning family to be present in the employee or ownership
systems. Or at the very least. this pattern should be highly correlated with the
functionality and health ofthe FOB system (e.g. FOBs that follow the structural
assumptions will have higher levels of satisfaction across the FOB system).

This exploration ofthe Three Circle Model may highlight significant limitations
to the Three Circle Model. It is likely (given the theory discussion above) that there is
variation in the boundaries proposed in the Three Circle Model (e.g. some FOBs allow
more communication and interaction across subsystems than others). If this is found to be
true for the businesses in this study there will be two additional points of interest: 1) why
do FOBs vary in the strength of their boundaries? and. 2) does boundary strength

variation effect individuals within the FOB?

21

The second point of interest then is “what is the affect of varying boundary
strengths in FOBS?” Given the discussion of the current state of FOB theory above. there
are likely three issues that influence the strength of the boundary. These issues are: 1) the
value orientation, 2) the level of adaptability, and 3) the level of cohesion within the
owning family. There does seem to be some evidence in the research that the level of
adaptability influences the strength of the boundary (Dyer, 2006; Lansberg & Astrachan,
1994; Zody et al., 2006), but how and to what extent is unclear as well as our
understanding of the influence of cohesion and value orientation. Some researchers have
attempted to study the role of cohesion (Lansberg & Astrachan. 1994), but the results
have been limited which may be a product of the methodology used and a lack of
exploration of interactions between cohesion and adaptability. Furthermore, the idea of a
value orientation for a FOB is very new and has not been tested. Therefore. we
understand that the adaptability of the family influences the structure of the FOB, but we
still do not fully know how adaptability, value orientation. and cohesion work together to
influence this boundary.

The third point of interest to be explored is how variations in the strength of the
Family-Business boundary influence individuals within the FOB. We can measure this
influence with two outcome variables. The first is the level of satisfaction. For example,
does a permeable boundary increase or decrease the level of satisfaction of an individual
within a FOB? Theories have suggested that a permeable boundary has a negative effect
on satisfaction (Dyer, 2006), but some preliminary research seems to suggest the opposite

(that a permeable boundary increases satisfaction (Zody et al.. 2006)).

22

While both of the explanations above address the relationship between boundary
strength and satisfaction, there is likely a relationship between boundary strength and
individual values. For example, previous research suggests that F OBs excel at uniting
individuals within FOBs in regard to values and goals (Galvin et al.. 2007), but the ability
to unite individuals may be contingent on the boundaries within the FOB. This study will
explore this unity issue by measuring individuals‘ level of agreement on his/her FOB
value orientation (e.g. does an individual see his/her FOB as being closer to the business
or family side of the value continuum?)

For both outcome variables (perceptions and satisfaction). there is likely an
interaction between the two and variability based on an individual’s position in the
system (e.g. owners may have higher levels of satisfaction in comparison to employees
even when we control for other FOB level characteristics). Both of these issues (an
interaction between perception and satisfaction and the individual's position in the

system) will be explored in phases 1 and 2 of this study.

Research Questions, Hypotheses and Variable Definitions

Specific Questions and Hypotheses
This study first tested the assumptions of the Three Circle Model, then moved on
to qualitatively explore an expanded version of the Three Circle Model. The first phase
addressed specific research questions. The exploration of these questions generated

testable hypotheses which were explored in the second phase of this study.

PHASE 1: STEP 1

23

Specific Aim 1: To evaluate the validity of the Three Circle Model’s assumptions
and its ability to explain interactional patterns within FOB systems.
Hypothesis 1: The Three Circle Model does not fully account for all possible
variations in FOB communication structures.
While the Three Circle Model is the most referenced theory within the
field, little research has been done to evaluate its practical significance. The first
phase of this study tested the structural assumptions of this model directly by
measuring communication patterns within FOBs using SNA and compare those
interaction patterns to the assumed interactions within the Three Circle Model.
For example, the Three Circle Model assumes that there is a boundary for family.
employee and ownership interactions. This study measured this assumption for
each FOB. It was hypothesized that if communication patterns fit these then
assumptions the Three Circle Model would be seen as a valid picture of actual
functioning within FOBs.
PHASE 1: STEP 2
In Phase 1: Step 1. the Three Circle Model was found to be helpful, but limited in
explaining functioning with FOBs. Therefore this second step within Phase 1 explored
integrations to this model that have been previously purposed in the literature. which
Show promise due to their foundation in systems theory. and which have credible levels
of acceptance within the field.
Specific Aim 2: Expand the Three Circle Model validity through the inclusion of

family dynamics. value orientations. and boundary creation.

24

Hypothesis 2.]: Satisfaction increases as value orientation moves closer to the business
side of the continuum.

The first hypothesis within this step sought to explore the integration of Value
Orientations within the Three Circle Model. More specifically. as explained in
Distelberg and Sorenson (2009), the point where a FOB falls on a value continuum has
implications for functionality. For example, when a FOB is closer to the family side of
the value continuum it is likely that FOB members support the family‘s goals and
development over the business goals and development. This hypothesis suggests that
FOBs that follow this side of the value continuum will have lower levels of satisfaction
when satisfaction is measured as an average level of satisfaction across all FOB
members. This is due to the majority of FOB members being non-family employees.
Non-family employees will decrease the aggregated level of satisfaction in FOBs when
they perceive that their FOB favors the family development over the business.
Hypothesis 2.2: Satisfaction varies by subgroup membership.

As eluded to in Hypothesis 2.1, individuals within FOBs may vary in satisfaction
due to where they are in regard to the Three Circle Model Subgroups. For example. if
the average value orientation ofa FOB is high (closer to the family side ofthe value
continuum) family members may have higher levels of satisfaction. but non-family
employees may have lower levels of satisfaction.

HypothesisZ.3: Employee groups with higher value orientations (closer to the family
side of the continuum) than the owning family will have lower satisfaction.

Hypothesis 2.2 suggested that satisfaction varied by subgroups. This hypothesis

suggests that value orientations vary by subgroups. Furthermore this hypothesis assumes

25

that if this variation accounted for subgroup membership alone. the level of satisfaction
in the employee group will be lower.
Hypothesis 2.4: Cohesion of the owning family is positively related to satisfaction.

Hypothesis 2.4 attempts to explore the integration of family dynamics and
particularly the family dynamic of closeness (i.e. cohesion). This integration was
purposed first by Davis and Stems (1996), and has been tested by Lanberg & Astrachan.
(1994) and Zody et al.( 2006). These empirical tests have provided limited support for
the inclusion ofcohesion. but the use ofsingle rater methodology within these studies
may have limited the explanatory power of this concept. as the scale used for measuring
cohesion often requires multiple raters to achieve a quality measurement (Thomas &
Ozechowski. 2000). The theories and studies of closeness with FOB imply that close
owning families work better together in FOBs and that closeness within the owning
family directly influences the entire FOB system.
Hypothesis 2.5: A rigid boundary for family communication will reduce satisfaction.
Hypothesis 2.6.“ A rigid boundary for family communication will increase the distance

between employee and family value orientation perceptions.

Hypothesis 2.5 and 2.6 attempt to evaluate systems theory within F OBS directly.
One of the critiques of the Three Circle Model has been that it does not take into account
the general systems theory assumption of variations in systems boundaries. Furthermore
the field has consistently debated the “right" strength for boundaries between the family
and business systems. Theorists tend to purpose that a rigid boundary between the two
systems is optimal (Dyer, 1986; Flemming. 2000). but empirical research highlights the

importance of a permeable boundary between the two (Olson et al.. 2003: Zody et al..

26

2006). Hypothesis 2.5 reflects the empirical research which has consistently shown that
rigid boundaries between the family and business reduce satisfaction within the family
which increases the level of conflict between family and non—family employees.
Hypothesis 2.6 integrated the empirical research on boundaries with the value orientation
concept (Distelberg & Sorensen, 2009).
HypothesisZ. 7: Adaptation is positively related to satisfaction.

Research (Lanberg & Astrachan, I994; Zody et al. 2006) and theory (Davis &
Stems, 1996) suggest that the level of adaptability in the owning family is directly related
to the FOB’s level of health. For this study, satisfaction was used a measurement of
health. While satisfaction may not cover all aspects that can be considered “health“ it is a
good litmus test for the level of functionality within a FOB. If the FOB is not functioning
well it is likely that individuals within the FOB will not be happy with many aspects of
the FOB system. The satisfaction scale used in this study measured an individual’s level
of satisfaction with the owning family, how conflict is handled within the business, the
strategic direction of the FOB. and the level of satisfaction with employees within the

FOB.

PHASE 2: STEP 1
Specific Aim 3: Test the new expanded model for its ability to explain the relationship
between owning family dynamics and satisfaction.

Many of the concepts within Phase 1: Step 2 were found to be valuable
integrations to the Three Circle Model. In addition, the exploration in this step pointed to

some possible interactions between concepts. These interactions are very important to the

27

 

field and to date no study has attempted to measure the interaction ofthese concepts.

Phase 2 explored the previous concepts. along with the interactions to provide further

evidence of the validity of the new expanded model developed through the exploratory

process in Phase 1: Step 2.

Httpothesis 3.1: The distance between an individual‘s perception of their F OB's value
orientation and the actual value of the FOB is negatively related to an individual‘s
level of satisfaction with their FOB.

Hypothesis 2.1 above showed that value orientation is a strong predictor of
satisfaction at the FOB level. Explorations of hypothesis 2.6 showed that satisfaction at
the individual level is positively related the unity of value orientations across an
individual FOBs. Therefore if an individual does not share a similar value orientation as
their FOB colleagues. then their level of satisfaction will likely be lower.

Htpothesis 3.2: Subgroup members vary in their level of satisfaction
The exploration of Hypothesis 2.6 showed that value orientations at the individual

level vary a great deal within FOBs. The Three Circle Model assumes that values and

perceptions vary by subgroup membership. This hypothesis explores whether this
relationship actually exists in FOBs.

lltpothesis 3.3: Different family system types produce varying levels of satisfaction

within the business.

Hypothesis 2.4 showed that cohesion (owning family closeness) has a relationship
with satisfaction. While hypothesis 3.1 from above will accounted for some the variance
in satisfaction by measuring the Value Orientation differences within F 083. hypothesis

3.3 tested the role of owning family cohesion in the presence of varying degrees of

28

differences in value orientation. In other words. while it was shown in hypothesis 3.1 that

individual differences in value orientation can predict some variance in satisfaction. the

owning family’s level of cohesion will also predict variance in satisfaction.

Hypothesis 3.4: Businesses closer to the family side of the value continuum have lower
levels of satisfaction.

Hypothesis 2.] showed a strong relationship for satisfaction and value orientation
at the FOB level, but it was also shown that value orientation at the individual level is
influenced by many variables which were addressed in hypotheses 3.1, 3.2 and 3.3.
Therefore, after accounting for all of the concepts above. is there still variance in
satisfaction that can be explained by the FOB level value orientation alone?

PHASE 2: STEP 2
Specific Aim 4: Test the new expanded model for its ability to explain the relationship
between owning family dynamics and value orientations.

Much of the exploration in satisfaction from Phase 1: Step 2 and Phase 2: Step 1
showed that value orientation at the individual level is fluid. In other words, individuals
can change their value orientation regardless of the value orientation of their FOB. Even
the null model in this current step showed that 75% of the variance in value orientation is
accounted for at the individual level. Furthermore. hypotheses 2.1 and 3.4 provide strong
evidence that FOBs with a total value orientation closer to the business side of the value
continuum have higher levels of satisfaction. Exploration of hypotheses 2.2, 2.5. and the
final model in Phase 2: Step 2 all show that value orientation at the individual level is
fluid. In other words. FOBs can change their FOB level value orientation by unifying the

value orientation of their employees and individual FOB members. Taken together, these

29

findings suggest that the F 085 who wish to reduce their overall value orientation should

begin by looking internally at their individual FOB members. The following hypotheses

provide some insight into how a FOB might reduce value orientations within their

system.

Ht-pothesis 4.1: Subgroup membership will affect the value perception of individuals
within FOBs.

This model begins by acknowledging that while some things may be done to
change Value Orientations, there may be some constants that are not easily changed. For
example employees on average tend to have higher value orientations than owning family
members. This may be an unchangeable structural issue. Therefore, this model starts by
accounting for the variance accounted for by structural subgroupings, and then attempts
to measure the following hypothesis.

Hypothesis 4.2: Access to family communication will decrease an individuals the value
orientation.

The previous model in Phase 2: Step 1 showed that individuals who are in
disagreement with the average value orientation within their FOB have significantly
lower levels of satisfaction. Therefore it is important to understand how a FOB can unite
value orientations within their FOBs. Phase 1: Step 2 provided many points which
support previous research that states that a permeable boundary between the family and
business systems is the best options for a FOB system. In this case the boundary is seen
as a way to unite or divide value orientations within a FOB. This hypothesis measures
this boundary through the use of social network measurements of communication

regarding the owning family. In other words a boundary is seen as rigid if employees do

30

not have access to communication regarding the owning family. Ifa rigid boundary exists
in a FOB than employees who are cut off from family communication will have
noticeably higher value orientations in comparison to their FOB colleagues.
Hypothesis 4.3: The value orientation of the owners will be positively related to
individual value orientation.
While value orientation at the individual level is important, this hypothesis seeks
to understand the role of owners with varying value orientations. For example, do

owners with lower value orientations also have employees with lower value orientations?

Variable Definitions

Family Owned Business (F OB): A business is a FOB if 1) the ownership
members and the family system members perceive themselves as a FOB.
and 2) ifa family possesses the majority ofthe shares. Nonfamily
businesses are defined as businesses that do not perceive themselves as
FOBs and in which a family does not own the majority of the shares.

(Jorissen, et al.. 2005)

Subsystem: According to Taguiri and Davis (1982) there are seven subsystems within
FOBs. Three larger systems; family system members. ownership system
members, and business system members. Because these three systems overlap,
there are four additional subsystems: the family-owner subsystem. the family-
business subsystem, the business-owner subsystem and the final subsystem which
is an overlap of all three systems, the family-owner-business subsystem. For the

purpose ofthis study and its exploration ofthe three circle model. an individual

31

can only be a member of one subsystem at a time. Operationally we measure this
variable by self reports and then verify self reports with information from the
business owners.

Owning Family System: The sum of individuals within the family subsystem. or
subsystem members in the family-ownership. family—business. or family-
ownership-business subsystems. This is a broader definition than nuclear family
as second and subsequent generations of ownership will have multiple nuclear
family systems within the owning family system.

Firm Size: Firm size is the size of the business itself. There are two measures of firm
size: the gross profit for 2007. 2008. and projected for 2009. and the number of
employees within the business.

Generation of Ownership: Generation of ownership is measured by how many
successions have taken place in the FOB. For example a founder stage FOB
would be in the 151 generation of ownership: when he or she transfers ownership
to his/her children, the children would be the 2nd generation of ownership.

Value Orientation: Is a continuum. with F 085 who favor the family system goals only,
on one end and FOBs that favor only the business systems goals on the other end.

Cohesion: Is the cohesion scale in Olson (1985). This is a measure of an individual‘s
perception of the systems closeness and distance in regards to emotional
connection. When the scale is group mean averaged. the result is the systems
level of cohesion.

Adaptation: Is the scale in Olson (1985) for a system's level of flexibility. This

continuum ranges from rigid (lack of flexibility) to chaotic (overly flexible). The

32

scale is an individual‘s perception ofthe system‘s adaptability. When the scale is
group mean averaged, the result is the system’s level of adaptation.

Boundary: Is conceptually an interaction or communication barrier. In this case. a
boundary between the business and family systems would make it difficult for
communication to flow from the family to the business (and vice versa).
Operationally, this boundary will be measured using social network tools such as

centrality, density and block modeling.

33

CHAPTER II: LITERATURE REVIEW

Introduction

Family businesses provide a benefit to both the family and the business systems
(Anderson & Reeb, 2003; Haynes ct al.. 1999; Kaye. 1991; Olson, 2003: Stafford et al.,
1999), especially when combined in the right way. For example, FOBs often use valuable
resources from the family to outperform other businesses. and FOBs provide greater
employment and wealth opportunities to owning families in comparison to other non-
FOB families (Gersick et al., 1997. Sharma, 2004). Unfortunately. it is also clear that if
the family and the business do not function well together. serious problems can develop
(Dyer, 2006; Olson et al.. 2003; Sharma. 2004). There are many anecdotal stories ofthe
business system tearing the family system apart and the family destroying the business
(Fleming, 2000; Gersick. et al.. 1997; Lansberg. 1992). The question that has driven the
field for the last three decades is, “How do the family and the business function in a way
that optimizes the benefits for each system?” (Shanna & Nordqvist, 2008).

There have been many attempts to understand this overlap between family and
business systems. Earlier theories (Davis & Sterns. 1981; Taguiri and Davis. 1982:
Ward, 1987) stressed concepts such as interdependence. adaptability. and unity (often
referred to as commitment or cohesion). These theories where based on General Systems
Theory (GST) (Bertalannaffy 1969). and reflected the complexity and variability
associated with a GST lens. However. much of the research rooted in these theories has
not followed GST principles in their methodologies. For example. the last three American

Family Business Surveys (Astrachan et al.. 1997; 2003: Galvin et al.. 2007) have sampled

34

F OBS on issues directly linked to the interdependence of the family and business systems.
While these findings are important, these studies sample only one representative from
each FOB. This limitation can be found in almost all empirical studies that measure
boundaries, adaptation, unity. or cohesion (Astrachan & Shanker. 1994; Zody et al.,
2006). Furthermore, there has been very little empirical exploration of the foundational
theories within the FOB literature. For example. while the Three Circle Model (Taguiri
& Davis, 1982) has gained wide acceptance (Gersick et al., 1997‘), there has been no
attempt to study whether the assumptions within this model hold. true for real life FOBs
(e. g. are there seven definable subgroups within a FOB, and do these subgroups vary by
the characteristics described in the Three Circle Model?)

This study explored these foundational theory assumptions (subsystem
boundaries, adaptation, cohesion, and unity in values) by employing a family systems
perspective related to FOB functioning. To do this effectively. the current study used
methodologies that evaluated not just one or two representatives of a family/business. but
which explored the perspective and experiences of all individuals within the FOB system.
Additionally this study highlighted the importance of accounting for family system
variability. The findings from this study will encourage practitioners, theorists, and
researchers to consider the family systems effect as equally important as some known
business system effects (such as the effect of varying industries (.Ioriseen et al., 2005),
management styles (Sorenson, 2000). and generation of ownership (Sonfield & Lussier.

2004; Sonfield et al., 2005)

35

Defining a Sample Population

Any study which explores FOBs must begin by defining the target population
(Astrachan, 2003; .lorisen et al.. 2005) because the definition and subsequent findings
have significant impacts on not just future research and theory but also public policy and
governing bodies (such as the IRS. and legislative bodies) (Astrachan & Shanker. 2003).
Also, how a researcher defines a FOB changes the measurement and findings of
empirical studies. For example the census definition of FOB changes the prevalence
measurement significantly from other, broader definitions (Astrachan & Shanker, 2003;
US. Census 2002). Jorissen, Laveren, Martens and Reheul (2005) proved that definitions
of FOBS change FOB versus non-FOB comparisons. For example. Teal. Upton and
Seaman (2003) used three criteria to define FOBs

Founder and families of the founder must control at least 50% of voting

shares, a member of the founding family must serve as CEO and the firm

must have at least one family member as an internal or external director

(Teal et al., 2003, pp. 181 ).
In comparison, Coleman and Carsky (1999) simply defined FOBs as any business that
has an owning family with a 50 percent or larger stakeholder position. Jorissen et al..
(2005) and Astrachan and Shanker (2003), have both suggested that conflicted findings in
research about businesses are solely due to measuring two different sections ofthe FOB
population, rather than true FOB versus non-FOB differences.

In an effort to minimize the effects of sampling error based on inaccurate FOB
definitions, many attempts have been made to present formulas to unify the field‘s

definitions. Some researchers have proposed that the definition of FOB should focus on

36

the level of influence an owning family has on a business (Astrachan et al.. 2002). while
others have focused on the number and role of family members within the business
(Astrachan & Shanker, 2003). While these and other approaches have provided helpful
frameworks for defining F OBS, most research studies have not utilized these tools. One
limitation of these definitions is the use of structural and influence criteria in defining the
sample. This study assumed that F 083 differ in their structural organization and
hypothesized that structural variations influence individual perceptions and in turn are
influenced by owning family dynamics. Therefore this type of definition would have
clouded the findings due to sampling criteria that were similar to the intended variables of
interest.

The definition of FOB used by this study is inclusive, and based on subjective and
objective measures. while not limiting sampling by the number of family members or the
level of owning family influence. This is the most common sampling process in the FOB
literature, and is also used in this study. .lorissen et al.. (2005) proposed the following
definition of FOBs:

We classify firms as family firms ifthey perceive themselves as family

firms and if a family possesses the majority of the shares. Nonfamily firms

are defined as firms that do not perceive themselves as family firms and in

which a family does not own the majority of the shares. (pp. 234)

Important to this definition is the observable and subjective components. The observable
is the percentage of family shareholders. and the subjective portion is the perception of
being a family business. Both elements should be included in any definition of family

owned business, as they seem to be affected by different independent variables and

37

perceptions. Also, it may be these perceptions or the subjective elements ofthis system
that account for a great deal of variance in conflict and satisfaction (Olson. ct al.. 2003:
Zody et al., 2006). Additionally, family system researchers have frequently cautioned
against using objective definitions of family alone. For example, Boss (1987), Vayda
(1983) and Bubolz and Sontag (1993) have all attempted to define the family. and each
have concluded that inclusivity in the definition is important; some even suggest that the
family should have the final authority in defining themselves as a family or not (Boss,
1987; Bubolz & Sontag, 1993).

For the purpose ofthis study. Jorissen et al.‘s (2005) definition will be used as it
allows for the sample to define themselves based on their own perception of family and
family business. However, it also incorporates a minimum amount of control by
including the objective qualifier (e.g. the owning family has to have a majority of the
ownership), which in larger businesses simply means that the total stakeholdership of the
owning family is a larger percentage than any other stakeholder. not necessarily 51%. For
example, the owning family could hold 12% of the shares as long as no other individual

holds 12% or more).

Satisfaction in F083
Since the beginning of the FOB field. research has focused on understanding how
family businesses obtain success or achieve satisfaction. There are two problems with the
way in which this research has addressed this issue. First, success is often defined for
these systems through measures such as return on assets, growth in sales, annual sales,
profits, number of employees, and survival rate (Dess & Robinson, 1984; Kalleberg &

Leicht, 1991; Miner, 1997). In studies like these. specific variables are created. and ifthe

38

family business reaches an a priori threshold. then they are considered successful. The
limitation with these studies is that the field has yet to understand what family businesses
perceive as success or what the meaningful goals are for individual FOBs (Castillo &
Wakefield, 2007; Sharma, et al., 1997; Sharma & Nordqvist, 2008). Therefore these a
priori success measures may not be the measures ofsuccess each FOB uses internally.

Other studies have let family business representatives report their level of
perceived success, which is often measured through likert scale items asking respondents
to rate their level of satisfaction (Dane et al.. 1999: Danes et al.. 2002: Zody et al.. 2006).
While this practice addresses the issue of self perception of success more directly, the
limitation has been that these studies often rely on one representative from a family
business to report for the entire family business. The problem is that perceived success
varies depending on who you ask within a family business (I'Iienerth & Kesser. 2006:
Olson et al., 2003). where an owner may have a different perception of success than
his/her spouse, co-owners. or employees.

Human Ecological Theory can help us understand these sometimes conflicting
findings. Through this lens, satisfaction is a perception held by an individual or group.
This perception is informed by a belief system held by an individual or group. and that
belief system includes; a) perceptions of goals (or identifying meaningful goals). b) the
availability of resources to meet the goals. and c) the fulfillment of goals (Bubolz &
Sontag, 1993). In other words. there is a belief system behind the tangible or objective
success indicators, and therefore the objective measures of success and the belief system
work together to create a perception of success. Using this as our frame for understanding

perceptions of satisfaction, we see that there is a difference between others' perceptions

39

of success (a priori objective measures ofsuccess) and self perception ofsuccess. with
self perception of success being more closely related to satisfaction. Also. since FOBs
differ in what they perceive as meaningful success (which we could call self perception
of success) (Dean, 1992; Hamilton, 2006; Wong, McReynolds, & Wong, 1992), using
objective a priori measures do not allow us to understand the self perception of success.
Therefore, in this study, self perceptions of success will be measured through a series of
likert scale items for each individual within the system. This process allowed the
researcher to obtain individual self perceptions. and through group “meaning" to obtain

group level perceptions of success.

Structure in FOBs

It was illustrated in the first chapter (Figure 1.1) that the Three Circle Model
(Taguiri, & Davis, 1982), is made up of three larger systems that overlap within a FOB,
(the ownership, family. and business systems). This model allows seven distinct options
for subgroup membership. meaning that individuals can be a member of multiple
systems. For example an individual can be a member of the family and business systems
(e.g. a teenage son of the owning family who is employed in the business). a member of
the ownership and family system (e. g. a mother in the owning family system who is also
the CEO), a member of the business and ownership systems (e. g. an employee who also
holds a minority share), and a member of all three systems (e. g. an entrepreneur who is
the father of the owning family, works as an employee but holds the majority of
ownership). Each one of these individuals has a distinct role in their FOB and each

position influences the FOB in different ways.

40

The question that has yet to be answered in the literature is how individuals and
groups influence FOB structure, and conversely. how does FOB structure affect
individuals within FOB systems. This study proposes that FOB structure is a moderating
variable. meaning that individual independent variables (in this case. perceived FOB
value orientation and individual satisfaction) are influenced by the owning family
dynamics (e.g. adaptation, cohesion). Therefore the chosen FOB structure is influenced
by the family system dynamics. and the FOB structure influences the individuals within
the FOB. For example, a family system which is enmeshed (high in cohesion) and rigid
(low in adaptability), may produce a FOB structure with a rigid boundary between the
family and the business systems. This structure is likely to create a situation where family

members have a higher level of satisfaction than non-family members.

Overlap Between the Family and the Business systems

Businesses that share an overlap with a family system contain unique “familiness”
or idiosyncrasies related to the owning family (l-Iabbershon & Williams. 1999) that give
it a certain uniqueness. This uniqueness has been attributed to F 083 success in growth
(Anderson & Reeb, 2003: Beehr, Drxler & Faulkner. 1997; Daily & Dollinger, 1992;
Gallo, Tapies & Cappuyrns, 2000; McConaughy et al., 2001), opportunities for business
ownership in minority populations (Astrachan et a1 1997: Galvin et al., 2007). and higher
survival rates in the five to seven year startup period (Anderson & Reeb. 2003: Chrisman.
et al., 1998: Sharma & Rao, 2000; Sonfield etal.. 2005).

Even though FOBs generally are more successful than non-FOB businesses in

growth and the initial startup period, how successful a FOB is seems to be due to the

41

F OB’s ability to facilitate the overlap between the business and the family system. For
example many studies have looked at resource transfers between the family and business
systems. These studies have indirectly shown the effects of variance in boundary strength
between these two systems (Haynes et al.. 1999; Kaye, 1991; Olson et al., 2003; Stafford
et al., 1999; Zuiker, et a1, 1998). For example, some FOBs allow very few resources to
move from the family to the business (strong rigid boundary) and others allow a great
deal of resources to move across the boundary (diffuse boundary).

A problem develops in FOBs when the owning family begins to feel taxed by
their relationship to a business. or when they feel that the business has taken over their
family (in other words there is diffuse boundary between the two systems). In cases like
these, stress develops in the family (Amarapurkar & Danes, 2005; Dane et al.. 2002) and
that stress easily flows through the diffuse boundary into the business system (Cole.
2000; Danes et al., 1999; Haynes et al.. 2007: Masuo et al., 2001; Zody et al., 2006).

Conversely. a rigid boundary seems to have as many problems as a diffuse
boundary. While it has been shown that a rigid boundary increases business
performance, it also creates high levels of dissatisfaction, anxiety. and conflict within the
family system (Olson, et al.. 2003; Zody et al.. 2006). and limits the family resource
transfers that help FOBs outperform non—FOB businesses (Anderson & Reeb. 2003;
Beehr, Drexler & Faulkner, 1997; Daily & Dollinger, 1992; Gallo, Tapies & Cappuyrns,
2000). The best option seems to be a semi-permeable boundary where resources are
brokered between the two systems rather than restricted or flowing too freely.

While the permeability of the Family-Business system boundary is predictive of

satisfaction and conflict. what resource is transferred seems to have as much impact as

42

the FOB’s boundary strength. The meaning and value tied to individual resources is
predictive of the FOB’s perception of success (Cole, 2000; Haynes et al.. 2007; Masuo et
al.. 2001; Zody et al., 2006,). For example. a FOB that values the growth and
development of the family system will place a higher value on family resources (such as
family time), where as a FOB that places a higher value on business system goals will
value business resources (such as CEO salaries) (Olson et al.. 2003). Therefore an
understanding of effective boundaries between the family and the business system is
more complex than measuring how much of a resource, or what type of resources are
transferred from the family to the business or vice versa. Rather, an understanding of
system boundaries includes the permeability of the boundary and the value orientation of

the FOB system.

Overlap of F amil 1', Business and Ownership systems

While the overlap between the family and the business system is complex. the
overlap between family, business. and ownership systems is even more complex.
Unfortunately this situation has been confounded in the research with the developmental
stage of the business, and most of the research in this area is focused on the founders of
family businesses, as this overlap (owner, family, business overlap) is most apparent in
the startup developmental stage of a family business (Gersick et al., 1997: Shanna. 2004).
During other times in the business development cycle we see individuals occupy all three
systems less frequently. For example, when businesses move from the single owner to
the sibling ownership phase, family members are diverted to the business system

(become employees) or the ownership system. It is less likely. as the business grows, to

43

see individuals within the family occupy both the ownership and a position within the
business system (e.g. be a CEO and hold a sales position) (Gersick et al., 1997).

Research also has shown that founders have a significant effect on the values.
performance and culture of their firms (Anderson et al. 2003). Founders who occupy all
three systems add tremendous value to their families and businesses. Anderson and Reeb
(2003) as well as Anderson. Mansi, and Reeb (2003) report that founders outperform not
only non-family CEOs, but also successive generations of family C EOs. But founders are
under a great deal of pressure to perform. A seminal study in the comparison of family
versus non-family C EOs was McConaughy"s (2000) study, which showed that family
founders have longer tenures (17.6 years compared to 6.43 years) and receive
approximately $565,000 less in total compensation than their non-family CEO
counterparts. Feltham. F eltham. and Barnett (2005) found that most organizations depend
heavily on the leadership ofthe founder with most making the majority ofthe decisions.
and 57% of founders operate largely alone, with fewer than two key managers to help
with the business.

Expanding our focus beyond founders and into all individuals who occupy the
overlap between the three systems throughout all the business developmental stages. we
see that the management styles of these individuals are important to the level of
satisfaction within family businesses. For example, individuals who are central to each
system but seek and value the input of all the individuals around them (termed participant
leadership) have the best success in terms of creating a functional business and family
system and also engendering satisfaction in all the family business members (Sorenson.

2000)

44

When these individuals are central to all three systems the effect seems to be that
they outperform nonfamily leaders (Anderson & Reeb, 2003). But "central” is a
balancing act. Individuals that are too centralized limit the FOB‘s effectiveness; for
example. F OBS do not perform as well when an individual in this position holds more
than a 12% stake in the firm (in publicly traded companies) (Anderson et al. 2003) and
stays in an ownership position too long (Zahra, 2005). Therefore these individuals need
to be central, as they drive the family business system and have the greatest amount of
influence on each of the individual systems, but they have to act as gatekeepers to each of
the systems (Morris et al.. 1997; Steier, 2001). When they hold the growth and
development of each system equally. and broker resources. rather than control resources.
they are fundamentally important to the success of each of the three systems as well as

the whole (Sharma. 2004*).

Boundaries and Cohesion within FOB Systems

Important to the discussion of the boundaries between family systems and
business systems is the work of Minuchin and Olson. Minuchin (I 974) originally
theorized that family boundaries vary from enmeshed to disengaged. Olson et al.,
(1979a; l979b) proposed that enmeshment and disengagement were two ends ofa
“cohesion” continuum. Therefore. disengaged families were defined as families that do
not feel connected to each other, and conversely, individuals within enmeshed family
systems have difficulty delineating their own ideas. goals. and values from others in their
system. In regards to permeable and rigid boundaries discussed in the family business

research (Zody et al.. 2006). enmeshed FOBs would employ a boundary between the

45

family and business system that is overly permeable. whereas rigid boundaries would be
similar to disengaged systems (very little flow of communication between the family and
business system).

From this theoretical foundation, Olson and colleagues (1979a) created a
statistical measure of this closeness and distance between individuals within a family
system, which they termed cohesion. This scale for cohesion was included in a family
systems assessment tool known as the “Circumplex model”. Over 200 studies of the
C ircumplex model have verified the importance of cohesion in family systems (Olson
2000). It is entirely possible that this measure can be a useful tool in understanding the
boundaries between the family and the business system. For example, there are more than
likely enmeshed and disengaged FOBs when it comes to the intersection between the
family and the business. An enmeshed FOB occurs when there is a highly permeable
boundary between the owning family and the business. Likewise a disengaged system
occurs when there is a rigid boundary between the family and the business. Zody and
colleagues (2006) have found that FOBs located closer to the enmeshed side of the
continuum had the highest reports of satisfaction. This study indicates that the boundary
between the family and business should be not be too rigid, and in fact that boundary
should be closer to the enmeshed side of the continuum. Olson and colleagues found this
same relationship, but also found that F OBs closer to the rigid or disengaged side of the
cohesion continuum also produced conflict within the family system (Olson et al.. 2003).

Unfortunately. many of the studies that measure cohesion in FOBs have been
somewhat unremarkable. For example, Lansberg and Astrachan (1994) attempted to

measure the effects of adaptation and cohesion on succession planning within FOBs

46

 

(using the FACES II, a version of the Circumplex model). In this study cohesion was a
significant predictor of succession planning, but in the presence of adaptation. cohesion
accounted for very little variance in succession planning. A similar effect was found for
family conflict in FOBs (Lee. 2007). Taking this into account, it may seem as though
cohesion is not a meaningful variable in FOBs, but other theory (Davis & Sterns. 1981;
Olson et al., l979a; 1979b) and family systems research (Olson 2000) insist that cohesion
is a factor in both family and organizational functioning. One possible reason for the lack
of significance in research studies on cohesion may be due to the problems with the
Circumplex model itself.

The relationship between cohesion and adaptation has long been debated.
Originally, Olson and colleagues (1979a: 1979b) argued that the relationship between
adaptation and cohesion was curvilinear. meaning that adaptation and cohesion form two
axes. Individuals who scored high on cohesion and high on adaptation were considered
problematic, and likely to exhibit numerous maladaptive symptoms within their family
system (similar for low cohesion and low adaptation). Therefore the ideal for family
systems was thought to be a good balance in both cohesion and adaptation, although.
since the inception ofthe Circumplex model. many have challenged this notion (see
Anderson & Gavazzi, 1990; Amerikaner, Monks, Wolfe, & Thomas. 1994; Dayley,
SowersHoag. & Thyer, 1991; Farrell & Barnes, 1993', Fristad. 1989; Green. Harris. Forte.
& Robinson, 1991; Hampson, I-lulgus. & Beavers. 1991; Perosa & Perosa. 1990; Pratt &
Hansen, 1987). Even Olson (1994) himself has conceded that the two scales in the
C ircumplex model are linearly related (meaning that the higher one is on cohesion and

adaptation, the less likely they are to exhibit maladaptive symptoms). But Olson (1994)

47

 

and others (Thomas & Ozechowski. 2000) have shown that the reason for the linearity
finding in the Circumplex is mostly due to the self report format of FACES I, II and III,
rather than the actual constructs or C ircumplex model itself. Furthermore. when multiple
raters are used to measure cohesion and adaptability. the curvilinear hypothesis is
supported (Thomas & Ozechowski. 2000). Since all of the cohesion studies in FOB
research have used one representative. it is not surprising that this field has experienced a
similar difficulty. Therefore the study of cohesion in FOB must rely on multiple raters
within the same FOB system in order to measure cohesion effectively.

Important Structural Issues for this Study

First, the structural characteristics of the ownership, business and family systems
are important to this study. How relationships function in these overlaps seems to have a
great deal of influence on the overall success and satisfaction ofthe whole. We do know
that FOBs who perceive themselves as successful have a defined structure within the
overlaps between family. ownership and business system. and the boundaries between
systems seem to be more permeable (rather than more rigid) (Zody et al., 2006).

Second, the centrality of individuals within the family business relates to the
overall health and success of the family business system. When an individual is too
centralized he/she is in danger of holding system resources too tightly. Individuals who
are central, but encourage cross system interaction and broker resources rather than
control resources, tend to produce family business systems with higher levels of
satisfaction in both the business and the family.

While there is some existing research on different types of structure within family

business. little is known about how family businesses choose or employ these structures.

48

The following section will outline a number of issues that are hypothesized to have an

influence on the chosen structure within a family business.

Values

According to Human Ecology Theory. Values are human conceptions of what is
good, right and worthwhile (Bubolz & Sontag, 1993). Values can be religious or spiritual
in nature, such as what is wrong or humane. But they also are deeply rooted in our day-
to-day functioning and help us prioritize our resources. Each ofthe three interconnected
entities that make up the FOB system have their own values (Bubolz & Sontag. 1993;
Davis & Stems, 1996; Gersick et al., 1997). The challenge for a family business is
related to how to incorporate the values ofall three systems and produce a value
orientation for the FOB system as a whole.

Two lines of research have given us some idea of the values within FBEs. The
first is Agency Theory research. The primary concern of research in this area is finding
mechanisms where individual and collective values can be united. so that individuals are
more inclined to subjugate their individual values for the betterment of the collective
(Gomez-Mejia et al., 2002; Schulze et al., 2001). Second, Resource—Based, theories have
indirectly led us to a broad understanding of the values inherent in family businesses.
Although resource-based research does not specifically address values, Human Ecology
Theory tells us that the decisions regarding the transfer of resources are driven by the
ecosystem values (Bubolz & Sontag. 1993). The conclusions from these lines of research
show that resource flows in these systems are rarely equitable. They usually favor either
the family system or the business system (Haynes. Onochie & Muske, 2007; Gomez-

Mejia et al.. 2002; Schulze et al.. 2001 ). Some researchers have titled this phenomenon

49

the “duality of economic and family ties“ (Blance-Mazagato, de Quevedo-Puente &
Castrillo, 2007 p. 200). An appropriate assessment from a these studies is that there is a
variance between F 085 in what they value (e. g. the health of the family system. the
health of the business system. or the health of the entire family business).

In an earlier work, this author proposed that family businesses vary along a
continuum of values (Distelberg. 2008; Distelberg & Sorensen. 2009) and then
subsequently tested this theory using the 2007 American Family Businesses Survey
(Galvin et al., 2007). In this study, values were explored on a continuum. with one end of
the continuum representing FOBs that valued the family over business goals, and who
supported the family over the business through privileging employment decisions and the
transfer of resources to the family. On the other end of the continuum lay family
businesses that valued the business over the family. In this study, value orientation did
not predict measures ofsuccess. but did influence what success goals were valued. For
example. in regard to succession goals, FOBs that lay closer to the business side of the
continuum tended to value selling the family business outside of the family, whereas
family business on the family side tended to value transferring ownership of the business
within the same owning family. This study concluded that a continuum of value
orientations does exist across family businesses.

The limitation ofthese studies (Distelberg. 2008; Gomez-Mejia et al.. 2002:
Haynes, et al., 2007; Schulze et al., 2001;) and others is that they rely on a self report of
values by one family business representative. It is likely that the real value orientation of
a family business involves more than an overt self-report of values by one or more

individuals within the family business. It is possibly even more complex than a sum of

50

the values of all individuals within the family business. Identifying the value orientation
more than likely involves assessing the weighted sum of the values within the family
business, because some individuals may have a greater influence on the total value
orientation, such as founders or managers.

The current study hypothesizes that values influence which boundaries are
employed within the family businesses. For example. Distelberg (2008) found that
“business-first” FOBs tend to desire selling the business outside of the family. whereas
“family-first’ FOBs tended to desire not only keeping the family business in the owning
family, but also dividing the ownership equally. whereas FOBs in the middle of that
continuum preferred keeping the business in the family but dividing ownership based on
individual characteristics (the desire of individuals to become owners, or the amount of
time and effort an individual previously put into the business). It is possible that
disengaged family systems correlate with the business-first value orientation and that
enmeshed family systems correlate with the strong family-first end of the continuum. If
this is the case. then not only is the boundary between the family and the business
important, but also the value orientation of the owning family to business growth and

family business satisfaction.

Agreement on Values
In Distelberg’s (2008) study, the actual value orientation explained much less than
the “agreement of values” (agreement between owners, family members, employees and
clients or customers). According to the findings of the American Family Survey (Galvin

et al., 2007), more than 80% of family businesses report a high degree of unity in values.

51

This means that the representative of the family business reported that the employees.
family members, ownership, and customers all shared similar values to the owning family
values (Distelberg, 2008). When this is the case (a family business with agreement in
value directions on each level). FOBs report a higher level of optimism for the future. and
they have an easier time reaching an agreement between generations regarding the future
ownership ofthe business (e.g. sell the business or divide ownership across the family
equally).

While the scale used in Distelberg’s (2008) study for the “agreement of values”
measured only the representative's perception of agreement of values across the family.
the employees, and their clientele. this scale hints at the notion ofcohesion. This scale is
not a measure of family cohesion, but it is appropriate to assume that family systems with
a healthy level of cohesion also would share similar values. What is not clear is the
relationship between enmeshed family systems and value orientation. It is likely that
enmeshed family systems have a high degree of value agreement, but it is also reasonable
to think that there might be a disagreement in enmeshed family systems on value

orientation.

Adaptability
So far we have discussed FOBs as if they were static: in reality. a certain level of
adaptation must exist within each family business. Certain boundaries that were
employed during one generation of ownership, or during one stage of the business
developmental life span, may not be functional during another stage. A healthy level of

adaptation within FOBs will allow FOBs to adjust their values. boundaries. and structure

52

 

to accommodate the new goals and challenges in the new generation or stage ofbusiness
development.

In previous studies of family business values, one of the major foundations has
been the role of family adaptability (Lansberg & Astrachan, 1994; Davis & Stems, 1981;
Distelberg & Sorensen, 2009). Human Ecology defines adaptability as the . .behavior of
living systems that changes the state or structure of the system, the environment. or
both. . .Adaptation is a necessary process for the growth and progressive integration of
living systems” (Bubolz & Sontag. 1993. p. 433). Olson et al., (1979a; 1979b) added to
this idea of adaptability to their Circumplex model. In this model. this axis is a continuum
with overly flexible and rigid family systems as the two ends of the continuum. In other
words, family systems that adapt too much are chaotic. There is very little continuity in
the system, as it takes very little to change the structure of the system. Conversely, rigid
systems do not adapt enough. Certain environmental and developmental events require
that systems adapt to some degree to survive. Rigid systems refuse to adapt even in the
face of negative consequences to the system. This idea could be adapted to FOBs.

From organizational theory, adaptability is often referred to as an organization‘s
flexibility. A flexible organization has a structure that allows the organization to succeed
under environmental pressure and unpredictability (Ackoff, 1977: Eppink. 1978). It has
the ability to make structural changes quickly. To make these structural changes. an
organization has to be “decentralized” in decision making. with a high degree of
permeability of boundaries and collaborative partnerships (Bahrami. 1992: Krijnen: 1979:

Overholt 1997). In other words. certain boundaries allow for adaptability. and certain

53

structural characteristics of organizations facilitate adaptability better than other types of
boundaries and structures.

These ideas have been examined in the study of FOBs. For example. there are
many positive benefits to the centrality of owners. but adaptability is limited when
owners are too centralized (especially when the FOB is larger, as it often is in second and
third generations (Anderson et a1, 2003; Anderson & Reeb, 2003; McConaughy, 2000;
Zahra, 2005). While organizational concepts like formalization (rigid boundaries), and
centrality within organizations decreases adaptability in organizations (Aiken & Hage.
I971; Corwin, 1972; Damapour. 1991). the key to health in FOBs seems to be a balance
between centrality and decentralization. where the ownership system is central to the
FOB but it also allows others throughout the FOB to use and transfer resources (Burke,
2007; Hatum & Pettigrew, 2004).

From both the Family Systems and Organizational perspectives. an organization
must be what General Systems Theorists call an open system (Bertalanffy, 1969). Open
systems allow for change within the system based on new information that is introduced.
Conversely, a closed system does not allow new information into the system and
therefore, since systems like to maintain a steady state or equilibrium. they will not
change without new information. Therefore the structure around a system. or the
characteristics of a system that allow (or do not allow) information to enter a system are
determinants of the system’s ability to change or adapt.

The concept of adaptation crosses every one of the previous ideas discussed in the
previous sections. According to SFT. the quality of a family system is based on its ability

to shift and change structures and functions (Minuchin. 1974). Family systems that do

54

not change when external or internal environmental changes require a shift produce many
problematic symptoms. If a family business does not possess a healthy level of
adaptation, they may not be able to make structural, boundary. or value shifts. This is
evident in studies like Anderson et al. (2003) and Zahra,( 2005), where the leaders of the
business, family, and ownership systems were unwilling to train and introduce new

leadership. and in the process reduced their family business’ profitability and satisfaction.

Other Variables to Consider
The following section provides some specific demographic issues that have been
shown to provide variance in the family business population. Therefore. when doing
research in this area the following variables need to be controlled for. and the effect of

these demographics should be made explicit.

Firm Size

Intuitively the size of the firm would have an effect on many variables relevant to
family business research. For example. the strategies used. and the tensions in both
business and family systems will likely vary by firm size. especially when firms differ in
greater numbers (e. g.. 10 employees versus 10,000). An often criticized feature ofthe
family business literature, that there is very little (or no) delineation between Wal-Mart or
Ford (both considered family businesses under some definitions) and the local mom and
pop restaurant down on the corner. While it can be argued that these two extremes are
just two ends of the same developmental continuum (Gersick et al., 1997), if this
demographic issue is not controlled for, the results may be a function of the firm

differences, and not necessarily the actual variables of interest, especially when the

55

objective is a comparison of family businesses versus non-family owned businesses
(.lorissen et al.. 2005).

To further examine the effects of this issue. Lussier and Sonfield (2006) recently
conducted a test of changes in family businesses as they grow and provided a map of
differences between small and large family businesses. They found that overall. larger
family businesses have significantly (p <0.05) more non-family members within top
management and make greater use ofoutside consultants. advisors. and professional
services when compared to smaller family businesses. Additionally. while the larger
family businesses exhibit less conflict and disagreement between family members. they
also spend more time in strategic management activities. and use more sophisticated
methods of financing.

Using a slightly different approach to understanding the limitations of failing to
control for family business size. Jorissen. Laveren. Martens and Reheul (2005) found that
controlling for firm size eliminated differences often found in the literature between
family businesses and nonfamily businesses for strategies used. networking. perception of
the firm's environment. long-term planning. nonfinancial control. growth. and
management training. This finding also was held up in a comparison between standard t-
tests versus multivariate forms. which included controls for firm size. The indication here
is that firm size (as well as other demographic variables) can be effectively controlled for
through multivariate methodologies.

For the purpose of this study. the size of the family business will be determined in
the initial contact with the owner. Size will be defined as the total number of full time

and part time employees as well as by the previous year‘s revenue. Using two different

56

measures of size will provide two separate variables, and both can be examined to

determine any differential effects.

Gender

The research on women in FOBs is in an early developmental stage. As a result
there is little that can be said regarding women in these systems. According to the
American Family Business Survey (Astrachan et al., 2003) a growing number of women
are entering family businesses and taking leadership positions. Businesses founded after
1980 are more likely to be women-owned (21.1%) than those founded before 1980
(14.1%). Since women’s leadership in family businesses is a fairly new development,
studies of women in leadership are confounded by the age of the firm and the generation
of ownership. and therefore it is difficult to determine the actual effect of gender in
family businesses. Most of the differences in studies comparing men led versus women
led family businesses report differences in managerial style and debt and equity practices.
These factors are more than likely a function of firm age rather than gender (I-Iaberman &
McTarvish, 2005; Sonfield & Lussier. 2005).

While there is little direction in the current research about differences between
men and women, it does seem that a female owner’s perception of well-being is tied to
her ability to balance both her family and business values and goals. Also, a female
owner‘s well-being seems to be related to the income received from the business (Lee.
Danes & Shelley, 2006). Finally. two very interesting trends have developed. and require
further exploration: first women in family businesses seem to be overextended with 25%

of women working at home. at the family business, as well as at another place of

57

employment (Lee, Rowe. & Hong, 2006). These women are more likely to be the primary
manager of both their family and business (Masuo et al., 2001). Secondly. studies of
succession in female run F OBS suggest that women that receive ownership from their
father are more successful than when they receive ownership from their mothers (Dean &
Vera, 2005). It was suggested that this succession issue is due to gender stereotyping.
where the daughter is expected to be similar to her mother. but allowed to be different or
unique from her father’s management style.

Therefore controlling for gender effects involves an exploration of interactions
between firm size. firm age. and (when the focus is succession) the prior generation.
Later we will see that controlling for gender also involves an exploration ofinteractions

between. countries. geographic locations. and industry.

Industry

Industry is a specific section of an economic sector (e.g. manufacturing. retail.
service, technology) is the grouping ofbusinesses by the services or products they
perform/provide. The North American Industry Classification System lists 1.107 different
industries within the North American economic sectors (US. Census Bureau, 2007).
While the classification of industry has become extremely sophisticated. and most
businesses follow this system as it relates directly to legal and tax issues. far too few
studies have focused on industry effects for FOBs. A handful of studies have focused on
a specific industry (Danes & MeTarvish 1997; Stewart & Danes. 2001). While these are
helpful to that industry, the generalizability of these studies beyond the sample

population is unknown. One recent study (Jorissen et al.. 2005) found that controlling for

58

industry eliminated differences often found between family businesses and nonfamily
businesses. One important aspect of this study was its ability to control for industry
differences using multivariate methodologies.

When controlling for industry. Westhead and Cowling (1997) found that family
businesses versus nonfamily businesses are equally growth oriented and equally export
focused. whereas Donckel and Fronlich (1991), who did not control for industry. found
that family businesses were less growth oriented, more risk averse. less active in
networks, and less export oriented. Also, while Daily and Dollenger (1992) found that
family businesses and nonfamily businesses had equal growth when there was no control
for industry. Gallo (1993) found that there were lower growth levels in family businesses
when controlling for industry. Therefore. the question ofhow industry interacts with
other family businesses outcomes is still unknown, but there is enough evidence to argue

that the industry should be controlled for in all family business studies.

Geographic Location

The research on the effects of geographic location has produced both limited and
sometimes conflicting results. There are at least two levels of influence associated with
the geographic location of a family business. First, the higher level is associated with the
country location. For example. Sonfield and colleagues (2005) and Sonfield and Lussier
(2005) found no differences across four countries for succession planning and strategies
in first, second, and third generations. In a similar study. Lussier and Sonfield (2006)
found that in comparison to French FOBs, U.S. F 085 have a smaller percentage of

women family members working in the business and less conflict and disagreement

59

between family members; the researchers also found that larger US. companies (in
comparison to larger French companies) spend more time in strategic management
activities, and used more sophisticated methods of financing. This study and others
(Astrachan, 2003; Jorissen et al., 2005) have suggested an interaction between country.
firm size, firm age, and gender. Therefore when controlling for country level differences
one should consider the interaction between the country level. and firm size. firm age.
and gender.

The location differences within the boundaries ofa country (e.g. urban versus
rural locations, or east versus west, or north versus south) is another demographic
variable often overlooked in the current literature. An example of the problems
associated with not controlling for this level oflocation can be found by comparing
Westhead and Cowling ( 1997) and Donckel and Fronnlick (1991). When controlling for
the location, Westhead and Cowling (1997) found that family businesses and nonfamily
businesses are equally growth oriented and equally export focused, but Donckel &
Fronlich (1991) did not control for location and found that family businesses were less
growth oriented and less export focused. Furthermore, Jorissen et al., (2005) found that
controlling for this level of geographical location eliminated differences often found
between FOBs and NonFOBs. Therefore controlling for both levels of location is

important.

Age affirm
Some examples of the problems associated with not controlling for firm age can

be seen by comparing Teal and colleagues (2005) to Westhead and Cowling (1997). and

60

to Gallo (1995). Teal et al., (2005), and Westhead and Cowling (1997) did control for the
firm age and found equal levels of growth between family businesses and nonfamily
businesses, but Gallo (1995) found less growth in family businesses when not controlling
for firm age. Also, when controlling for the firm age, Westhead and Cowling (1997)
found that family businesses are equally growth oriented and export focused, but Donckel
and Fronlich (1991) found that family businesses were less growth oriented (did not
control for firm age). The importance of controlling for firm age is further explored and

supported by the work of Jorissen et a1 (2005).

Summary

This study begins by exploring the assumptions of FOB structure within the FOB
literature. It compares the theory with actual F OBS and integrates owning family
dynamics (eg adaptability and cohesion) and value orientations into this exploration of
FOB structure. It is hypothesized for this study that the owning family’s level of
adaptability and cohesion effects the strength of the boundary between the family and
business system. It is also hypothesized that the strength of the family-business boundary
influences the individuals within the FOB.

This study"s strength over previous studies regarding family dynamics and
boundaries is the inclusion ofa value orientation and the methodology employed. The
value orientationis a new concept for the field and this study will explore the effects of
varying value orientations on FOBs. The methodology will sample a wide range of
F 085 to account for the issues of FOB size, gender of ownership. generation of

ownership and industry. Also the methods allow for sampling multiple representatives

61

from within the same FOB rather than relying on one representative, allowing for a much

more trustworthy picture of each FOB.

62

CHAPTER III: METHODOLOGY

Introduction
This study explored the role of family systems dynamics in FOBs through
carrying out the following four specific aims:
1. Evaluating the validity ofthe Three Circle Model‘s assumptions for
communication structures within FOB systems.
2. Expanding the Three Circle Model”s validity through the inclusion of
family dynamics. value orientations. and boundary creation.
3. Testing the new expanded model for its ability to explain the level of
satisfaction within and across F083.
4. Testing the new expanded model for its ability to explain the variations
within and across FOB value orientations.

Each one ofthese aims addresses the role of family systems in FOBs which has been
overlooked in the FOB literature (Distelberg & Sorensen, 2009; Sharma & Nordqvist,
2008). These aims also will directly or indirectly evaluate the effect of variations in FOB
structures. evaluate the Three Circle Model for the first time (Taguiri & Davis. 1982).

and examine key variables (values. adaptation and cohesion) central in FOBs.

Methods

Sampling Procedures
Given the exploratory nature of the study and the access to funding and resources.
the study was limited to one state. The first step in identifying a sample population was

63

to contact organizations within the state that served F 085 (e. g. Nonprofit membership
groups such as the local Chamber of Commerce, the Family Business Alliance and the
Family Owned Business Institute). These organizations offer membership to FOBs and
provide educational programming as well as networking services to their members. These
organizations were briefed on the study and the potential benefits to their member
businesses, and were encouraged to advertise the opportunity to their members. The
researcher then followed up these advertisements with an email or telephone call to the
business owners and invited them to participate.

A total of 63 FOBs were made aware of the study, and 23 business owners
expressed. interest in participating. Once a business owner expressed interest in
participating, the researcher met with that individual and discussed the study process. A
total of 12 business owners decided to not participate due to increasing economic stress,
planned layoffs. or general uncertainty about the future ofthe 2009 economic
environment, and l 1 businesses agreed to participate and completed the entire data
collection process.

Prior to collecting any data from the business, the owner was asked a series of
questions to determine whether the business met the inclusion criteria ofthe study. The
following two inclusion criteria were evaluated prior to beginning any data collection:
First, the FOB needed to meet the following definition of a Family Owned Business:

A business is a FOB if the ownership members and the family system

members perceive themselves as a FOB, and if a family possesses the

majority of the ownership shares. Nonfamily businesses are defined as

64

businesses that do not perceive themselves as F OBS and in which a family

does not own the majority ofthe shares. (Jorissen et al., 2005).
Secondly, the business needed to provide a 70% response rate in all subgroup areas
(family members. ownership members. business members). FOB gate keepers who did
not believe a 70% response rate was possible were not included in the study. These
two criteria allowed 1 1 F OBS into the sample population. and 492 individuals were

surveyed.

Data (,‘ollection

Once a business owner had given the researcher permission to conduct the study
within the business, the researcher obtained a roster of names for employees. owners and
family members. At the same time, the researcher conducted a brief interview with the
owners for the purpose of collecting the business level demographic information (e.g.
revenue, number of employees. generation of ownership). At the conclusion of this
meeting, the researcher collaborated with the business owner to develop a plan of action
for collecting data from the employees. In all but two cases the plan involved an
advertisement by the owner and a series of emailed and mailed invitations to take the
online or paper version of the survey. Advertisements by the owner ofa business were
carefully planned with the help of the researcher, so that the advertisement met two goals;
1) make employees aware of the study, and 2) highlight “voluntary” and “confidential”
participation (employees were aware that they were not required to participate. that
participation or a lack of participation would not affect their employment. and their

participation was confidential, in that only the researcher would see their responses to

65

survey items). For each of the three methods of survey administration (email. mail.
onsite) participants were given an informed consent form (Appendix E) which detailed
the risks and benefits of participation. During onsite administration this consent form was
read out loud by the researcher and time was given to address any concerns or questions.
The Informed Consent also included information about financial compensation. A lottery
was held for each business. For each business one $50 gift card per every 75 employee
was given at random to a participant. A random number generator was used to determine
the winning participant.

There are a total of three surveys that were used for this study, one for business
level variables, one for all participants. and one specifically for family members of the
owning family (see; 1) business owner interview (Appendix A). 2) participant survey
(Appendix D), and 3) family member additional survey items (Appendix C).

Business Owner Interview (Appendix A)

During the initial meeting with a Business Owner, the researcher conducted an
interview using the Gate Keeper Interview in Appendix A. This interview served three
purposes: 1) collaboration with the business owner in obtaining access to the sample
participants, 2) obtaining a roster of all possible sample participants within the FOB, and
3) collecting business level demographic information. These variables will be used in the
case study portion ofthe analysis (Specific Aim 1). Covariate items from the Gate
Keeper Survey are: age of business. generation of current family ownership. industry,
gross profit for 2006, 2007 and 2008. construct a family tree to identify family members.

and obtain a list of current employees.

66

Participant Survey
All participants in the study were asked to complete the Participant Survey

(Appendix D). This survey includes a number of demographic variables as well as the
scales for Value Orientation, Satisfaction and the Network Communication items
(addressed in detail below).
Family Member Survey

While all participants received the Participant Survey. individuals who were
identified as “family members” received an additional set of 20 questions (Appendix C).
These additional questions were used to measure the Adaptability and Cohesion levels of
the family system. These two scales (Cohesion and Adaptability) were taken directly
from FACES III (Olson et al., 1985). The primary purpose of these two scales is for

business group comparisons in Specific Aim 2. 3 and 4.

Data Imputation

Once all data were collected, the researcher inputted each individual‘s
information into two separate computer programs. First. the researcher entered the
network communication items in Ucinet 6.0 (Borgatti. Everett & Freeman. 2002). This
program was used to produce the sociograms (using the Netdraw function) in specific
aim 2. This program also produced the centrality and density data used in specific aims l-
4. Secondly all of the data were put into HLM 6.06 (Raudenbush & Bryk, 2002). This
program was used to develop and test the models used in Specific Aims 3 and 4. SPSS
15.0 was also used to clean data and transfer centrality and density data from Ucinet 6.0

to HLM 6.06.

67

Ownership Validation
After completing the data imputation and analyzing all I 1 business’ data. the
researcher returned to the ownership of each business and reported general findings
regarding the overall study and the location of the owner’s business in comparison to the
study findings. This process added validity to the findings as all of the owners confirmed

the assessments of the researcher for their businesses.

Study Participants: Individuals

A total of 492 individuals completed the survey. These participant responses were
used for fitting the models in Specific Aims 3 and 4. While only 492 individuals
physically took the survey and provided actual responses. due to the social network items
in the survey it was possible to have individuals represented within the networks without
having that individual physically take the survey, and as a result the network data
represents 853 individuals. Therefore. the sociograms, centrality, and density data are
based on the sample population of 853, while the actual models in Specific Aims 3 and 4,
as well as the Value Orientation and Satisfaction variables are based on the sample of
492.

One of the largest contributors to the difference between the network N and the
sampled N is the economic downturn of 2008 and 2009. All of the businesses sampled
were in the process of reducing their number of employees. Therefore. while terminated
employees were not available to take the survey, sampled employees maintained
communication with these terminated employees and nominated them in the network

data. For example while person A (employed) took the survey. they may have nominated

68

person B (previously employed) in their communication network. Therefore person A
and B were included in the network N but only Person A was included in the sampled N.
The participants were divided among three subgroups: Owners, Family Members.
and Employees. Table 3.1 below represents this distribution. It should be noted that an
individual can qualify for two or more subgroups, as an individual may be an owner but
also a member of the owning family. and employed by the business. For example. all but
five owners also were family members. and approximately 60% ofthe family members

also were employees.

Table 3.1: Participants by Subgroup

 

 

 

 

 

 

 

Subgroup Frequency Percentage
Family 59 12.0 _ -*
Owner 38 7.7
-___Ena%1.9.\£e_ __ £16: 94-5 _,
Nee;

 

 

 

Subgroup Definitions
Ownership System .Itlembers. Individuals within sampled FOBs were considered a

member of the ownership system if they maintained a stakeholder position (own stock in
the business) and/or they hold a seat at either a governance board or board of directors.
Family System Members. An individual was considered a member of the owning
family system if he/she was related to the owner ofthe business or owning family
through blood marriage or adoption. Individuals also were considered a member ofthe
owning family system if the family system considered them a member ofthe owning

family.

69

Employee System Members. Individuals were considered members of the business
system if they receive compensation for services they provided for the FOB. Most
commonly these individuals were employees of the FOB.

Individual participants were given these definitions and first asked to self select in
or out of each group. The participant’s response was then verified by the business
owners, and cross checked with the roster of employee, family and owner members

obtained at the first interview with the business owner.

Study Participants: Businesses
The sampled businesses represented a wide variation of generation of ownership.
industry, revenue size. employee size and gender of primary owner. Table 3.2 below

represents the demographic variation of these businesses.

70

Table 3.2: FOB Participant Demographics

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Company Industry Owner’s Generation Employees Revenue
Gender of Three
Ownership Year
Average
(in
thousands)
1 Children Female 1 13 1,700
Education __ __
2 Residential Male 2 8 2.100
1__ _.__ -2.“ _-B:rno_ciflis____-_.__f__.___ __ _-___ __ __ __ _ .--___-_.
3 Agriculture Both 4 104 17,000
4 Wholesale Male 1 100 12.000
Distribution
5 Commercial Male 2 24 24,842
Real Estate ”_g * .______2___,
6 Whole Sale Both 2 500 89,876
H_____ _ _~*_”_H_"12i_str_ibution
7 Tourism Male 3 18 2,100
8 Funeral Male 2 20 4,867
Services
9 Children‘s Female 1 7 174
Arts/Ed. , A . ___W
10 Finance Male 2 8 10.500 .
1
11 Finance Male 2 9 15.424” i
1..“ ___ m _ - 2 ___-_,._ 2,_---,2-2______2 _-__- 1
Measures

Value Orientation Scale.

The value orientation scale has been normed using the American Family Business

Survey (Galvin et al., 2007) in Distelberg (2008). The actual items and associated alphas

are presented in Table 3.3 below. This scale was used in Specific Aims 2, 3 and 4. Using

the discussion of FOB values in Distelberg and Sorenson (2009), this scale can be used in

different ways depending on the level of analysis. For example. when evaluating an

individual’s score on this scale, one is actually measuring the individual’s perception of

71

 

his/her F OB’s value orientation. When evaluating this scale as a mean of a subgroup. one
is measuring the subgroup‘s perception of the F OB’s value orientation. When this scale is
averaged across an entire FOB sample the score is considered the actual value orientation

of the FOB.

Table 3.3: Value Continuum Items

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Likert Scale Response Factor
Paired items Loading
1 2 3 4 5 6 0t

. 7 ______ _..__ __
A manager’s qualifications Family members are given . 0.827
(education, experience, etc.) preference in hiring and
are the only characteristic promotion decisions
considered in hiring and
promotion decisions __
All employees are Family members are paid more 0.812
compensated (excepting than non-family members in
dividends) based solely on comparable positions
their position and performance
This company is a business. This company is a family. which .71 l
which happens to employ happens to be in business
people from the same family together
The owner(s) primarily get The owner(s) primarily get .826
financial and professional satisfaction from working with
satisfaction from this business: family members; the financial
working with family is a bonus rewards from the fir are a bonus

Cronbach Alpha for Scale 0.805
N 638

 

 

 

Satisfaction Scale.
Since the beginning of the FOB field. research has focused on understanding how

FOBs obtain success or achieve satisfaction. Many studies have used self assessments of
satisfaction as an indication of success (Danes et al.. 1999; Danes et al.. 2002; Olson et
al., 2003; Zody et al., 2006). In these studies a scale is created based on a select number

oflikert scale items. In most cases these items reflect a combination of satisfaction with

72

the three systems in the FOB. Unfortunately there does not currently exist an accepted
assessment for satisfaction. Therefore the researcher created a scale based on the items
commonly used in the literature.

Prior to administering this scale it was pilot tested with the 20 individuals with
experience in FOB issues (i.e. 16 business owners. 2 family business organization leaders
and 2 family business researchers). These individuals all believed that the seven items in
the scale accurately measured satisfaction within a family business, therefore providing
face validity to this scale. This scale was used in all four Specific Aims. The individual
items as well as reliability estimates are presented below in the Dependent Variable
section.

Network items.

These items were used directly and indirectly to address each one of the specific
aims. More specifically. these items were used for the Dependent Variable in Specific
Aim 1 and the sociograms in Specific Aim 2, and the density and centrality values in
Specific Aims 2, 3 and 4 were calculated based on these items. These items reflect the
social network analysis portion of this study and as were constructed through an
exploration of social network literature.

The typical fashion of evaluating relationship ties in networks through SNA
involves asking respondents (nominators) to nominate individuals with whom they have a
relational connection (Wasserman & Faust. 1994). Most surveys ask the nominator to
either choose from a list ofindividuals or recall individuals from their memory.

Recently, two important ideas have developed in the SNA literature that relate

directly to this study population. First, Marsden (2005) has noted that the typical methods

73

of asking nominators to identify a nominee are ineffective in densely packed groups
(groups with a great deal of interaction on a regular basis). FOBs could be defined as a
dense group in terms of interactions and therefore Marsden (2005) recommends using a
more precise item, in this case asking participants to nominate interactions that are
“meaningful” and happened within a finite period of time (three weeks was used for this
study). Secondly, SNA researchers have found that if given the ability to choose the
number of nominees reported (instead oflimiting the nominations to a specific number of
nominees), nominators will average between three to five nominees (Marsden. 2005).
Therefore it is more efficient to provide space for up to six nominees.

The three network items are:

1. In the last three weeks whom have you had a meaningful conversation with

regarding the [INSERT OWNING FAMILY NAME] family. or discussed
issues specifically related to the owning family?

2. In the last three weeks whom have you had a meaningful conversation with
regarding the day to day functions of the business (e. g. job responsibilities.
problems with coworkers. production changes. time off)?

3. In the last three week whom have you had a meaningful conversation with
regarding the overall strategy and future of the business (e.g. strategic
planning, succession planning. initiating or changing governance boards)?

Each of these items reflects the division between family (item 1), employee (item

2) and ownership (item 3) communication patterns. Theoretically. if the Three Circle
Model is correct. employees should have little to no values recorded for items 1 and 3.
FACES III.

The Family Member Participant survey is the FACES III assessment. FACES 111

contains the Cohesion and Adaptability Scales which were used as independent variables.

74

This survey is derived from the Circumplex model (Olson et al., 1979a; 1979b). The
Circumplex model has been revised four times (Olson. 2000). The FACES 111 format is
the shortest of these formats (20 items) and has the most research validating its reliability
and validity. Table 3.4 below shows the individual items. scales and associated alphas.
There are two scales within FACES 111. one for family adaptability ((1 = .62) and one for

family cohesion (O. = .77).

75

Table 3.4: FACE III

 

Items for FACES nr "7

 

Cohesion Items

(1 = .77
X = 39.8 SD = 5.4

 

.\J

 

10.

Family members/eel very close to each
other

Family togetherness is very important
Supportiveness

Family members ask each other/or help
Family members consult otherfamily'
members on their decisions

Family Boundaries

Family members/eel closer to other

family members than to people outside

thefamily

We like to do things withjust our
immediate family

Time and Friends

Family members like to spend free time
with each other

We approve of each other 's‘friends
Interests and Recreation

When ourfamily gets together. for
activities, everybody is present

We can easily think of things to do
together as afamil y

 

76

 

Factor Loading

.60

.47

.51

.48

.49

.39

.69

.43

.54

.43

 

 

Table 3.4 con’t

 

FACES III

 

Adaptability Xa :le SD 4 7

 

Factor Loading 2

 

 

l I. Different people act as i i .35
leaders in ourgfamily

 

I 2. It is hard to identify the .38
leader(s) in ourfamily
Control

 

13. The children make the .34
decisions in our family

 

l 4. In solving problems, the .37
children ’s suggestions are

followed
Discipline

 

 

I 3. Children have a say in their 7 .48

discipline

 

l 6. Children and parents 7 .37
discuss punishment together
Roles and Rules

 

l7. Ourfamily changes its way ii i i T45
of handling tasks

 

I 8. We shift household .38
responsibilitiesfrom person
to person

 

19. Its hard to tell who does .34
which household chores

 

 

 

 

20. Rules change in our family .36

 

Dependent Variables
Value Orientation
Value orientation is a continuum of family businesses based on whether the FOB
system values the family side or the business side of the entire FOB system. Previous
research using the American Family Business Survey (Galvin et al.. 2007; Distelberg.

2008) revealed that family businesses do vary in regard to their preference for the

77

business or family side of the FOB system. The scale in this study is unique in that it is
the first time all members of the system have been measured. For the sample population
table 3.5 shows that the total Cronbach Alpha for the scale is .698, and each of the four
items load equally well using Cronbach Alpha Factor Analysis. Table 3.6 shows the

distribution of this variable and table 3.7 provides descriptive statistics.

Table 3.5: Value Continuum Reliability

 

 

 

 

 

 

 

 

Item Cronbach Alpha
Value Item 1 .612
Value Item 2 .603
Value Item 3 .651
Value Item 4 .660
Total Alpha .698
N 492 -

 

 

 

78

Figure 3.6: Value Continuum Histogram

Frequency

Table 3.7: Value Continuum Descriptive Statistics

60

 

 

0.00 5.0 10.0

15.0

Mean = 14.1
Std Dev = 5.24
N = 486

 

20.0 25.0

Value Orientation

 

 

N Valid 486

Missing 6
Mean 14.1379
Std. Error of Mean .23752
Median 15.0000
Mode 16.00
Std. Deviation 5.23623
Skewness -.025
Std. Error of Skewness .111
Kurtosis -.445
Std. Error of Kurtosis .221
Minimum 4.00
Maximum 28.00

 

 

 

79

 

Satisfaction

The satisfaction variable was constructed using seven questions on a likert scale

of 1-9. Table 3.8 shows the items and the associated factor loadings. Table 3.9 provides a

visual representation of the variance, and table 3.10 provides the descriptive statistics for

the satisfaction scale. Taken together this is a very strong scale with total alpha for the

scale of .91.

Table 3.8: Satisfaction Scale Items

 

 

 

 

 

 

 

 

 

 

 

Item Cronbach
Alpha
1. Your level of satisfaction with your involvement with the .894
business
2. Your level of satisfaction with the ownership/management of .887
the business _
3. Your level of satisfaction with the employees within the .907
business
4. Your level of satisfaction with members of the owning family .894
5. Your level of satisfaction with the amount of conflict i .896
throughout the business ‘ ;_____ _ _ , __ >__ __ ___ a
6. Your level of satisfaction with the future direction ofthe .896
business ___ _ _ -
7. Your level of satisfaction with how problems are solved within .890
the business
Total Alpha .909
N 484

 

 

 

80

Figure 3.9: Satisfaction Histogram

 

50
Mean = 47.4
Std Dev = 10.96
40 N = 484

30

Frequency

20

10

 

 

 

0.00 20. 00 60.00
Satisfaction
Table 3.10: Satisfaction Descriptive Statistics

N Valid 434

Missing 3
Mean 47.4421
Std. Error of Mean .49300
Median 49.0000
Mode 63.00
Std. Deviation 10.95590
Skewness -590
Std. Error of Skewness .11 1
Kurtosis _14g
Std. Error of Kurtosis .222
Minimum 9.00
Maximum 63.00

 

 

 

81

 

 

Independent Variables
Cohesion

The Cohesion scale is constructed using the ten items from FA (”ES [I] (see
description above). Since only family members reported on Cohesion there are only 58
individual values. These values. when used in the analysis section, are aggregated across
each owning family system, providing one score for each family system.

Cohesion is a measure of the closeness and distance within a family. Higher
scores indicate a family system that is too close (often referred to as enmeshed, Olsen et
al., 1985). Low scores indicate a family system that is distant. Three decades of research
using this scale suggests that functional family systems score between the two (X = 39.8
SD 2 5.4). For the sample population in this study, the )T =~ 40.67 with SD 6.5. A T-test
indicates that these values are statistically similar to the National Average (1 =1.02. 4f;
57.p : .31). This suggests that on average the families in this study do not vary
significantly from the general population in closeness. Tables 3.1 l and 3.12 provide the

visual variance in the sample population as well as descriptive statistics.

82

Figure 3.11: Cohesion Histogram

 

   
   

 

 

Mean = 40.7
Std Dev = 6.51
6 - N = 58
>. 4 ‘
0
C
0’
3
u-
2
IL
2 -
0 -
25.0 30.0 35.0 40.0 45.0 50.0
Cohesion

Table 3.12: Cohesion Descriptive Statistics

 

N VW 58—
Missing 434
Mean 40.6724
Std. Error of Mean .85571
Median 41.0000
Mode 40.00‘I
Std. Deviation 6.51688
Skewness -.307
Std. Error of Skewness .314
Kurtosis -1.016
Std. Error of Kurtosis .618
Minimum 28.00
Maximum 50.00

 

 

 

 

83

Adaptability
The Adaptability scale is constructed using the corresponding ten items from

FACES 11] (see description above). Since only family members reported on Adaptability
there are only 57 individual values. These values when used in analysis are aggregated
across each owning family system providing one score for each family system.
Adaptation is a measure of the flexibility and rigidity within a family. Higher
scores indicate a family system that is too flexible (often referred. to as chaotic Olson et
al.. 197%). Low scores indicate a family system that is too rigid or resistant to change.
Three decades of research using this scale suggests that functional family systems score
between the two (X = 24.1 SD = 4.7). For the sample population in this study the X =
28.14 with SD = 5.2. A T-test indicates that the sample population families are
significantly different from the National sample (I -—'5.89. 91/4 56. p <1 0.001). This
suggests that families in family businesses are more adaptable than the average family
system. Tables 3.13 and 3.14 provide the visual variance in the sample population as

well as descriptive statistics.

84

Figure 3.13: Adaptability Histogram

 

12—

  
   

Mean = 28.14
Std Dev = 5.18
N = 57

  

10-

Frequency

 

 

10.0 15.0 20.0 25,0 30.0 35.0 40.0
Adaptability

Table 3.14: Adaptability Descriptive Statistics

 

 

N Valid 57

Missing 435
Mean 28.1404
Std. Error of Mean .68617
Median 29.0000
Mode 29.00
Std. Deviation 5.18045
Skewness -.691
Std. Error of Skewness .316
Kurtosis .654
Std. Error of Kurtosis .623
Minimum 14.00
Maximum 39.00

 

 

 

85

Centrality (Reachability)
In order to compare one business’s communication structure to another in

Specific Aims 2. 3 and 4. a network measure was used to quantify the patterns of
communication. A number ofeentrality measures are available for this comparison. from
the most basic, “Degree Centrality” (Wasserman & Faust, 2007), to “Reach C entrality”
(Hanneman & Riddle. 2005). Degree Centrality is a simple count of all ofthe connections
to and from individual i. This measurement wOuld be helpful and is used to measure
communication density in Specific Aims l and 2, but in practice it is easily manipulated
by mediating factors (Hanneman & Riddle, 2005). For example, in the survey.
individuals were provided enough space to nominate six individuals for each of the three
network items. Therefore. if individual A took his/her time to think about six
nominations but individual B was rushed and only thought of two individuals. individual
A would have a higher rating than person B simply because of the effort put forward on
the survey. To some degree this problem can be solved by looking at “In-degree" versus
“Out-Degree“.

In-Degree is a sum ofall the nominations to individual i. by all individuals i'.
where as Out-Degree is the sum ofnominationsfi'om individual i to all individuals i'.
Therefore a higher In-Degree for person i can be conceptualized as being a person that
many people talk to, while a higher Out-Degree conceptually means that person i talks to
a lot of others.

While In-Degree has fewer issues in regard to sampling error. it only accounts for
person i’ nominations and does not account for the position of individual i'. For example,

In-degree centrality measures how many connections individual i has but does not

86

consider “to whom” individual i is connected (e.g it is different to be connected to four
employee friends than it is to be connected to one owner).

One measure that does account for whom an individual is connected. is “Reach
Centrality.” Reach centrality is the sum of all ii’ connections along all possible geodesic
paths for individual i, and it weights these connections by how many steps away the
connection is from individual i. For example, if i is connected to j andj is connected to k.
then i is connected to k throughj. Therefore ifj is a highly connected person then it is
more advantageous to be connected to j than have multiple connections to other’s with
few connections. Furthermore, Reach Centrality weights each step. so the number of
connections from i to j is divided by 1, but the connection from i to k throughj, is two
steps and therefore divided by two (weighting is equal to the l/n where n = the number of
steps). Conceptually this measures an individual’s reach to all individuals in the system.
or similarly one’s access to all individuals in the system. A higher rating of Access
(Reach Centrality) means an individual can access all the individuals in the system better
than an individual with a lower Access (Reach Centrality).

In-Degree and Reach C entrality are used to quantify the four communication
patterns for each business. Each business produced four communication networks:
family issues (Family Access), and employee issues (Employee Access), ownership issues
(Ownership Access), and the sum of these three networks creates the “Total
Communication Network” (Total Access). In each case the standard Reach Centrality was
used rather than the Normalized value which is common in comparison of multiple
distinct networks (Hannamen & Riddle, 2005) because the normalization of the standard

numbers is accomplished by dividing the standard value by the total number of geodesic

87

vertexes. The same thing can be conceptually accomplished by using the “group-mean
centering“ function in HLM 6.06.

Family Access. A matrix of interaction between person i andj was constructed for
each business by asking each participant to nominate up to six individuals associated with
his/her business that he/she talks to about issues concerning the "owning family“. Once a
matrix for a business was constructed. a sociogram was created and Reach Centrality
measures recorded. Family Access is then the individual’s Reach C entrality for the
Family Communication matrix. Table 3.15 and 3.16 provide the histogram and
descriptive statistics for Fan-wily Access across all businesses. In table 3.15 there are a
large portion of individuals who have a low value for F umily Access. This makes
conceptual sense, as not all individuals in the business would have access to family
communication. This positively skewed histogram is a problem for the normal
distribution assumptions of HLM and will be addressed in the analysis section. (A similar
problem exists for Owner Access).

It should be noted for this measure as well as the other "access" measures. the
value of the measure is based on the larger network (or the N of 853). Even though only
the individuals who participated in the survey were recorded for modeling purposes. the
survey process brings in individuals even if they do not take the survey themselves. This
is the primary reason for only including businesses that can provide access to 70% of
their employees. By sampling 70% of the employees we can be relatively confident that
the sampled social network of that business is a fair representation of the business and

there are not structural holes due to sampling error ((Wasserman & Faust, 1994).

88

Reach Centrality requires a symmetric matrix and therefore each matrix had to be
made symmetric by taking the larger of the column or row values of the matrix.
Substantively this is acceptable because if communication happens from person A to

person B, it also happens from person B to person A.

Table 3.15: Family Access Descriptive Statistics

 

 

N Valid 492

Missing 0
Mean 5.7167
Std. Error of Mean .29874
Median 1.0000
Mode 1.00
Std. Deviation 6.62646
Skewness 1.258
Std. Error of Skewness .110
Kurtosis .427
Std. Error of Kurtosis .220
Minimum 1.00
Maximum 26.83

 

 

89

 

Figure 3.16: Family Access Histogram

 

300 '1

Mean = 5.72
Std Dev = 6.63
N = 492

200

Frequency

100

 

 

 

 

0.0 5.0 10.0 15.0 20.0 25.0 30.0

Family Access

Ownership Access. For each business a matrix of interactions between person i
and i ' was constructed by asking each participant to nominate up to six individuals
associated with their business with whom they talk about issues concerning the future
direction or strategy of the business, typically conversation that the owners would have
with individuals. Once a matrix for a business was constructed, a sociogram was created
and Reach Centrality measures were recorded. Ownership Access is then the individual’s
Reach Centrality score for the Ownership Communication matrix. Table 3.17 and 3.18

provide the histogram and descriptive statistics for Reach Centrality across all businesses.

90

Table 3.17: Ownership Access Descriptive Statistics

Frequency

 

 

N Valid
Missing

Mean

Std. Error of Mean

Median

Mode

Std. Deviation

Skewness

Std. Error of Skewness

Kurtosis

Std. Error of Kurtosis

Minimum

Maximum

 

492

8.3704
.48059
1.0000

1.00
10.65994
1.404
.110
1.013
.220
1.00
46.52

 

 

Table 3.18: Ownership Access Histogram

300

200

 

.l

 

Mean = 8.37
Std Dev = 10.66

 

10.0 20.0 30.0

Owner Access

91

 

 

40.0

50.0

Employee Access. For each business a matrix of interaction between person i and
i’ was constructed by asking each participant to nominate up to six individuals associated
with his/her business with whom they talk to about issues regarding the day to day
function of the business. Once a matrix for a business was constructed a sociogram was
created and Reach Centrality was recorded. Employee Access is then the individual‘s
Reach Centrality for the Employee Communication matrix. Table 3.19 and 3.20 provide
the histogram and descriptive statistics for Reach Centrality across all businesses. There
is a noticeable bimodal distribution of this histogram. This suggests that the values are
non-randomly varying for Employee Access. It could be hypothesized that owners have a
higher access on average compared to family and employees. Either way this bimodal
distribution violates the normal distribution assumptions within HLM and will be

addressed in the analysis section.

Table 3.19: Employee Access Descriptive Statistics

 

 

 

N Valid 492

Missing 0
Mean 31.1468
Std. Error of Mean 1.16007
Median 23.0095
Mode 1.00
Std. Deviation 25.73166
Skewness .348
Std. Error of Skewness .110
Kurtosis -1.310
Std. Error of Kurtosis .220
Minimum 1.00
Maximum 91.51

 

92

 

Figure 3.20: Employee Access Histogram

 

100 .-
Mean = 31.15
80 ‘1 Std Dev = 25.73
N = 492

Frequency

 

 

 

0.0 20.0 40.0 60.0 80.0 100.0

Employee Access

Total Access. Once the three previous matrices had been constructed the final
matrix (Total Communication) was constructed by summing the previous communication
matrices. Rather than being a binary matrix, each cell in the matrix has a strength
weighting of 0-3. A score of three would mean that the relationship from i to i ' exists
across all three communication groups (family, employee, and owner). For example, if
person A talks to person B about the family, the employee issues, and ownership issues,
cell A-B would equal 3, but if Person A only talks to person B about employee issues,
then cell A-B would equal 1. This strengths weighted matrix is used in Specific Aim 2 for
the density measure, but the tests in Specific Aim 1 required that this matrix be a binary

symmetric matrix. Therefore the weightings were removed (3 and 2 become 1, 0 =0),

93

and the matrix was made symmetric by taking the larger value of either the row or
column.

Tables 3.21 and 3.22 show a bimodal distribution (similar to the bimodal
distribution of the Employee Access variable). Due to this bimodal distribution it cannot
be used as is in the HLM models. But the individual block modeling techniques and
individual centrality scores can be used in Specific Aims l and 3 due to the process of

examining one business at a time rather than the total group.

Table 3.21: Total Access Descriptive Statistics

 

 

N Valid 492

Missing 0
Mean 39.7876
Std. Error of Mean 1.39932
Median 30.1800
Mode 1.00
Std. Deviation 31.03841
Skewness .299
Std. Error of Skewness .110
Kurtosis -1.262
Std. Error of Kurtosis .220
Minimum 1.00
Maximum 116.35

 

 

94

 

Figure 3.22: Total Access Histogram

 

80 -
Mean = 39.8
Std Dev = 31.04
N = 492

Frequency

 

 

 

0.0 20.0 40.0 60.0 80.0 100.0 120.0

Total Access

95

Data Analysis Procedures

The following is a step by step process for exploring the Specific Aims and
hypothesis of this study. This analysis process follows two phases. Phases 1 addresses
Specific Aims 1 and 2 while Phase 2 addresses Specific Aims 3 and 4.

Phase 1: Step 1 explored the data’s ability to support the Three Circle Model. An
ANOVA like process will give a model fit comparison for each business. The purpose of
this step is to support the hypothesis that the Three Circle Model does not fully account
for the flow of communication within FOBs.

Phase 1: Step 2 begins with an exploration of seven research questions across the
1 1 businesses using the measured values for Value Orientation, Sc‘tt‘is/action. Cohesion
and Adaptation. This step includes a case summary of each of the 11 businesses and also
a presentation of their network structures. These structures will be measured in different
ways in an effort to show each business‘s ability to support or failure to support each of
the research questions. Conclusions from this step will generate hypotheses that will be
tested in Phase 2.

Phase 2: Step 3 will fit a multilevel model for predicting satisfaction within and
across businesses. This step uses the information from Phase 1 (Steps 1 and 2). and
builds a HLM to test the validity of the findings within the previous steps.

Phase 2: Step 2 will fit a multilevel model for predicting an individual’s value

orientation perception.

96

CHAPTER IV: RESULTS

Phase 1: Step 1
Specific Aim 1.‘ Evaluate the validity of the Three Circle Models assumptions/or

communication structures within FOB systems.

H1: The Three Circle Model does not fully account for all possible variations

in FOB communication structures.

To test this hypothesis the Total Communication matrix was created by summing
the three network communication matrices (Family ("ommunication, Employee
Communication, and Ownership Communication). To test this hypothesis a model was
fit for each business. This model is a block modeling technique where an “Expected”
matrix is created by randomly placing communication ties within a matrix. A new matrix
is formed by correlating this “Expected” matrix with the “Observed” or Total
Communication matrix. This new “autocorrelated“ matrix is considered the Dependent
Variable and the subgroups within the Three Circle Model are regressed on the
autocorrelated matrix. C onceptually the fit of this model tells us whether the subgroups
explain the communication patterns, or whether members of the subgroup prefer to talk to
each other or across subgroup (ANOVA like). The Three Circle Model (Figure 1.1,
Chapter 1) provides the subgroupings used in this test. Specifically family member = 1.

Owners = 2, Employees 3, Family members who are Owners are = 4, Owners who also

97

are Employees : 5. Family members who also are Employees = 6. and individuals who
fit into all three groups = 7. This model was fit for each business.

Table 4.1 shows the adjusted r-squared and Chi-Squared significance for each
business. It is clear from this exercise that the Three Circle Model does have some
explanatory power (e.g. companies 3., 4. 6. 8 and 10 are significant). It is also evident
from this exercise that even when the model is significant, it does not explain a great deal
of the variance in communication patterns. For example, the model fits for company 8.
but the adjusted r-squared is 0.018. This means that to some degree individuals within
subgroups talk to each other more than they talk to individuals outside of their subgroup.
but these subgroups only explain 1.8% of the variations in communication patterns. There
are two exceptions to this finding. The Three Circle Model fits well (statistically
significant) and explains a fair amount of the variance (r—squared > 0.10) for companies 3
and 10. This model is a very good fit for Company 10 (adjust r-squared 0.42). But if we
also consider the average level of satisfaction and the fit of the model for companies 3
and 10, we notice that both of these companies have significantly lower levels of

satisfaction.

98

Table 4.1 Three Circle Model Fit with Total Communication Matrix as Dependent

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Company Three Circle Model Satisfaction
FIT
R-squared (p) )T 2 47.7
1 0.006(0.61) 55.7(7.7)
2 0.000(0.99) _ 49.5(8.1)
.3 j- 0.10(<0.001) 46.3(1 1.8)
4 0.003(0.02) 49.4(10.7)
5 <0.001(0.74) 57.5(6.7)
6 <0.001(<0.001) 46.4(10.7)
7 0.005(042) 43.3(7.8)
8 0.018(0.04) 46.2(9.8)
9 <0.001(.43) 56.9(6.4)
H“_10 0.42(0.002) _ _ M3571 14.3)”
' l _ __ _- _m __Q-_90_3(0-_30_)_1_.__-__ __4. 19.11241

 

Taken together, the Three Circle Model is relevant for communication patterns.
but it does not explain a lot of the variance in communication patterns. It is likely that
other variables account for a greater percent of the variance. Furthermore. there may be a
negative relationship between the fit of the model and satisfaction. These findings
suggest that the next step in phase 1 will be helpful in adding understanding to how and

why communication patterns vary within businesses.

Phase 1: Step 2
Specific Aim 2: Expand the Three Circle Model validity through the inclusion offamily
dynamics. value orientations. and boundary creation.
The process of reaching this specific aim begins by imputing the network data for
each business into a computer program. Ucinet 6.0 (Borgatti. Everett & Freeman. 2002).

This program allows the user to construct N X N matrix for communication networks.

99

For this study four matrixes were created for each business: one for the family
communication, one for employee communication, one for ownership communication
and one for the total (or sum) of the previous three matrices. This program also
generated the social network measurements. Netdraw (Borgatii. 2005) (a function within
Unicet 6.0) was used to construct the visual sociograms for each sampled businesses.

The following section will provide a brief summary of each of the 1 l FOBs. For
each business there is a short narrative that discusses the pertinent findings (in
relationship to the hypotheses for this section). These findings reference the included
sociograms and summary table, which follow each narrative.

Each business will have two sociograms (one for the Total Communication, and
one for the Family Communication). The employee and ownership sociograms are not
presented below, they are included in the appendix (APPENDIX F).

Also. each FOB summary will include a table ofthe company level data. The
table includes the values for Satisfaction, Value Orientation and the Cohesion and
Adaptability of the owning family. This table also includes a few social network
measurements such as each subsystem’s density within the total communication network.
and Joint Count measurement of the family communication network. The Satisfaction,
Value, Adaptability and Cohesion scales were developed using a series of items on the
survey and discussed in detail in Chapter 111.

Since SNA measurements can vary. it is important to summarize how each
measurement was calculated in this phase. F irst, density conceptually is a measure of the
degree of communication in a given group. This measure is used on the total

communication network, which is a strength based matrix (values range from 0-3).

100

Therefore if everyone in the family talked to everyone in the family on all three
communication measures (family. employee and ownership communication), then the
density of the family subsystem would be D = 3. Conversely, if there was no
communication between any family members on any of the three communication
networks the density for the family subgroup would be D = 0. The equation used to

determine the density is:

D _ L L = the number of 1ines(connections) present in
_ n(n — 1.) the subgroup
n = number of individuals within the given
subgroup

Second, the Joint Count measurement is similar to an ANOVA method. in that it
measures variation within and across groups and provides some comparison of whether
the variation can be attributed to a subgroup. Unlike ANOVA. it does not use variance
components. but rather it compares the actual count of interaction in a measured matrix to
a randomly generated matrix of interactions. The process begins by creating a matrix with
random interactions. This random matrix (or the Expected matrix) is compared to the
measured matrix of interactions. The measured matrix, or Observed matrix. is each
FOB’s Family Communication matrix. By comparing the Expected to the Observed
matrices we can make some judgment about whether, and to what extent. communication
exists within a subgroup. Two numbers are presented in the table from this process. The
first is the difference between the Observed and Expected connections for each group (e. g
Observed — Expected). Therefore, if the number is positive. and high for the family
group, we would say that the interactions within the family group are larger than could be

expected by random. or conceptually. there is a group called family. it is a meaningful

101

category because if it were not we would see the Observed value be equal to or less than
the Expected values. The difference between the Observed and Expected matrices can
also be tested with a Chi-squared test of significance. In other words. is the difference
between the observed and expected connections is larger or smaller than could be
expected by chance alone (Chi-squared uses 3 dfi one for each group. family. family-
nonfamily interaction. nonfamily). The second number presented for each group in the
table is the observed over the expected ratio. Therefore, if the ratio is a 3 for a family in
company A, we would say that being a family member in company A provides three
times more interactions than seen by random. This allows us to compare the difference
across companies because if the ratio for Company B is a 10, then we conceptually can
say that the family members in Company B interact more than family members in
Company A (this can also be verified by the density measures discussed above).

This process allows hypotheses to be generated about family communication in
each business. For example a null hypothesis would be that there is no flow of family
communication from the family subgroup to the nonfamily subgroup. We can measure
this by looking at the Joint Count analysis of the family-nonfamily interaction group. In
this case the null hypothesis would be supported if there is a low (negative) value for the
interaction group (and it would be statistically significant, indicating that this negative
value is lower than we would expect by chance alone). F urthermore, if there is an
extremely rigid boundary for family communication we would see no connections

between the family and nonfamily group. or a 0 for the ratio (O/Expected = 0).

102

Company Summaries

COMPANY 1:

This business is a relatively young business ( l 6 years) in its first generation of
ownership. The female owner (1001 from figure 4.2.1, and 4.2.2) provides education and athletic
training for young children. The owner’s daughter ( 1002) and sister (1009) are employed by the
business also. On average there are 13 employees within the business. The three year revenue
average is 1.7 million. Therefore we would conclude that this is a relatively small business.

Within this FOB there have been discussions ofthe daughter (1002 in figure 4.2.1)
taking over the second generation of ownership in partnership with a valuable program manager

(1007). But the owner believes that a succession is not likely for at least another 5-10 years

103

Figure 4.2.1: Company 1: Total Communication

/.-1011

.V: 1016 "' Tr l

t _
Owner i // l
\ Mfr/FOWRKR ,’ .
\. ”Fame-m“ \ T a“; ‘ .
if”! (I If \ ’3'! Rmfifi 1004
1 i l
\\ t
’ \X l

 

 

 

- Employee
{5* Not in business as employee or owner

0 Family

D Not family

 

 

 

 

104

Figure 4.2.2: Company 1: Family Communication

}>1014

[11018

‘1' 1003

 

 

 

“11004

[1015

 

 

- Employee
Not in business as employee or owner

0 Family

[:1 Not family

 

 

105

From table 4.2.3 we see that this business is closer to the business side ofthe continuum
(9.2), but the ownership and family members see it even closer to the business side in comparison
to the employees (9.6 versus 6.0 and 6.5 respectively). Overall everyone in this business is very
happy (satisfaction 55.7 compared to the average 47.7, t-test significance < 0.001). There is a
difference of opinions in satisfaction from the owner, family members and employees. with
employees being slightly less happy than the owner and the family members. Even so. these
employees are happier than the average employee ofa family business.

In regard to the family dynamics, this family is very close (cohesion = 43.5). They also
are more adaptable than the average American family (adaptability = 26.7) but slightly less
adaptable than the families in this study. What is interesting about the communication patterns of
this business is that the density of family communication and employee communication across all
three types of communication is rather low (D = 1.0 for both). This suggests that there is not a
great deal of within group communication for this FOB (the ownership group density is not a
measureable number due to there being only one owner).

When we look specifically at the family communication we see that family members
communicate between themselves (10.0. p < 0.001), and family communication exists between
the nonfamily group, but there is a semi—permeable boundary between the two groups.
Substantively this means that family communication does not flow freely to the nonfamily
groups, although it is not a complete cut off as illustrated by the ratio of 0.5 (close to 0 but > 0)
and the family communication sociogram (figure 4.2.2). In this sociogram we see that 1001,
1002 and 1009 (family members) do communicate to the employee group, but there is a visual
symmetry to this picture, with all family members on top and all nonfamily members on the
bottom. This means that family members talk to each other. employees talk to each other. but in

comparison. there is less between group communication.

106

In summary. this company does have a boundary for family communication, but
it is not a cut off. Key family members 1001 (owner), 1002 (daughter) and 1009 (sister)
pass family communication on to the nonfamily members. The danger is that the family
group does not pass a lot of family communication to the employees, but the employees
communicate about the family (2.9 with ratio 0.63). This could lead to a cutoff. and more

than likely incorrect information regarding the owning family circulating within the

employee (or non-family) group.

107

88% 3 3:35.433 23 S can 633.36 B: u a:

638:“. 329:3 8 66.2556.» 8:33:34 Reine

 

 

 

 

 

 

 

~86 V R a... Sums macaw 6.8563 8 contemseo oozeegzmma >83 u a
36 v Q s :62: “.8563 an S eoflxemfieo eeeeeSnma :63 u a
345.6
as: .3 $293.2
2. 6:30 35 v a o: 5.5% ma
6.33” 6:35.33
:2 ...... a .2 3E5 3 8635 cacao". ._<zo_e<z
c3 .. _ a- 920.25 ~85 u a $65.3 mun—25. E
as am aghaoz 3 , beam 8.33.. 58.66
Aflovowam mean—-30 8:80 Steam—55550
Sign: 33. main:
5332:5850 man—ah macaw—am we 36:6: mom—:25: bait
6.0.8 o 63:8 2: :45. ago
.28 o 336% seamen been
m:
a 88”.: 8.2%.? 8.835
:5...
a 9.5.3 5.3.5. 8:356
68 o 833 semi 6:30
93 a 833 2.33: base
m:
6 3.36.5 can: 8635
8...:
a 3:? 39.: .8825 83>
See 882$: 882$:
£8
.ufi — mean—.50 =a 3.9.34 ecu—=55 E8—

 

 

..Ddfifism ~ 33:80 mdé 2%...

108

COMPANY 2:

This business is 25 years old. Its primary service is residential remodeling. This
business recently (January of 2008) completed a transfer of ownership from the first
generation of ownership (individual 2005 in Figure 4.2.4 and 4.2.5) to the second
generation (a son) (individual 2001). This business employs eight individuals (4 of
whom are family members). The founder and his wife are still employed by the business
and provide administrative and sales support. The current owner‘s wife is also employed
by the business as a sales representative. The three year average revenue for this business
is 1.9 million. As this is a business directly affected by the economic issues of 2009. the
employee count was reduced from l5 to 8 from January 2008 to March 2009.

Important to note in the total communication sociogram and the family
communication sociogram is that the new owner (2001) controls almost all of the
communication between family members and non-family members. This suggests that a
boundary for family communication exists. Also important to note is the role of
individual 2002. This non-family employee is communicating with three individuals
outside ofthe business (2010. 2009 and 2008). When this was explored in greater detail.
it was found that these individuals are contractors that are used regularly for the business
(e.g. electricians. plumber). This seems like a natural and innocent path of
communication until we look at the family communication sociogram (figure 4.2.5). In
this picture. employee 2002 is talking to the same contractors about the owning family.
For this study we did not ask what was being communicated. rather with whom one
communicates. Therefore this communication path may be innocent as well. but seeing

that there is a relatively strong boundary for family communication between the family

109

members, and non-family members it is possible that the information received by the

contractors may not be completely accurate.

110

Figure 4.2.4: Company 2: Total Communication
2006

2005

2010 2007

2002

2011

 

- Employee

- Not in business as employee or owner
0 Family

|'_'] Notfamily

 

 

 

111

Figure 4.2.5: Company 2: Family Communication

2005

2010

2003

a... ’

 

2002 2%

 

 

/
[12008
2004
2007
- Employee
- Not in business as employee or owner
0 Family
D Notfamily

 

 

112

According to the measurements in table 4.2.6, this business is closer to the business side
ofthe value continuum (l 1.4 p <0.001). There is some disagreement on this issue with employees
reporting a value orientation higher than the family (13.7 versus 7.0). In other words, employees
tend to view the business as closer to the family side ofthe value continuum in comparison to the
family members. Also, while everyone in the business is rather happy (satisfaction 49.5, p =
0.001 ). employees are less satisfied than the owner and the family. but they are as happy as the
average employee in a family business.

This family is not as close as the average FOB family or the average American family
(cohesion = 35.8, p <0.001). This would lead us to believe that cut offs within the family exist.
This is confirmed by the owner who identified an older brother who used to be employed by the
business but was let go and has no contact with the family since that time. While this family is not
very close, they are adaptable (32.8 p < 0.001 ). Therefore they are not adverse to change. This is
evident by the relatively easy transfer of ownership from the founder to the son.

While the family is not very close. their communication is rather good as the family
member density ofthe total communication is 2.1. There is a developing cut off for family
communication as noted in regards to the family communication sociogram (Figure 4.2.5). This
is further verified by the Joint Count analysis for family communication. This measure tells us
that the family communicates between family members, the employees communicate between
themselves, but there is relatively little between group communication (-4.3, p #001), the ratio is

dangerously close to 0 (0.19) suggesting a rigid boundary for family communication.

113

macaw 2% ~§E>~VS one 8 oak 33:95 3: u a:

£353. 320.5»: 8 :oflbmfioo mogomxwzmma Reine

 

 

 

 

 

 

 

3e 6 v Q 2.. =88 $22M goofing S zoflsamsoo oogombzwmn 284 u a
3 6 v a ._ S85 ”outage :6 S coatumsoo oo§o$=m~n 23¢ u a
Cévofivm
Ammzam 12225.2
2 .055 83. v a 0: SEE m...
3». £33.52.
ca :3 EEE :.~ 8.835 among q<zofi<z
2 .c an? 380.25 Sod v a $83.3 mqm2<m mn—
m2 3. biécoz :.~ éefi awn £35
AoBvoamm mom—-30 8:80 netwomnsfifiou
x232 Eon. win:
nowaomuafifioo bmfiam magma—m ac ham-Sn 833:3 £53.—
nmoo o 9mm 84 Ct: 3:30
98 o sauna 9°52; £5“...
8
a 35:4 8.2:”: 8335
25.:
a 2.82.. 8.35... aauoefiam
98 o 3: €3m2 3:30
38 o 8.83 2.3%: $5:
m:
a 8.35.. 2 an: 8935
:5... v
a 6.84.: 3.9.: .8525 2.3»
$2.: 852%: Emvzfi:
50h
cacao—haumm n bin—Eon. E. 3984 1.56.35 88—

 

 

baggm N @888 6.3a 033.

114

COMPANY 3

This agricultural business is an older business with roots dating back to the 1850‘s. It is
currently in its fourth generation of ownership. This is the first generation to have multiple
members of the owning family share ownership. The previous three generations of ownership
have consisted of one member ofthe family holding all or the vast majority of ownership. The
previous generation (3221 from figure 4.2.7 and 4.2.8) still maintain some ownership and is
physically present in the business. The current generation of owners are the four daughters of
3221 (3223, 3219, 3217, 3216), and two of the daughter’s husbands (3215 and 3218) also are
active in the business and share ownership.

This business has been under a great deal of stress from the 2009 economy and reduced
its employee head count from 120 to 98 in the two months data were being collected from this
business. Over the last three years this business has produced an average of$17 million in
revenue, but this number is dropping quickly due to raising cost of goods sold (specifically the
price of gas, corn, and increased government regulation).

Important to note regarding the sociograms is that communication is centered around the
family and extends in several distinct branches of communication paths (particularly apparent in
figure 4.2.8, but also apparent in the employee and ownership networks in Appendix F figures 6.5
and 6.6). This pattern is not that unusual when we consider that this business has multiple farm
buildings and employees usually stay in one location. The danger ofthis communication structure
is that information has to pass through numerous individuals before it reaches the final individual

in the path. In organization literature this could be seen as siloed organizational structure.

115

Figure 4.2.7: Company 3: Total Communication

3196

3205

13197
.3204 /3198

3213 3234 319913210
/
3206

[3209‘ /3173/3231:l3129
//43162
93220 $3174 3128 /'3130

I3227 123 .73155 H3228

3124 J3167 3216

       

 

- Employee

- Not in business as employee or owner
0 Family

[:1 Not family

 

 

 

116

Figure 4.2.8: Company 3: Family Communication

\13223
3123

 

I319 3146 \
'3156 3178
i3120 3111

- Employee

- Not in business as employee or owner
0 Family

[I Not family

 

 

 

 

117

According to table 4.2.9 below, this business is closer to the family side ofthe value
continuum (16.3), which is supported by the entire sibling cohort having ownership in the
business and communication being centered around the family (as seen in the sociograms). There
seems to be a strong agreement across the family, owners, and employees on this position in the
continuum. There also is an agreement on the level of satisfaction, with everyone scoring near
46.3, which is lower than the average level of satisfaction (p = 0.02). It is difficult to tell from this
single FOB whether this lower level of satisfaction is due to the economic difficulties and the
recent layoffs within this business, or whether there is a relationship between the family value
orientation and satisfaction.

This family is about as close as the average FOB family, but more adaptable (30.1
p<4.6). This closeness is further supported by the high density of family members in the total
communication (1.8). The Joint Count analysis ofthe family communication in this business also
tells us that the family members communicate among themselves a great deal. Taken together this
family is close, they like each other and talk to each frequently.

What is interesting about this business is that while there are higher levels of
communication within the family there is also a lot of intergroup communication, or
communication from the family to non-family members (family-nonfamily interaction 7.1). This
frequency of communication does not carry over into communication between non—family
member (non-family intercommunication = -25.8). Conceptually this would tell us that there is
not much of a boundary between the family and nonfamily for family communication. This
relates back to the family communication sociogram (figure 4.2.8) where visually the family is
centered in the middle ofthe family communication. and branches ofnonfamily members are

connected to this center group.

118

3.9% S 3333“»5 28 S can m3§§5 B: u a:

0322.. 3:83: 8 coagumsoo mozcombzma s83"...

 

 

 

 

 

 

 

390 v Q a... .32: macaw m§~§ S acetamfioe moacombzmmn as: u o
36 v Q _._ :38 93:33 :3 S nonfunfioc uuzcugema 3.2.» u a
sex: .vm
3 .830 83." a o: 33% mm
Sign 3:58.33
3% :52 age 2 859cm $.3de €22.22
2 E. usages no" a $859. 35.2% m...
85 tea- €88.52 3 been as}: 3.25
A332“; lava—-30 3:20 Borgia—5.30
it“: 13 a. ”En:
:ezauisfifienv bmfium magma—am he bicen— memfiaas 25:...“
£3 a 359.3. 2: 9.: 3:30
n3 o 32:3 5:58 been
m:
a a: 53¢ 8.2%.? 8835
S...
a A533.“ 8.55.. aceoafiam
as o 5%.: $302 5:30
as c 5%.: :92: been
m:
a. aimé can: 0%035
:5... v
a 8.32: 89.: 5:925 33>
an: 892$: 882$:
.50...
consoumaufi m mean—=80 =a 3e3< FEE—5w E3—

 

 

013 Eaesm m gage q? 23.

119

COMPANY 4

Company four is a wholesale distributor. selling a specific line of machinery for office
use to universities, hospitals and other large businesses. This business is in its first generation of
ownership. On average it employs 100 employees (3 of whom are family members). Figures
4.2.l0 and 4.2.1 1 show the position ofthe two owners. One is a family member (4182) and the
head of the owning family. His wife (41 83) and two sons (4 l 84 and 4185) also work in the
business and are in line to receive ownership ofthe business. The other owner (4143) is not
biologically related to the owner but is considered a member ofthe family by the owning family.

This business has been less affected by the economic climate of 2009 and has increased
revenue while expanding cost of goods sold over the last five years. The three year average for
revenue is $12 million.

The sociogram for total communication is rather impressive since this business operates
in three separate cities. each of which is more than 50 miles apart. Given this geographic
limitation, we would expect to see separate branches of communication for each location. similar
to the pattern in company 3. Rather we see a fairly integrated picture (figure 4.2.l0). This tells us
that even though the business is geographically separated, the communication patterns overcome
this separation. Two other points should be noted. First, in both figure 4.2.10 and figure 42.] I
(also in the employee sociograms in Appendix F figures 6.7) the two owners are not connected
directly but communicate through others. The only network in which they share a direct
connection is the ownership network (Appendix F, figure 6.8). Also the family communication
sociogram (figure 4.2.] I) shows that many individuals are connected to the family
communication network, but there are four other chains ofcommunication that are not connected
to the main family communication network. One could assume that these separated chains have a

high probability of circulating information about the owning family that is not accurate.

120

Figure 4.2.10 Company 4: Total Communication

4177 I4138

4122 \
4115

I4180

  
   

 

I4129
4125
4108 . 4 //
I4195 /I4153 ; V . 417s
l4146 L
7"»- ' 4170
94188 F \‘4136
107 -—-I4121
#4194 4 4197
[4159 179
/191 4196 IBM
4104 4189
4190
- Employee
- Not in business as employee or owner
Family
U Not family

 

 

 

121

 

Figure 4.2.11 Company 4: Family Communication

/4114

 

I4196
4131 Q4191
5 4137 173

411 ‘ 185
I \
4138 44188 7 184‘ 4164 4156

- 4183 \
04188 ‘1 149 4128
1314189 (M!) f.
‘1 4174
4113 4147 111
E4187
4179 I4162
143
‘ 4145
I4126
\u139 “94
\l4171
\4125
\14176

- Employee
- Not in business as employee or owner
0 Family
U Not family

 

 

122

Table 4.2. l 2 shows that this business is closer to the business side ofthe value continuum
(9.8). The owners and family members tend to see the business even closer to the business side in
comparison to the employees, but even the employees believe that this business is closer to the
business side ofthe continuum (10.1). The total level of satisfaction is slightly higher than
average, but the owners and family members are significantly higher than the employee‘s level of
satisfaction.

The owning family is very close (cohesion 45.0), and the level of adaptability is average
in comparison to other FOB families. This closeness is further supported by the high level of
density in the family subgroup (l.8).

The Joint Count analysis shows a similar boundary for family communication in
comparison to company 3 where the family members talk to each other (I27) and to the
employees (4.3), but employees do not communication with each other (-1 7.0). What is different
from the boundary found in company 3 is that the family communication sociogram (figure
4.2.1 1) shows a boundary between some employees and the main family communication group.
In this picture there are four chains of communication outside ofthe main centralized
communication. and this conceptually means that these employees talk to each other about the
family but do not receive communication directly from the family subgroup. The average level of
satisfaction in these out off chains is 34.4. which is significantly lower than the average
satisfaction in this business (49.2. p <0.001). In addition, the mean value orientation ofthese
chains is l2.4 which is higher than the average (9.6, p<0.00l ). This suggests that when a
boundary for family communication exists. the perceived value orientation increases. and the

level of satisfaction decreases.

123

macaw S 3:33?“ 26 2 use 33:96 3: u a:

£853. 3:83: 2 coflueQEoo uoecogzmfi 983".»

 

 

 

 

 

 

 

Sc 6 v Q 2.. :35 SSE «83.33 S aoflxeusoo moreogtwfi 28¢ u a
36 v Q a :33 “.85....3 fie S cofiuumfioo oozuognm: 28¢ u a
siefi
am: .3 422222
3 .236 Sou a o: 45.3% E
3.33" $33.32
v.2 :52 iii 2 833m $.3de 4222.22
3 2. 380.25 83v a $83.8 Had—2% m:
«we :3 _- bgesz 3 been. 5%.? £8.39
Acevosam mxméno 8:80 5533:5590
5.552 .53. mg»:
3333:3800 Ema—ah 9.9.3.5 .3 .37.qu 8:55.?— .mzfiah
3:5 c v 8.39% a; :4: .230
g .8 c v 6.92m 5:38 éefl
m:
a $53.? as so: 8935
a...
a $33.3 555.. 5:833
as o v 3.9.0 sauna 3:30
3:5 o v 5%: 2.9:: been...
see c 3&2: can: 81815
:5... v
a 9.35 3.9.: .3525 2.3»
See 882$: 532$:
£0..—
ooaaofiaumw v hang—~80 =a 9.9.34 FEE—5m 83—

 

 

23. 58:3 v >538 2.3 03¢

124

COMPANY 5

Company five is a 32 year old business dealing mostly with leasing business properties.
This business is in its second generation of ownership. The founder (individual 5135 in figure
4.2.13) has little contact with the business and his two sons (5105 and 51 12 in figures 4.2.13 and
4.2.14) hold equal shares of ownership and are primarily responsible for the day to day operation
ofthe business. This business. while affected by the 2009 economy, is growing revenue. but
expects a decrease in 2009. The average revenue in 3 years is $24.8 million.

The sociograms below show that the current owners are very central to the
communication in this business. Also a few non-family employees (5109, 51 19, 5123, 5104 and
51 15) are highly connected in the communication network. There also are many individuals not
in the business that receive communication; for example, individuals 5129, 5130, 5132, 5131,
5126, 5128 and 5127 are not family members and not employed by the business. Some are
consultants, while others are contractors that are used frequently. Finally, the family
communication sociogram (figure 4.2.14) shows a similar pattern as Company 4, where there is a

centralized hub of communication and four separated, or isolated chains of communication.

125

Figure 4.2.13 Company 5: Total Communication

 

 

5125
115131 5107
E15132 512 5128
M5127
05130 5115 5111
I5129 5134 I
4," ‘1. 5106
05135 i 31’ ‘ .41 77 5100
5135 ’7 ‘~ 5109 \
’ 5123 ' 5101
5121 5103 / 511°
5104, 5120

/
\ 5124 \'
5118 5113 \ 5116
/ \ \
5102

I5117

 

 

- Employee

- Not in business as employee or owner
0 Family

[I Not family

 

 

126

Figure 4.2.14 Company 5 Family Communication

I5121

\I5124 #435125

[15126

\l5107

[15127

\5111

 

/15132
/5115
\5131

[5122

 

- Employee

- Not in business as employee or owner
Family

[I Not family

 

 

 

127

According to table 4.2.15. this business is closer to the business side ofthe value
continuum (l 1.5, p<0.001), The owners and the employees share a similar perception of the value
orientation, while the family sees the business even closer to the business side of the continuum
(9.6). This difference is not statistically different from the average. Across the business, everyone
is very happy, with an average satisfaction of 57.5. This is significantly higher than the average
(47.7. p< 0.001).

The owning family is very close (cohesion = 43.4, p=0.02) and has an average level of
adaptability in comparison to the other FOB families, while higher than the all American families.
This closeness is further supported by a fairly high density within the family subgroup (1.9).

The family communication boundary in this business is similar to Companies 3 and 4.
with family members talking to family members (7.2), to employees (4.2). but employees do not
frequently talking to each about the owning family. Similar to company 4, the family
communication sociogram (figure 4.2.14) shows four chains of communication within employees
that are not connected to the family subgroup. The individuals in this chain do not vary from the
average for satisfaction, but they do have a much higher value orientation in comparison to the
entire group (16.0, p<0.001). This suggests that a boundary for family communication increases

the value orientation ofthe employees.

128

macaw S 3:3;an mac 2 can m3§§6 B: n a: 335:. 320.5»: 8 coauumEoe moauoSzm~u~ao~1~ue

 

 

 

 

 

 

 

Sa 6 v Q .1. «Bus 38% 985% S garages monaogcma 933 u c
36 v R e :38 anufig -u S =ontemfioo muzeomhtmmu :83 u a
sex: .em
2 .230 :30 a o: 353% mm
Artmvméu 5:33:23
3: 1.3. 3E8... Wm coho—95. $.3de A<ZOE1<Z
2 .3. 6.85.35 83.0 a @8598 man—2% m".
33 *m._ T $85.52 a; DEB..— Am. 319. comma-.00
mocvoumm Aim—-30 230.5 5353.58.50
5.552 .33. win:
get—33:88.5 man—am mzouwnsm he Ema—on 8:535 bmfiah
m:
a Gimfim 84 $12... .0550
m:
a €89 G 96:de beam
8
5 8.3.3 as _ CNS 8535m—
33. v
a 9.83% 8.25;. 8333.5
m:
a 8.992 $.33. 8£5
a: c as: 2.33: his...
m:
p 5%; _ 2.3m: .3»an
3...: v
a sire: 89.: 838.5 2.3,
$2.: 892$: 882$:
mM—Om
033555 m b.2380 =a nmEo< FEE—Sm Evan

 

 

bugm m 38:80 m—Né 035,—.

129

COMPANY 6

Company 6 is a 16 year old business in wholesale distribution. This business is growing
exponentially, and is in its first generation of ownership. The current owner (6297 in figure
4.2.17) employs his three children (two sons 6298, 6296, and one daughter 6314). There is some
talk of succession in ten years, with the daughter taking the operational leadership. She already
has begun to purchase shares of ownership from her father. It is currently unclear right now how
ownership stakes will be divided across the three children when the succession is complete.

On average, this business has 500 employees, and generates an average of $90 million in
revenue. This large size makes identifying visual patterns from the sociograms somewhat
difficult. When we look closely at the total communication sociogram (figure 4.2.16) we can see
a similar pattern to company 3, where there is a centralized hub of communication and branches
coming off this hub (it is difficult to see in figure 4.2.16 because ofthe size ofthe picture, but
more apparent in larger versions). This branching (like Company 3) follows the six different
departments within the business. which suggests the same limitation noted in Company 3 with
siloed businesses.

The family communication sociogram (figure 4.2.17) shows a centralized network of
family communication which includes non-family employees. It also shows 12 separated chains

of communication.

130

Figure 4.2.16 Company 6: Total Communication

 

 

E3 Not in business as employee or owner

0 Family

- Employee
[I Not family

 

 

 

131

Figure 4.2.17 Company 6: Family Communication

“239-46202 .6204

//la« \\'“B”'an
law
5421 6141 /.6541
'“W as 'fim 41
Gu?pun i\\\uw1 »M%-eifi% j71I6510 ‘Iaa
an

.n 7::Iew

an ‘Hfin an

  

'. ,ma ’//3$M. an

1 551 A6552

Ina ". I
an ////IM%16m2 "' l‘i‘n‘6%2 \\\Iaa H9
'5 an
097 /pnm an 6M2 ”‘elea
new Q”\\Iea fifi
use an //.mwr
an 520
as
an
[6258

an an
an
au .ulan‘/Jfifl
a
a In an
an Ina

I6328
/
6208 [36224 6543 $3qu5173
5322 6392 I6535

-__

- Employee
- Not in business as employee or owner
0 Family

1:] Not family

 

 

 

132

This business is closer to the family side of the value continuum (l5.04. p < 0.001).
There is a great deal of disagreement among the ownership, family and employees regarding this
value orientation. Employees see the business at a 15.] l. but family members see the business
closer to the business side ofthe continuum (10.7. p<0.001). and owners see the business even
closer to the business side than family members (8.4, p<0.001). On average individuals within
this business are slightly less happy than the sample population mean (46.4. p = 0.04). Employees
show an average level of satisfaction (in comparison to other family employees). Owners show a
much greater level of satisfaction (57.2). and family members show the lowest level of
satisfaction (46. l ).

The owning family is not as close as other family business families (38.0. p=0.003). and
the level of adaptability is lower than the other families (25.3. p<0.001 ). They are still more
adaptable than the average American family. The density (1.8) of family communication
somewhat contradicts the low cohesion ofthe owning family. Furthermore. the Joint Count of
the family communication is very high (17.0). Therefore the family talks frequently. but members
do not feel close to each other. The Joint Count of family communication also shows that while
family members talk to each other, they also include employees in that communication.

Similar to companies 3. 4. and 5. the employees rarely talk to each other about the
owning family. Also similar to companies 4 and 5. the sociogram for family communication
(figure 4.2. l 7) shows that while there is a central density of family communication. there is also a
boundary for some employees, creating 12 cut off chains of communication between employees.
Like company 4 the employees without access to the family communication have lower
satisfaction (44.4, p<0.001). and like companies 4 and 5 these same employees have a higher
value orientation (l6.3. p<0.001). This finding adds support to the hypothesis that employees cut
off from family communication have a perception ofthe business as closer to the family side. and

have a lower level of satisfaction.

133

' m

macaw 5 BaEhfiS 28 8 3% Nirvana» B: u a:

«ESE. 3295»: S zefiscmfioo mo§oS=MB amino

 

 

 

 

 

 

 

Sad v Q .3 :35 annm mazmaao S zofixcufioo oo§oS=M~u as: u a
36 v Q a :35 «85.43 -e S :oflgumfiou moauogtma as: n a
$65.3
$9.2 izoE<z
S .230 83. v a o: 3.3% m“.
@553 E3833
:8 25.: £8“; 2 8335 $.3de u<zofi<z
o6 1&2 9653:: mood" a $65.3. man—.25, m:
85 imam- z=§m=oz 3 £55. @533 8.8.50
moBvotax ear—-30 3:80 neuaomasfifieu
£82 .33 was.
5332......an man—ah 9.9.3.5 he mam-.0: «382.3 35.3
sac o v 9%? 8.. at: 3:30
m:
a 32:9. A3308 éafl
m:
n 82:34 Sosa: 02835
3...
a 83...: 8.55.. 5:336
nae o v :dfiw $.33. 5:30
9:5 o v $.92: 2.33: beam
59. o 2.2 2.33; 8335
8...: v .
a 9.9.9.2 an: 3 538:: 2.3,
dual» 892$: 55232
amon—
ounaoumnumw e b.2330 =a $9.3fl oufio>< So:

 

 

9.88% 9 >538 33 03¢

134

COMPANY 7

Company seven is 35 years old and primarily operates in the tourism industry.
This business is seasonally dependent and fluctuates its employee count from over 100 in the
summer to less than 20 in the winter. This business has been divided into separate wholesale
distribution, real estate, dining, car washing and park services businesses. The owners suggested
that this happened to provide a separate business for each ofthe four children ofthe current
owner. The business on average generates $2.1 million in revenue.

The current owner (individual 7107 in figure 4.2.19 and 4.2.20) is 72 years old and still
holds a 51% ownership of the business. The remaining ownership is divided between three ofthe
owner’s children (individuals 7101, 7103. 7105) and three spouses (individuals 7102. 7104.
7106). The fourth child (7103) is no longer in the business and has cut off relationship with the
family. This individual did speak with the researcher over the phone. He is not happy with the
owning family, and reported that taking part in the actual survey would bring up too many
difficult emotions for him. The owning family all described this individual as the "black sheep"
ofthe family. While this individual did not take the survey. he is noted in the sociogram because
others nominated him in their survey responses. Primarily individual 7105 (brother) keeps in
contact with him.

It also should be noted the individual 7105 is seen by everyone in the business as the
current leader ofthe business even though his father (7107) still hold the majority share.
Individual 7105 and his wife (71 1 1) are the only family members who are regularly onsite at the

main buildings ofthe business.

135

Figure 4.2.19 Company 7: Total Communication

 

 

7122
O 7102 7103
7107 7104
[7119
7101
_i 7105
um 41
7113 , 7115
/ / 7109
I7117 7111 . I 7108
, 7110 — 7112
/-I7121
7115 7120
I7124 \7123
- Employee
- Not in business as employee or owner
0 Family
I] Notfamily

 

 

 

136

Figure 4.2.20 Company 7: Family Communication

7108

7117

\7110

7113

 

[7118

7112

 

7120

7116

 

- Employee

- Not in business as employee or owner
Family

D Not family

 

 

 

137

This business is closer to the business side ofthe value continuum (see table 4.2.21
below), and the employees, family members and owners all agree that the score is an I 1.6. While
they agree about the value position, the level of satisfaction is rather low (43.3). and the owners
have a lower level of satisfaction compared to the other subgroups (39.0).

The owning family is rather distant (cohesion 32.8, p<0.001), which was verified by the
owners when the researcher returned to discuss the survey results. The current owner (7l07) and
his son (7105) (a major figure in the operation ofthe business) told the researcher that the family
members do not work well together. Also, this family is not very adaptable in comparison to
other family business families (25.2. p<0.001). But they are similar in adaptability to other
American families. This lack of closeness is further supported by the low family subgroup density
of 1.].

Generally there is little communication in this business, and ofthe communication that
exists regarding family matters, it typically stays within the family, except for what the 7105
releases. Also 7 l 05 is seen as the owner ofthe business by the employees because he and his wife
(7| 1 l) are the only family members physically onsite at the business on a daily basis. This cut
off in family communication. as well as the cut off with the 7l03 (discussed above) is
characteristic of low cohesion family systems (Olson, 2000). In this business, the family
dynamics seem to be mirrored in the business. with the same pattern of communication seen in
the family communication sociogram (figure 4.2.20) and in other communication sociograms
(Appendix F). Each one in some way shows 7l05 at the center of communication between the

employees and the rest ofthe owners (siblings).

138

macaw 5 BafiEhS 86 8 ~36 «33.63 B: u a: £553. 3:83: 8 contamsoe noncombzma value

 

 

 

 

 

 

 

Ba 6 v Q 2.. =32 macaw unmfiag 2 nonsense.» mocuogzma 224 u a
36 V A a :33 285% :3 S zoflsemsoo eunuuScma 363 u a
A58 _ .vm
. . 3m: an €22.22
5 _ 6:30 .8 o v a o: 35.2% mm
an" 3:35.52
3... :36 reg 3 8835 $.3de €22.22
35 IV. 9.80.85 :56 v a 23:.de mAmE<m mm
m2 2. bgmaoz 2 age as 38:5
Afiovouam 95-30 3:80 netwomasfifiou
Mina—z .35. mg»:
netwoiafifiou bun—ah Qantas—5 no 5655 8:535 ban—am
as o G. :93 84 at: .255
m:
a. A? came A332; Sea
a:
a about 8.2K: 88.3.".
5.: .
a 3.3%? 3.3: 5‘ cocoaunflam
m:
a 2.32: 2%va— .8550
m:
a 83.2 2.33: Qsaa
m:
a 8.33 _ 2.83; 8335
was ,
a 5%: _ - 3.9.: 8:925 2.3,
$3.: Emvzfiz 882$:
much
oouaoumaufi r manna—:0 =a 329d. 1.55.55 88—

 

 

baeesm a %53550 a? ”BS

139

COMPANY 8

Company eight is 18 years old. It provides mortuary and funeral services. Its primary
business is running five funeral homes, but other distribution and real estate businesses have been
added over time. One of the current owners (Individual 8001) reported that these off shoots
happened out of demand and not to provide separate businesses for the siblings in the current
generation. This business is in its second generation of ownership, and rapidly approaching a
third. The current generation (2"d generation) consists of two brothers (individuals 8101 and
8102) ofthe founder (individual 8124). Individual 8101 has three adult children. two of whom
are owners in the funeral business and real estate businesses (individuals 8123 and 8121). One is
an owner in the real estate business (8134). The son (8103) of 8001 is believed to be the next
owner of the funeral and distribution businesses. In total the business employs approximately 25

individuals. Over a three year average. this business generates 4.9 million in revenue.

140

Table 4.2.22 Company 8: Total Communication

'31“ 8134
”112 8122

.8113
8127 /8126

I8117

 

 

 

 

[8132
8124 8135
M131 \
. 8120
new , fl f ‘1
I 8123
8107
[8129 -
8121
8104
I8128
[8114
- Employee
- Not in business as employee or owner
0 Family
[I Not family

 

 

 

141

Table 4.2.23 Company 8: Family Communication

8128
8115

. 8121

/ 8126

I312?

8125

 

 

 

 

 

678134
122
- Employee
- Not in business as employee or owner
0 Family
[I Not family

 

 

 

142

This business is closer to the family side ofthe value continuum (l5.7). The employees
and owners agree on this position. but the family sees the business closer to the business side of
the value continuum than the employees. In general, individuals in the business are less satisfied
than other individuals within family businesses. The ownership has the lowest satisfaction, and
the family and employees agree on the average level of satisfaction (43.3). When questioned
about the low level of satisfaction in the ownership, one of the adult children in generation two
discussed the difficulty of knowing who was responsible for what tasks in each ofthe three
business entities. In general the owner believes that all the children get along well and work
together, but dividing out responsibilities explicitly between business entities has been difficult.

The owning family is fairly close in comparison to other family business families (42.3, p
= 0.06). They also share a similar level of adaptability with other family business families. The
family subsystem group density verifies this closeness with a high density of2.2.

The family communication is very dense in the family subsystem (10.4) and there is
communication between the family and nonfamily. There is little communication between

nonfamily members.

143

uaoxm S BaEéhS one 8 ~33 3535816 32 .I. a:

3559.... 3:83: 8 nonsense.» 8:335: “83".»

 

 

 

 

 

 

 

SQ 6 V E .3 :32: 35% magmas S carcass wonuoSzma ~83 u a
as v Q _._ .82: «.85an an S zofiueufioo mocuoStma :62 u 3
Eye _ .QN
3 .830 «no u a o: 353% me
5." £25.52
m: 1:2 3:5. 3 8335 $.3de 122222
S; a.» gages 861 a $83.8 35125 E
86 :3 _- €5.52 3 been nu. 8:25
Acevozam 96-30 3:80 netwoisfifioo
5.22 .53. use.
5333:8890 .555— nsenuasm no man-:5 8:52:2— mafia..—
nS c 5 :33 3:31.: 3:30
as... A2 so? 9°:de $53
as 3.3%? 8.2K: 8335
:5... v
a 85an 555.. 2233.5
m:
a €352 $802 .230
as o 33:2 2 .32: been
98 o $.32 : .33; 8335
:5... v
a 8.35.3 3.9.: 5:925 2.3,
N82» amvz<m2 89252
£0...
oongwmnwmm a AER—:50 =5 829< Biz—3m E8—

 

 

ozfl museum w sg§§o 2.3. 23.

144

COMPANY 9

Company nine is in its first generation of ownership. The current owner (9001) offers arts
education to young children. The current owner employs her adult son (9003), and there are two
other sons (9004 and 9005) not involved in the business. On average this business generatesjust
below $200,000 in revenue. The current owner is the founder ofthis business and has a great deal
of interest in theater and dance. This business generates its revenue by having children take part
in a six week program where they learn a play. The business then charges admission to the final
performance.

Individuals of interest in this business are 9001 's spouse (9002) who is an accomplished
Christian music artist and provides much ofthe music for the plays. Individual 9003 (9001‘s son)
takes care ofthe administrative duties. Finally this business has two locations which are separated
by approximately 50 miles. Individual 9006 manages the second location. All others are
volunteers (but referred to as employees in the following). as this business does not have a large

enough revenue stream to support other employees.

145

Table 4.2.25 Company 9: Total Communication

 

99004
9M6
09002
- Employee
- Not in business as employee or owner
Family
D Not family

 

 

 

146

 

Table 4.2.26 Company 9: Family Communication

09W
I9008
”310

9011

 

- Employee

- Not in business as employee or owner
0 Family

D Not family

 

 

 

147

9009

9002

This business is very close to the family side ofthe value orientation (I60), and the
employees, family members. and ownership agree on this position. Individuals in this business
are typically happier than employees of family businesses (56.9, p<0.001).

The owning family is very close (cohesion = 44.5, p<0.00l ), and has an average level of
adaptability in comparison to other family business families. The family communication is dense
in the family subsystem (D = 2.2). The Joint Count analysis does show a boundary between the
family subsystem and the employees (see figure 4.2.26). There is one employee attached to the
family communication (9009 in figure 4.2.26), and this employee has a higher satisfaction than
the two employees cut off from the family communication (employee 9006 and 90] I have a
mean satisfaction of 52 which is a significant difference of p 1 0.09).

Interesting to the value orientation and the family boundary is that the disconnected
employees have a mean value orientation of 8.0 (compared to the company mean of 16.0). This is
a reverse situation from businesses l-7 were being distances from family communication
increased the value orientation. It would seem that the relationship between value orientation and
access to family communication is more complex. In this case the cut off increased the difference
between the ownership value orientation and the employee value orientation. It would appear
from this business that a disconnect from family communication will decrease satisfaction
(similar to companies 4. 5. and 6) and increase the difference in value orientation from the
employees to the ownership (the direction most likely depends on the value orientation ofthe
ownership, i.e. high value orientation ofthe ownership will result in a lower value orientation in

the employees and vise a versa).

148

Quogm S 333.565 one 8 our o3§§5 3: u a: 3&5»... 3:83: 9 copious?» outuognma >830.»

 

 

 

 

 

 

 

3a 6 v Q 2. .82: Raoum 985.43 S towtemsoo 8:332»; 3.8.» u o
36 v Q .. :32 .983an -u S coutomsoo monooSzmz as: u a
pic _ .VN
Amdvfiwm 4<ZOE.<Z
m3 3:30 93 u a o: 3.3% mm
ms" 5:33.233
m2 3 3:5“. 5 8338m— Avémdm ._<zo_e<z
$5 on. _- 9.20.35 83v a $93.2. 35.4% E
3.6 w _ . T bEamcoZ Nd bison mév :38an
93333“ mxméfio mmzohO 5:53—55:50
5.552 Each “Ev...
533358—50 man—ah 98595 .3 mam—:5 8:53.»: wan—ah
m:
a $.8qu an 3...? 3:30
nmm o 8.83% 5:38 3:5
8
a 8.3.3 Scams. 88.3w
:5... v
a @653 5:35.. 5:85am
pg c 25808 Av.mvm.m_ 3:30
m:
a 8%.: 2.32: .25...
a: c 3%.: 2.33; 8.835
25.: v
a 6.3.3 $9.: 5:925 33>
$2.: amvzfi: 832$:
amon—
oouuoumawmm a manna—o0 E. 39.3. 9.5235 E3—

 

 

Efigm a 3.3800 add 2an

149

COMPANY 10

Company ten is 18 years old, in its first generation of ownership. The founder (individual
10006) employs his son (10007) who is in the process of buying out the ownership from 1007.
Their primary industry is finance. This business employs five other (nonfamily) individuals
(10001, 10002, 10003, 10004 and 10005). Individuals 10008 and 10009 are the spouses ofthe
current owners.

The family sociogram in figure 4.2.29 is particularly revealing of the level of conflict
between the family and non-family employees in this business. In this figure it is very clear that
the family communication has a strong rigid boundary between the employees and family
members. To some degree the same pattern is apparent in the total communication sociogram
(figure 4.2.28). 111 this picture there is a visual symmetry, with the owners on top and the
employees on the bottom. What is also shown in this picture is that the employees talk to the
father (10006), but they are not talking to the son (individual 1007). This hints at problems with
the succession process. Conceptually this tells us that the employees still see the father as the
primary leader, and it even shows that the employees are not willing or possible able to develop

lines of communication with the future owner, the son (10007).

150

Table 4.2.28: Company 10: Total Communication

5 11!!
I m m

010009

 

i 10005

111101

I: mom / ‘V

10002
1W3

 

- Employee

- Not in business as employee or owner
0 Family

D Notfamily

 

 

 

151

Table 4.2.29 Company 10: Family Communication

 

1310010 10001
10003
.1 ..
10005
10004 11007
10006
10009

- Employee

- Not in business as employee or owner

0 Family

U Not family

 

 

 

152

The business is closer to the family side ofthe continuum (15.7, p<0.001). There is a lot
of disagreement on this perception, with the ownership seeing the business closer to the business
side ofthe continuum (l 1.5, p , 0.001 ). The level of satisfaction is very low for this business. and
the lowest of all 1 1 businesses sampled. The ownership and family is very satisfied (54.0), but
the nonfamily employees are very dissatisfied (37.0).

The owning family is very close (cohesion 45.0) and has a similar level of adaptation to
other family business families (31.0, p<0.36). The density of family communication is about
average (D=1.5), while the ownership communication is very dense (D: 2.1). The Joint Count
analysis tells us that regarding family communication, the family subsystem does communicate
(2.13) although there is not much communication. The Joint Count also shows a very significant
boundary for family communication. The nonfamily group has a greater density than the family
group for family communication (5.33 versus 2.13). Also there is no communication between the
two groups (lnterGroup ratio . 0). This distinct boundary verifies the visual boundary seen in
figure 4.2.29. This situation adds support to the finding that employees who are cut off from
family communication have less satisfaction (37.7 compared to 54.0) and a disagreement between

the value orientation (15.7 versus 1 1.5).

153

$8.& 5 3333.55 28 8 use m3§~§u 3: u a:

335:. 320.5»: 8 toutumfiou muzaufiammu r8399

 

 

 

 

 

 

 

396 V R 1. ~32: macaw unufig S contamEou moxuogzmma 963 n a
.36 v a 5 =35 anufimae :u S tofigumsg mozuoSsma :83 u a
5.3::
3.8—.2 A<ZO~H<Z
Z 3:30 one u a o: 3.5% m...
in 3.33%:
Ea m2 325 2 00835 $3.3. 5.2932
3. LE- gauges 8qu a $33.9. miszm m:
3 .N than $5932 3 éefi 3... asafio
3823M quéno 8:80 5533:8500
£32 .83. 9:»:
gouaomaafifiou baa—am 9.9.3.5 no .035: woman—ia— 22.8,.—
p8 c 6.33% A3 3.: 3:30
as o €393 A332; 335
m:
n. And—v5.2“ S.o$m.$ coho—mam
:56 v
a 5.3.? 8.35.. geoeflam
as o 2.8%: $95 5:30
as o :dn: 2.32: baa“.
m:
a 9.95.: 2.33: 833m
8:... v
a 83.2 3.9.: 5:925 2.3»
32¢ 662$: 832$:
mm—Om
3:35..»5 3 hug—:80 =u 8964 Eacafim EB—

 

 

ozfl Eagm S $.38 ”on? 2%...

154

COMPANY l I

Company eleven is in its first generation of ownership and is 23 years old. The current
owner (l 1001 ) employees his two sons (individuals I 1002 and l l006) and two other non-family
employees (I 1004 and l 1005). The founder also has a daughter (I 1009) who is not employed by
the business. The primary industry is finance.

From the sociograms we can see that the founder and his two sons are in the middle of
the total communication and family communication networks. There also is a good balance of
communication going to the founder and his two sons, which is good to see since the founder is
considering a succession within the next five years. Also interesting to note from the family
communication sociogram (figure 4.2.32) is that the founder’s wife and mother ofthe two
brothers (individuall l008). and individual 1 l006's fiance' (individual l 10] l) are well connected
into the family communication network , but not well connected in the employee and ownership
networks (see Appendix F. figures 6.21 and 6.22). Conversely, the founder’s daughter
(individual 1 1009) is well connected in each ofthe three communication networks. even though
she is not employed and does not hold any ownership within the business. This creates a nested
family group within the family system. It would seem that the founder and his three children

have a special bond that other family members (spouses) do not share.

155

Table 4.2.31 Company 11: Total Communication

0111010 11°05

11009

11002

 

 

 

[11004

 

- Employee

- Not in business as employee or owner
0 Family

D Notfamily

 

 

 

156

Table 4.2.32 Company 11: Family Communication

 

”1°05 11007
11002
11008
/
it“
11011
/ \ A
11011 V"
11006
11009
(711010
- Employee
- Not in business as employee or owner
0 Family
D Notfamily

 

 

 

157

111114

This business is very close to the family side ofthe value continuum (18.5), and there is a
great deal of agreement between the owners. family members. and employees. The lcvel of
satisfaction is somewhat low (41.33 p<0.001).

The owning family is not very close (cohesion 37.8, p<0.001), but they are more
adaptable than most family business families (30.3, p<0.003). There does not seem to be m uch
family communication from the family to employees or within the employee group (Joint
Comparison close to 0 for both). Although the Joint Count tells us there is little betu een group
communication, the sociogram (figure 4.2.32) shows that both nonfamily employees are
connected. Therefore we would conclude that the family communication is low, but there does

not seem to be a rigid boundary between the family and employees.

158

35% S BSeEVS 23 8 2% 33336 B: u a:

3&5»... 3:83: 9 zoflgumfioo outcombtmmu unmiuo

 

 

 

 

 

 

 

~86 v Q 1. =33 Anew wastage S routomfioo unauognmmu $83 0 a
36 V & _.. :35 .035an an S contoufioo 3§o$§n 283 u a
ASS—.3
Ada—am A<ZOF<Z
E 5:30 83 u a o: 552% me
man banana—3..
a: a? £5". 3 3335”,. $.33 q<zo:.<z
mo ”3.. 380.35 Scan a $83.8 55% mm
3 8o- bgmeoz 3 baa; ”.5 832.5
AoBVozam guano 8:80 :ouaoiaafiou
E32 .531 we»...
notgfiafifiou bun—am magma-i .3 >325: mom—52in— mafia."—
nfi c Aoocmmdm 8.51:1 .055
as o a; :33 9238 Sea
a? o 8.23.3 8.2K: 853m
3:... v
a 6338.: 53:5. actuaflam
as o 3%.: 9531 .830
m:
n 8.32: 2 Si: 551
m:
a 8.03: 2 do: 8335
:56 v .
a 963: a 9.: 5:925 2.3»
See 592$: 892$:
mach
3:35.35 3 manna—o0 =a 323‘ Ewen-3m ES—

 

 

Efigm n: 3.8800 ”mmdé 03.31

159

Discussion o/‘Research Questions

2.] Satisfaction increases as the value orientation of the FOB decreases

There are obviously many factors involved in the level of satisfaction within and
across businesses. For example. businesses that are doing well financially probably have
higher levels of satisfaction than businesses that are struggling. But even given these
other outside factors, there seems to be a strong connection between the overall value
orientation of a family business and the overall level of satisfaction. Figure 4.2.34 shows
this strong negative relationship using the company level summaries. When we regress
the business value orientation on satisfaction we see a very strong adjusted r-square of
0.3 88. which means that nearly 40% of the variance in satisfaction across businesses is
due to the value orientation ofthe business (i.e. with businesses that are closer to the
family side of the value continuum having lower levels of satisfaction than businesses

closer to the business side ofthe continuum).

160

Figure 4.2.34 Satisfaction and Value Orientation

 

60 - Company

3 1

' 2

§ 3

N 4

‘ 5

' 6

< 7

I: ' 8
.3 ‘ 9
9 O 10
g X 11
‘6 Fit
to \ Line

 

O
35 I I I I I I I

8.0 10.0 12.0 14.0 16.0 18.0 20.0
Value

 

 

 

2.2 Satisfaction varies by subgroup

There does seem to be strong evidence that satisfaction varies by subgroups.
Companies 1, 2, 4, 5, 6, 9, and 10 all have ownership and family levels of satisfaction
higher than their employees. (Companies 3 and 11 have equal satisfaction across the
groups). But in some cases (companies 7, 8) the ownership will have a lower level of
satisfaction. Interesting to this deviation is that company 7 is decreasing in revenue, and

company 8 is increasing in revenue, therefore it doesn’t seem to be the decrease in

161

revenue alone that creates the decrease in satisfaction. In general it seems that employees
have lower levels of satisfaction than the ownership and family, but there is another
factor that can change this relationship. This additional factor will be explored in the

following questions.

2.3 Employee groups with higher value orientations than the owning family will have

lower satisfaction

There is strong support for this hypothesis. This relationship was shown in
companies 1, 2, 4. 6, 8 and 10. Companies 3, 5, 7. 9 and l 1 had similar (statistically
similar) value orientations and levels of satisfaction between the family and employee
subgroups. Of the six businesses where this relationship occurred (employees with

higher value orientations than the owning family). the hypothesis was supported.

2.4 Cohesion positively related to satisfaction

While there is a positive relationship between cohesion and satisfaction, this
relationship is affected by the overall Value Orientation of the FOB. Figure 4.2.35
below depicts this relationship. In general. there is a positive relationship between the
closeness ofthe owning family and the level of satisfaction across the business. but close
families that also have a strong leaning towards the family side of the value continuum
produce low levels of satisfaction across the business. Figure 4.2.35 shows that

businesses in the lower right hand quadrant all have a high level ofcohesion and a low

162

level of value orientation. All ofthe companies in this quadrant have a level of
satisfaction that is above average. Conversely, company 10 (Upper Right Quadrant) has
a high cohesion but also a high value orientation. This business has the lowest level of
satisfaction in comparison to all of the sampled businesses. Figure 4.2.35 also suggests
that the while cohesion plays a role in satisfaction, value orientation has more influence.
For example, company 2 (Lower Left Quadrant) has a low cohesion and low value
orientation, but it benefits from a high satisfaction due to the low value orientation.

The least appealing relationship seems to be an FOB with a very close family and
a Value Orientation that favors the family side of the value continuum (upper right
quadrant), followed by distant families that favor the family side of the value continuum
(upper left quadrant), when compared to distant families that favor the business side of
the value continuum (lower left). The best option is the lower right quadrant in which

the owning family is close, but the FOB favors the business side of the value continuum.

163

Figure 4.2.35 Cohesion and Satisfaction

 

 

 

 

 

 

20.0 "
0 Company
17.5 " Q 1
x . 2
o O 3
v ' I 4
15.0 1 ‘ 5
f.) ' 6
K " I 4 7
’ 8
O
a .. n 9
1;: 12.5 x 10
4 <9 G 0 11
\ Fit
Line
10.0 ‘ 0
o Satisfaction
above
average
I i F I I 1 I
32.0 34.0 36.0 38.0 40.0 42.0 44.0
Cohesion

2.3 A rigid boundary for family communication will reduce satisfaction

There is evidence that a rigid boundary between family and nonfamily members
reduces satisfaction. Companies 2, 7 and 10 showed a fairly rigid boundary between all
family members and employees and demonstrated a significantly lower level of
satisfaction for employees (in comparison to the owners). Additionally, companies 7 and
10 had employees that scored significantly lower on satisfaction than the average
employee of a FOB. Furthermore, companies 4, 6 and 9 had employees that were visually
cut off from the family communication (using the sociograms). These employees showed

a significantly lower level of satisfaction than their connected

164

counterparts. Finally, employees where communication flows through a permeable
boundary from the family system to the nonfamily subgroup produced satisfaction scores
that are similar to the owner‘s (e.g. companies 3 and 1 1).

It seems that while the value orientation and the level of cohesion of the owning
family do have a significant effect on the value orientation of the business. a connection
to the family communication also can increase levels of satisfaction in nonfamily

employees.

2.6 A rigid boundary for family communication will increase the distance between

employee and family value orientation perceptions.

There is evidence that a connection to family communication (specifically having
access to family members and an ability to receive communication from that group about
the owning family) has an effect on the perceived differences in opinions in the
business‘ value orientation. Hypothesis 2.3 showed that in general employees have a
higher score for value orientation in comparison to the family members, but having
access to family communication from the owning family seems to reduce this difference.
The rigid boundary for all employees in companies 2 and 10 accounts for the much
higher value orientation of the employees in this business. and the employees in
companies 4, 5 and 6 that were visually identified as cut off (from the sociograms) had a
value orientation that was significantly different from the connected employees.
Furthermore. companies 3 and 1 l have permeable boundaries. and there is agreement

between the employees and family members for the value orientation of the business.

165

2.7 Adaptability is positively related to satisfaction

There does not seem to be a significant relationship between adaptability and
satisfaction with this sample population. This may be due to most of the businesses in
this study being significantly higher in adaptability than the national average (two
businesses‘ scores were higher than the mean but not statistically different from the
mean). The fact that all of the business scores were at or above the mean, and produced a
sample mean 4.04 points higher than the national mean (national mean = 24.10, sample
mean 28.14, t=5.89, cal/‘56. p <0.001) raises the question of whether there is a threshold of
adaptability for FOBs, especially those that are successful. In other words. does a family

have to be at or even above the mean adaptability to survive as a FOB‘?

Summary of Phase 1: Step .2

First. FOBs vary in regard to their overall value orientation. This variability has
an effect on the overall satisfaction ofthe individuals within the business. with businesses
closer to the family side of the value continuum having on average lower levels of
satisfaction. Also Hypothesis 2.4 shows that there is a relationship between the level of
closeness in the owning family and the average level of satisfaction. C loseness is
somewhat related to satisfaction where closer families have higher levels of satisfaction.
However close families have to be careful of forming a FOB with a value orientation that
is too far to the family side ofthe value continuum because FOBs with close owning

families and a value orientation closer to the family side have lower satisfaction levels.

166

The best option is to have a close owning family that has a value orientation that is lower
(closer to the business side).

While Cohesion and Value Orientation seem to have some relevance in
explaining across business variations in satisfaction, access to family communication and
subgroup membership tend to explain the within business variability for satisfaction and
value orientation. Hypothesis 2.2 showed that on average individuals within the
businesses vary in their scores for satisfaction by their subgroup membership. More
specifically, family members tend to have the highest level of satisfaction, followed by
the owners, and then the employees have the lowest levels of satisfaction. The level of
satisfaction is higher for employees who have access to family communication
(Hypothesis 2.5). Hypothesis 2.3 showed a similar trend for the value orientation for
individuals. In this case owners have the lowest score for value orientation (tend to see
their business as closer to the business side of the value continuum). followed closely by
the family members. Employees seemed to see the business as closer to the family side.
This difference of opinion becomes exaggerated when there is a rigid boundary for
family communication (when employees are cut off from family communication)

(Hypothesis 2.6).

Phase 2: Step 1
This final phase of this study will model the relationships found in Phase 1: Step
2. Since the relationships found in Step 2 were qualitative in nature, it makes sense to test
these findings using a quantitative methodology and in this way add support to the
findings in Phase 1: Step 2. Phase 2: Step 1 addresses Specific Aim 3 and fits a model for

the variations in satisfaction within and across the sampled businesses.

167

 

Specific Aim 3 .' Test the new expanded Three Circle Model for its ability to explain the

relationship between owning family dynamics and satisfaction

H 3.1 The distance between an individual's perception of his/her FOB‘s value
orientation and the actual value of the FOB is negatively related to an individual‘s

level of satisfaction with his/her FOB.

To test this hypothesis a baseline or unconditional model was created to compare

H 3.1 for its ability to explain variations in satisfaction.

Model I — Unconditional Model
Level]

Satisfaction” = 30,- + 7‘11

1801' = You + #0)

This model and all other models presented in this study will use Restricted
Maximum Likelihood (MLR). Raudenbush and Bryk (2002) as well as Kreft and de
Leeuw (1998) suggest that when models have larger populations on level-2 (J) the
difference between Full Maximum Likelihood (MLF) and MLR is negligible. but for
models with smaller J, MLF estimation will produce artificially low variance
components, as they are reduced by (J —F)/.'/ factor, where F is the total number of
elements in the fixed effects vector, y. This makes MLR a more appealing option for this

sample population of 492 individuals and l 1 businesses.

168

Using MLR estimation, this model converged in 6 iterations, indicating a
relatively good fit for this model. The y00(or intercept) was estimated at 48.39. The
estimated between business variance (or TDD) was 27.15. The estimated within business
variance was 02 = 1 10.74. The 95% confidence interval for the variance between
business intercepts of satisfaction is 48.39 i 1.96(27.15)1/2 = 58.59, 38.19. Based on
this covariance. the intra-class correlation is: ICC : 27.15/(27.15 1- 110.74) : 0.1968.
Therefore, approximately 20% of the variance in satisfaction is between businesses.
while approximately 80% is within businesses. In other words, while satisfaction docs
vary from business to business. satisfaction varies even more from one individual in
business j to another individual in businessj. This magnitude of variance between
businesses can be formally tested (H0 : r00 = 0), and is distributed using a x2 with 1-1
degrees of freedom under the null hypothesis. The present unconditional model takes the
values of x2 = 55.31 with Cir: 10 (. = l 1). This is highly significant (p < 0.001).

In summary, this model shows that more variance lies within businesses (80%)
than across businesses (20%). This means that there is variance to be explained in level
one and level two which allows us to use a multilevel model to test the findings from

Phase 1: Step 2.

Model 2 — H 3.1 The distance between an individual's perception of his/her F OB‘s value
orientation and the actual value ofthe FOB is negatively related to an individual‘s
level of satisfaction with his/her FOB.

Level 1
Satisfaction”- : 60,- + Eli-(Value difference”) + ru-

169

Leve12

180/ = 1’00 + #01

Bu 2 1’10
In this model Value difference is a group mean centered variable or Value”- — WI}.
Therefore this is modeling the distance an individual is from his/her business's mean
value (or the true value). 501 is the average level of satisfaction for businessj when we
control for the distance for person i's perception of value from the mean of their business
j value. Since the independent variables have been standardized by mean centering
(A7 = 0 )and setting the SD to 1. the intercept becomes the average satisfaction for
individuals in businessj for those individuals who have a score ofX. and one SD
increase in Value Difference produces a corresponding change in Satisfaction.

Model 2 converged in 6 iterations allowing the deviance ofthis model to be
compared with the deviance on the unconditional model. Adding Value difference to the
model created a better fitting model as can be seen by the change in deviance from the
unconditional model to Model 2. With adding one extra parameter. the deviance was
reduced by )8“ 32.2. (cl/'1. p < 0.001). It is also possible to determine the model‘s ability
to explain variance (or proportion of variance explained). This is accomplished by taking
the difference in variance from the unconditional model and the nested model (model 2).

100(m0del 1)— r00(mode( 2)

 

The equation for this is [5 =
TOUUnOdel 2)

The ,6 = 0.016 or 2% more variance is explained by this model. The estimated coefficient
for the Value Difference is -1.69 (cl/"479, p = 0.002). This finding supports the hypothesis

that differences in individual perceptions (in comparison to the businessj mean value)

170

will decrease Satisfaction. In other words. one standard deviation increase in Value

difference will decrease satisfaction by 1.69.

H 3.2 Subgroup members vary in their level of satisfaction.

Model 3
Levell
Satisfaction”-
: [30}- + Bil-(Value difference”) + BZJ-(Familyij)
+ B3j(Employee,-j) + B4j(0wner,-j) + ru-
Leve12

301‘ = 1’00 + #01'

511' = 1’10
321' = 1’20
531' = 1’30
54} = 1’40

To test this hypothesis three parameters were added to level one. Each parameter is a
binary value (1 for a member of the group, 0 for not a member) and therefore not
standardized. The model converged in 6 iterations. allowing for comparison with Model
2. This model did reduce the deviance from the previous model by 19.1 (16': 19.1. df'3, p
< 0.001), but the level 1 variance component increased by 1.1. while the level 2 variance
component decreased by only 0.01. Furthermore the t ratios for each parameter were not
significant (Owner I: 0.424. df.‘ 476. p = 0.671) (Family I = 0.11. df.‘ 476. p = 0.916)

171

(Employee t : -1.2. df.’ 476. p = 0.23). Therefore. the addition of subgroup members does
not make a significant contribution to explaining variance in Satisfaction after controlling
for individual differences in Value Difference.

A possible explanation for this difference is that once one controls for the
perception difference. it is redundant to explain differences for subgroups because
subgroups may vary consistently. For example, it may be that family members always
have a lower value orientation than employees as discussed in Hypothesis 2.3 above. This
will be explored further in Step 2 of this phase. For this step it makes sense to remove
the subgroup member variables and continue building a model using Model —- 2 as the

baseline model.

H 3.3 Different family system types produce varying levels of satisfaction within the

business.

To test this hypothesis the level 2 variance was tested. According to Model 1. we
know that 20% of the variance in Satisfaction is accounted for by between business
characteristics. To explain this variance and test Hypothesis 3.3, a fourth model was fitted
that included the family system parameters Cohesion and Adaptability. It made
substantive sense that each variable should be fitted for variance on the intercept as well
as variance on the slope of the Value difference variable. By fitting the slope we are
testing the hypothesis that levels of adaptability and cohesion within the owning family
affect the intensity of the slopes for each Businessj on each of the regressions of Value

difference on Satisfaction. For example if cohesion is found to have a significant negative

172

effect on the slope of Value difference. then we would say that the closer a family is. the
more intense the negative relationship for Value Difference and Satisfaction. In other
words, while having a different opinion from the mean value orientation reduces
satisfaction. it is reduced even more significantly in a FOB with a very close owning

family.

Model 4
Levell

Satisfaction”- : [301- + [ill-(Value difference”) + r”-
Leve12

301- = yoo + y01(Cohesion_j-) + y02(AdatabilityJ-)+ “01‘

BI]- = ym + y10(C0hesi0n_I-) + y20(AdatabilityJ-)

This model converged in six iterations. The change in deviance from Model 2 to
this nested model was x2 20.13. (if; 2. p <0.001. But the level 2 variance explained did not
change. Furthermore. neither variable produced a coefficient for the intercept that was
significant (Cohesion t = 1.27, of, 8. p = 0.24) (Adaptability t = -0.36 (if? 8. p = 0.73).
This suggests that neither Cohesion or Adaptability affect the intercept (or mean
satisfaction) for businessj. While these family dynamic variables did not explain mean
Satisfaction, Cohesion was found to have a significant relationship with the Value
difference slope of a -2.54; this relationship is further supported with a t-ratio of -2.04
(df. 475, p = 0.04). Therefore, the model was fit again using just Cohesion for the slope
of Value differenceSatisfaction. This cleaned version of Model 4 was a well fitting

model with a change in deviance of 10.84 (elf; l p = 0.001). This model explains

173

approximately 1% more of the level 1 variance than Model 2 (using the proportion of
variance explained).

This model partial supports Hypothesis 3.3, in that families that are closer (higher
on cohesion) increase the magnitude of the Value Difference to Satisfaction relationship.
The yl 1coeff1cient was -2.67 (t = -2.24, df478, p = 0.025). Therefore while having a
different perception of the FOB value orientation (in comparison to the group mean) will
reduce satisfaction, the magnitude of that reduction is increased as the level of cohesion

in the owning family increases.

H 3.4 Businesses closer to the family side ofthe value continuum have lower levels of

satisfaction.

Model 5
Levell

Satisfaction” = [30}- + fill-(Value difference”) + TU-
Leve12

1801' = 1’00 + y01(Value_j) + #0}

31; = 1’10 + rio(C0h88i0n.,-)

This model converged in 6 iterations allowing a deviance comparison with the
previous model. The change in deviance was 7.04. (x2 1.56. (if; 1 p = 0.008). Therefore.
after controlling for individual value perception differences and the effect of the owning
family system. 40.8% (using the proportion of variance explained of level-2 variance

174

components) of the between business variation in satisfaction can be explained by the
mean value orientation of the business. More specifically, the estimated coefficient is a -
1.19 (t = -2.34. df.‘ 9. p = 0.045). meaning that for every standard deviation increase in
businessfs value orientation. there is a decrease of— l .19 for satisfaction (mean
satisfaction for businessj).

The author stopped fitting this model at this point due to the reliability of the
model dropping below 0.70 (reliability of model = 0.69) also the deviance changes are
now relatively small. Taken together with issues associated with MLR and a small
sample size. further fitting of this model may produce biased variance estimates and
shortened confidence interval which would lead to type 1 errors (Raudenbush & Bryk.

2002).

175

Table 4.3.1: First Model Summary

 

Fixed Effects

 

Intercept
Value Difference

(Bl)

Null
Model; s.e.)

48.40(1.69)**

48.37(1.76)**
-1.69(0.53)**

Pinoacf4f‘

 

 

48.37(l.7)**
-2.08(0.56)**

f

hinders—

64.42(7.03)**
~2.08(0.56)**

 

 

Cohesion (y10) -2.67( 1.19)* -2.67(1.l9)*
Value (y01) -1.19(0.51 )*
Variance ‘ — — fl 7— _
Component
Intercept (#0) 27.15 26.58 26.69 15.80
Level 1 (R) 110.52 108.93 107.99 108.03
Model Fit
Reliability (BO) 0.781 0.779 0.781 0.692
Deviance 3666.227 3634.03 3628.67 3621.63
Deviance Change 32.20 5.36 7.04
(if

2 3 4 5

 

 

 

 

 

**p<00m
*p<005

 

 

Reliability for l’hase 2.“ Step 1 ."l’lodel

A box plot of the within business residuals can be used to determine ifthe

residuals are centered at 0. and that the variances are consistent across groups. Figure

4.3.2 shows that the residuals seem to be centered at 0. Also, a scatter plot of the

residuals against the fitted values is used to test whether there are problems with

heteroscedasticity. Figure 4.3.3 shows that there are no recognizable patterns. which

indicates that the assumption for heteroscedasticity are reasonably met. Finally a P-P plot

(Figure 4.3.4) ofthe level 1 residuals show that the data seem to normally distributed.

176

 

Figure 4.3.2: Box Plot of Residuals by 11 FOBs

 

20.1300"

 

11
l l

 

 

 

 

 

 

 

 

E
‘0 d
0
I-
E
010 432 T T
57
520.000" 01 0
g: 44', 3-4
26
128 .3331 27:
1')
306
40.000“ 0
T I l T l I T I 1 I I
”l 2 3 4 '3 El 7 8 El 10 1 1
121d

177

 

Figure 4.3.3: Scatterplot of level 1 residuals against fitted values

 

 

 

 

 

20.000—
00
(D 8 8;)
O O 0000 f)
. - 0%0 8 o O ‘ 8700
. r3 ii
" O - t
0.0UJ 00 ,3 C- ‘ 0
13 00 C) (.1 . O O (J
3 .c B C) C
3 00 0')
E 00
C
-20000— 0 C
O
Q ‘
O C ,
O
0
“40.000—1 0
1 1 I l I
40 000 45.000 50.000 35.000 60,000
fitval

178

Figure 4.3.4: P-P plot ofthe level 1 residuals

Normal P-P Plot of l1resid

 

08"

Expected Cum Prob

 

 

 

0.0-r

 

I 1 I I
0.0 0.2 0.4 0.13 0.8 1.0

Observed Cum Prob

Phase 2: Step 2
The second step in phase 2 seeks to address Specific Aim 4:
Specific Aim 4: Test the new expanded model for its ability to explain the relationship
between ou'ning‘faniily cit-I'namics and value orientations.
Phase 1: Step 2 produced a number of hypotheses that suggested that value
orientation is not only fluid, but is influenced by business level and individual level

factors. Additionally the first step in this phase as well as Phase 1: Step 2 showed that

179

there is a significant relationship between satisfaction and value orientation. Taken
together. if satisfaction is low in a business. the new expanded Three Circle Model would
suggest that this is due to value orientation being high. along with the interactions with
cohesion and adaptability. Therefore we can affect (or increase) satisfaction by
decreasing the value orientation. This Step explores the most efficient ways of changing a

value orientation.

H 4.1 Subgroup membership will affect the value perception of individuals within

FOBs

Model 1: Unconditional
Level 1

Value Orientation”- : 80,- + ru-
Level 2

301’ = 1’00 + 1’01 '1‘ #0)

Using MLR estimation. the model converged in 5 iterations. indicating a
relatively good fit for this model. The y00(or intercept) was estimated at 13.59 (t ; 14.95.
(ll: 10. p < 0.001). The estimated between business variance (or T00) was 7.47. The
estimated within business variance was 02 = 2292. Based on these covariances the intra-
class correlation (ICC) = 7.47./(747 1- 22.92) 0.246. Therefore 24.6% ofthe variance in
Value Orientation is between businesses while approximately 75.4% is within businesses.
The 95% confidence interval for the variance between business intercepts of satisfaction

is 13.59 i 1.96(7.47)1/2 = 8.23, 18.95. This magnitude of variance between businesses

180

can be formally tested (H0 : T00 = 0), and is distributed using a x2 with 1-1 degrees of
freedom under the null hypothesis. The present unconditional model takes the values of
111.67 with df= 10 (J = 11). This is highly significant p < 0.001. In summary. this model
shows that more variance lies within businesses (75.4%) than across businesses (24.6%).
This in itself is an interesting finding as one might assume that everyone in a FOB would
have a similar impression of the family versus business value, but this unconditional
model suggests that there is greater variation in value perception within a FOB than
across F 085. Since significant variance is within and across FOBs we can test
characteristics within and across businesses to explain this variance.

Using this unconditional model as a baseline. we can test Hypothesis 4.1: Does

value orientation vary by subgroup membership?

Model 2 — Subgroup membership will affect the value perception of individuals within
F OBs.
Level 1
Value Orientation”
= 50,- + [BU-(Family Member”) + Sal-(Owner Member”)
+ [Rm-(Employee Member”) + Ti}
Level 2

.80} = 1’00 + #01'

Blj = 1’10
321' = 1’20
33)" = 1’30

181

This model converged in 6 iteration, and the deviance change was 18.71 (de 3. p <
0.001). While this was a relatively better fitting model than the unconditional model. the
parameter estimates for Owner Member and Employee 1 ember were small and not
significant (Owner 1 ember 0.22. p = 0.862; Employee Member 0.18. p I 0.883). while
the estimate for Family Member was larger and significant (~2.88, p i 0.021). One
explanation is that all three variables are binary coded, and the Three Circle Model
suggests that there is overlap between the three groups. Therefore. none of the three
binary coded variables are a true dichotomy. In this sample the Family Member group is
the closest to a dichotomy. and arguably the most exclusive variable in comparison to
owners and employees. For example, there were 38 owners in the ownership group. and
35 of those were family members. making these two variables somewhat redundant. This
leaves the family and employee groups. but a similar problem exists with the employee
variable. the majority of family members in this study were also employees. Due to these
factors it made sense to isolate the family variable. Furthermore. from Phase 1: Step 2
access to family communication seems to be important to individual value orientation.
Therefore. controlling for family group membership while exploring family access will
allow us to test the hypothesis that access to family communication affects value
orientation even after we have controlled for the effect of being a member of the family.
A new model was run that contained only the family member variable. This new
model converged in 6 iterations allowing the deviance of this model to be compared with
the deviance on the unconditional model. Adding Family Member to the model created a
better fitting model as can be seen by change in deviance from the unconditional model

to Model 2 (x2: 14.10. df‘l, p < 0.001). In comparing the proportion of variance

182

100(m0del 1)— (00(‘m0d612)

 

explained from Model 2 to the unconditional model, using )6 = ,
t00(m0del Z)

, f5 = 0.032. In other words. 3.2% more of the level 1 variance is explained by model 2 in
comparison to the unconditional model. The estimated coefficient for the Family
.ll/‘lembers is —2.83 (deSJ. p < 0.001). Taken together when we control for family
members value orientation the average intercept is 14.48 (from cleaned model 2). Family
members tend to have a lower value orientation than other subgroups by a 2.83 point
decrease in intercept. Therefore. generally family members tend to see the business closer

to the business side of the value continuum in comparison to other subgroups.

H 4.2 Access to family communication will decrease the value orientation of an

individual.

To test this hypothesis a model was fit using the Family Access variable. There
are two problems with this variable. First (as was noted in chapter 3) the distribution of
the family access variable is not normal. it is positively skewed. This is because there are
often individuals within the business that do not have access to family communication.
Secondly the information gained in Phase 1: Step 2 revealed that individuals who
communicate about the family. but are not connected to the central communication (or
the owning family) have a higher value orientation. The limitation with the Family
Access measure is that it measures one’s “connectedness" but not what group one is
connected to. Therefore an individual could be highly connected to a group that is broken
off from the family group. For example in company 10 the employees are highly
connected with each other but not connected to the family. These individuals would

183

receive a high score for Access because they are connected to each other. but
conceptually they are not connected to the real family communication. The remedy for
this situation involved two steps. First is to create a categorical variable (0 = no access. 1
= access). This step does have a limitation in that we do not know how the strength of
access influences the value orientation just that having access is better than not having
access. The second step is to account for individuals with access. but not connected to
the family group. These individuals were visually identified using the family

communication sociograms and coded as 0 (no access).

Model 3 — Access to. family communication will decrease an individual 's value
orientation

Levell

Value Orientation” 2 BO,- + fill-(Family Member”) + B4j(l"amily Accessij)+r,-,-
Leve12

180/ : 1’00 + #0)

1811' = 1’10

541' = 1’40

This model converged in 6-iterations. and produced a deviance change ofx3 —‘
1488.476 (df'l. p < 0.001). indicating a much better fitting model than Model—2. Using
the proportion of variance explained by the equation. Model 3 explains 4.3% more of
level 1 variance than Model-2. This model is considered a much better fit. and tells us
that having access to family communication will reduce an individual‘s value orientation
by a -1.77. even after the effect of being a member ofthe family subgroup has been

184

controlled. It should be noted that the intercept has increased (15.49) to account for the

effects of family communication access.

H 4.3 The value orientation of the owners will be positively related to individual value

orientation.

To test this hypothesis a fourth model was nested in Model-3. This nested model

contained a variable for owner value on level 2.

Model 4 —- Owner Influence on Value Orientation
Levell

Value Orientation”- : 6’0] + fill-(Family Member”) + Bil-(Family Accessij)+r,-j
Leve12

.80)‘ = 1’00 + y01(0wner Valuel- ) + #0}

1811' : 1’10

341 = 1’40

This model converged in 6 iteration. with a deviance change x2 —- 8.39 (df'l. p <
0.001). This suggests a better fitting model in comparison to Model 3. Additionally the
associated coefficient for Owner Value was 0.63 (t = 4.48. df‘ 9. p ; 0.001). This model
explains 70.43% of the unexplained level 2 variance. In general, the addition of the
owner’s value orientation explains a great deal ofthe between business variance. and the
owner’s value orientation is positively related to an individual‘s value orientation. In
other words. one standard deviance increase in the ownership value orientation will

185

produce an increase of their FOB‘s mean value orientation by 0.63. This is not a large

change indicating that after we account for the subgroup. and an individual's access to

family communication there is little variance left for value orientation.

Since there is very little variance left on level two. the deviance change was small.

and the reliability has dropped to 0.60 the researcher stopped estimating the model here.

Table 4.4.1 : Summary of Second Model

 

Fixed Effects

 

Intercept

Family Member ([31)
Family Communication
(153)

Owner Value (y01)

 

Variance
Q om ponent

Intercept (110)
Level 1 (R) .

 

 

M90615:

Reliability (BO)
Deviance
Deviance Change

 

Null Model(s.e.)

 

l3.59(0.909)**

Model 2

 

14.48(0.98)**
-2.83(0.73)**

 

0.822
2919.606

1K)

 

hand—fir

0.840
2905.507
14.01

3

 

Model_3

15.59( 1 .3)**
-2.25(0.79)**
-l.77(0.82)**

0.85
1417.031
1488.476
4

 

M0491 +1

8.49(1.76)**
-2.26(0.78)**
-1.95(0.81)**

-1.95(0.81)*

' ‘ "'_—'1

2.47

21.22

0.60
1408.64
8.39

5

 

 

Reliability/or Phase 2: Step 2 Model

A boxplot of the within business residuals can be used to determine ifthe

residuals are centered at 0. and that the variances are consistent across groups. Figure 4.6

shows that the residuals seem to be centered at 0. Also. a scatter plot of the residuals

186

against the fitted values is used to assess for problems with heteroscedasticity. Figure 4.7
shows that there are no recognizable patterns. which indicates that the assumptions for
heteroscedasticity are reasonably met. Finally a P-P plot (Figure 4.8) ofthe level 1

residuals show that the data seem to be normally distributed.

187

Figure 4.4.2: Box Plot of Residuals by each ofthe 1 1 F085

l1resid

 

 

 

 

 

 

 

 

 

 

 

 

15000 e
21.7
o
v 1»
10.00CH
12'
C
23?
I.)
5000-1
{H I-‘N F q I f:
4 i
l i; D
l
I-I .
0.001% L "1 i
. 1-1 :34
i l 1 ~11:
6.000% T
1
i
i
-1 1:1 1:11] 1:1 -4 i LI
-19 000 r I I 1 I T I I I f
'1 2 3 E- 0 7 8 9 10 1‘1
12id

188

 

Figure 4.4.3: Scatterplot oflevel 1 residuals against fitted values

 

 

 

o o 0 0
18.000“
00 000000 000000000
0 O O Q Q o o O
O .
O O %UOO%OO%OO OO G
15000-1
O O O O
3 o oo‘boooooe 050090000
5; o o o o
“' 0 o o o c o o
1:1'000_4 (1‘ o (J o k) j
0 Q C- O Q
a o o o o o o o o
C (T) O C
O O o ‘0 O o o O
c o o o
9.001% 0 o o o o c C
O O O
o c o o o o
3000 I l I r I I I
-15 000 -10.000 601—0 0.000 5.000 10 000 15.000
l1resid

189

 

Figure 4.4.4: P-P plot oflevel 1 residuals

Normal P-P Plot of 11 resid

 

 

 

 

 

1 .04
08*
.0
O
I.
a.
g 0.131
0
"O
0
a
8
a 04‘
x
I.l.l
0.2-
013“ I 1 I I I
0.0 0.2 0.4 0.13 0 E: 1.0

Observed Cum Prob

190

Summary of Findings

Through this exploration of the Three Circle Model it was apparent that the
assumptions of the Three Circle Model, regarding subsystems and boundary are valid. but
limited. This study found that individuals who share a membership in the family.
ownership. or nonfamily employee groups tend to share similar communication patterns.
levels of satisfaction and value perceptions with their subgroup members. The limitation
with the Three Circle Model is that it does not account for the strength of the boundary
between subsystems. the value orientation of the FOB or the family dynamics of the
owning family.

The value orientation of the FOB tells us a great deal about the level of
satisfaction within the FOB, with FOBs closer to the family side ofthe Value Continuum
exhibiting lower levels of satisfaction. Furthermore. owning families that are close (high
level of cohesion) have greater levels of satisfaction throughout their FOBs. but owning
families that are close often produce higher value orientations and therefore diminish the
effects ofthe closeness. Within a FOB. the value orientation ofindividuals (or their
perception of the FOB‘s value orientation) is influenced by their subsystem membership
(family members have the lowest value orientation. owners next and employees have the
highest value orientation) and the strength of the family to business boundary. In
businesses where there is a rigid boundary between the family and business. the
individuals who are cut off from the family have a much higher value orientation. That
decreases their levels of satisfaction with their FOBs.

Therefore the best option for FOBs is to start with an owning family that is close

but also encourages a permeable boundary between the family business systems. This

191

permeable boundary will create a unity. or shared agreement for the F OB's value
orientation (and more than likely bring the value orientation closer to the business side of
the Value Continuum). All ofthese together will produce a FOB with high levels of
satisfaction.

Two areas were not explored in this study due to the fact that not enough variance
existed in the sample population to test these ideas. The first is the influence of
adaptability on this model. The families in this study shared similar levels of adaptability
and therefore there was not enough variance available to test the effects of adaptability.
Also there is evidence in the literature that a diffuse boundary between the family and
business will hurt the business. None of the F083 sampled showed a boundary that
could be conceptually thought of as diffuse. Therefore this study does not support or fail

to support the effects ofthese two issues.

192

CHAPTER V: DISCUSSION

Introduction

The general purpose of this study was to evaluate and strengthen the foundational
theory within FOB literature. the Three Circle Model. The mixed method approach
highlights the importance and limitations of the Three Circle Model. This study also
points to the significance of integrating the Three Circle Model assumptions with: 1) the
owning family dynamics of adaptability and cohesion, 2) the value orientation of the
FOB (whether the FOB values the growth and development of the business system.
family system, or a balance of both). and 3) the strength of the boundary between the
family and business systems. While the findings from the exploration of family dynamics
and value orientation are important and novel to the field. the findings regarding the
boundary strength are the most important addition to the current literature. and offer a
systemic solution to strategic plans requiring a movement along the Value Continuum.
Movements along this continuum are often necessary as the economic environment
changes (Distelberg & Sorenson. 2009).

The conclusions from this study, and in relation to the current field of literature
suggests that: l) in general FOBs with a total value orientation closer to the business side
of the value continuum have higher levels of satisfaction. 2) value orientations vary by
subgroups with employees seeing the FOB closer to the family side of the continuum in
comparison to owners and owning family members, 3) close owning families have FOBs
with higher levels of satisfaction as long as their total value orientation is the same or

below the sample population mean. 4) FOB family systems should be adaptable. 5)

193

satisfaction at the individual level (i.e. individual FOB members) is closely tied to the
degree of unity in the F OB‘s value orientation, conversely, FOBs with a large
discrepancy of value orientations at the individual level will have lower levels of
satisfaction. and 6) while owners influence the value orientation of individuals within
their FOBs. being connected or having access to family communication is a powerful tool
to unite value orientations across a FOB.

These conclusions are discussed in more detail below. as well as suggestions for

individuals working with or conducting research on FOBs.

Discussion of Results
The following section outlines the findings of each phase ofthe study. These
findings are grouped into two sections; 1) the measured limitations of, and the proposed
integrations to the Three Circle Model, and 2) integrating the study findings for family
dynamics (adaptability and cohesion). the value orientation of the FOB. and the strength
of the boundary between the family and business systems into the new Expanded Three

Circle Model.

Evaluating the Three Circle Model
The first Specific Aim of the study focused on testing the validity of the assumed
structure within the Three Circle Model (Taguiri & Davis, 1982). Hypothesis 1.] stated
that the model had many benefits. but that this model did not fully account for the actual
functioning of a FOB. In this study, functioning was measured by the communication
patterns within eleven FOBs. This study operationalized “functioning" as the total

communication matrix measured by the sum of the three network items within the

194

participant survey. This matrix is a good measurement of the functioning ofa system for
two reasons. F irst, general systems theory is rooted in cybernetics which relies heavily on
the assumption that communication is a function ofa system. whether that system is open
or closed. Closed systems allow communication to move within the system while open
systems allow communication to move within and across systems. Secondly. the total
communication matrix is a weighted matrix which is a more robust variable and accounts
for the strength of relationships within a system rather than simply measure whether a
relationship exists. Therefore the use of the total communication matrix as a
measurement of functionality is in line with general system theory. and it provides a
robust evaluation of functioning within 1085.

In summary ofthe first step in Phase 1. for many ofthe sampled FOBs the Three
Circle Model does indeed explain interactional patterns within businesses. There is some
proof that members of each subgroup interact with each other. suggesting some
commonality or substantive grouping similar to the Three Circle Model subgroups. Even
though there was some support for this model. the amount of variance that it explained
was minimal. For three FOBs where the subgroupings produced statistically significant
values. the groupings accounted for less than 1.8% ofthe total variance in
communication.

While most ofthe FOBs did not fit the Three Circle Model. two FOBs fit the
Three Circle Model well (more than 10% of the variance was accounted for by the Three
Circle Model). but the level of satisfaction across these businesses was dangerously low.

The Three Circle Model accounted for 42% ofthe variance in Company 10. but the

195

employees in this FOB showed the lowest level of satisfaction across all the employees of
all FOBS in this study.

The finding in this step provided support to the first hypothesis which stated that
the Three Circle Model may be a good foundation, but by itself it does not provide
enough explanation to be valuable without integrating other systemic concepts.
Furthermore the findings from company 10 and 3 suggest that the model might actually
be a model of dysfunction rather than health when the subgroup boundary assumptions

are followed too rigidly. This exercise added support to expanding the assumptions of the

Three Circle Model.

Expanded Three Circle Model

Chapter I discussed three possible variables that could be used to expand the
Three Circle Model: 1) the inclusion ofthe owning family‘s dynamics of adaptability and
cohesion. 2) the inclusion ofthe value orientation ofthe business. and 3) the inclusion of
system boundaries between the Three Circle Model subgroups. These variables are taken
directly from the current literature and each has been purposed as integrations to the
Three Circle Model, although they have little direct supporting empirical evidence. Step
2 in this study sought to evaluate these integrations and determine which. if any. have
value in expanding the Three Circle Model.

Specific Aim 2 explored these ideas through in depth case studies of each of the
l 1 businesses. This exploration utilized social network analysis. family science. and

FOB empirical tools to develop hypotheses that were tested quantitatively in Specific

196

Aims 3 and 4. The following section describes how each of these areas adds strength to
the existing Three Circle Model.
Subgroup Membership

As discussed above. the Three Circle Model does have limitations in explaining
communication patterns. but it should not be ignored as the subgroups within the model
do provide some insight into FOBs especially when we consider subgroup differences in
value orientations. Furthermore. this study did find some patterns consistent across F 085
which can be attributed to subgroup membership. These attributes are discussed below.

From Phase 1: Step 2 it was hypothesized (based on the case study explorations).
that the family and ownership groups have higher levels of satisfaction and lower levels
of value orientation (closer to the business side of the continuum) compared to their
employees. In most of the F 08s in this study the highest level of satisfaction was in the
family group. followed by the ownership group. and then the employee group. Similarly.
the lowest level of value orientation is often seen in the family group, then the ownership
group, with the highest in the employee group. In other words, family members tend to
perceive the FOB as more professional and business like than do their employees. who
tend to see the FOB as more informal and privileging members of the owning family. The
relationship between subgroup membership and value orientation was further supported
in Phase 2: Step 2, where it was found, that family members rate the FOB value
orientation 2.26 points lower (on a 28 point scale. Mean = 14.1. SD = 5.3) in comparison
to nonfamily members. In other words. family members. in general. see the FOB closer to

the business side of the value continuum in comparison to their employees.

197

While subsystems have some predictive significance. there are limitations. and
issues such as the owning family’s dynamics, the FOB’s value orientation. and the
strength of the family-business boundary. When these areas are combined with the
subgroup membership findings the Three Circle Model can provide more insight into the
functioning of F OBs.

Family Dynamics: Adaptability

The actual findings for adaptability in this study are inconclusive. There was no
measured relationship found for the owning family’s level of adaptability on satisfaction.
value orientation. or communication patterns. This is more than likely due to all of the
F 085 in this study scoring at or above the mean level of adaptability (compared to the
national average). Also. the scores for adaptability at the FOB level were relatively
similar which provided very little variance to explore.

Although the actual measurements for adaptability for this sample population did
not produce significant findings. we should not disregard the affects of owning family
adaptability. When we view the findings in this study alongside the conclusions from
other studies that used the same measure of adaptability (Burke. 2007; Lansberg &
Astrachan, 1994; Zody et al.. 2006), and studies that used measures that are conceptually
similar (Anderson et a1, 2003; Anderson & Reeb, 2003; Bahrami, 1992; Krijnen; 1979;
Overholt 1997; Zahra, 2005) it appears that adaptability is important for FOB survival. It
is likely that the reason lower levels of adaptability were not found in this study is that a
lower level of adaptability decreases the likelihood of survival for F085. and these
businesses (less adaptable FOBs) would feel the greatest pressure from the 2009

economy. Therefore. they would have declined to participate due to the enormous

198

economic stress during the data collection time frame. or possibly they failed to survive
as F OBS.

This study offers one hypothesis for future testing: There is a threshold/or FOB
family systems and adaptability. This study would suggest that family systems that do
not meet the average (and more than likely score below the average) for adaptability on
FACES III will have difficulty surviving as a FOB system. Future longitudinal studies of
FOBS could learn whether this threshold exists by studying F OBs with owning families
who have varying levels of adaptability. This methodology would have to identify F 083
in their early stages of development. as done by Davis and Stems (1981). as well as the
findings from this study suggest that families with lower levels of adaptability may not
survive past the initial startup phase.

Family Dynamics: Cohesion

Cohesion. or the level of closeness and distance within an owning family, does
add value to the Three Circle Model. Findings from this study suggest that the closer the
owning family, the higher the level of satisfaction across the FOB. However. there are
some limitations to this explanation. It was found in Phase 1: Step 2 that cohesion and
value orientation have an interaction effect on satisfaction. meaning that the positive
effects of higher levels of cohesion are reduced when that family has a FOB with a value
orientation closer to the family side ofthe Value Continuum. Figure 4.2.35 in Chapter IV
illustrates this relationship and shows that the danger associated with a close owning
family is that they may inadvertently privilege a high value orientation for their F OBS.
and when this happens, the positive effects of cohesion diminish. Therefore, cohesion has

a positive relationship with satisfaction as long as the value orientation of the FOB is

199

closer to the business side of the value continuum. This finding was further tested in
Phase 2: Step 1 where the level-l negative slope for value difference and satisfaction was
found to be magnified by the closeness of the owning family. More specifically the slope
for value difference on satisfaction was found to be -2.08. and the level-2 slope of
cohesion on value difference was -2.26. This tells us that the higher the level of cohesion
in the owning family the greater the effect of value differences on satisfaction. Or,
although we know that an increase in value difference will decrease satisfaction in an
individual, the decrease is more significant when the individual is in a FOB with a close
owning family. Therefore, FOBs with close owning family systems are good for the
FOB. But F OBs with close owning families need to be careful not to let their FOB also
develop a high value orientation (closer to the family side of the Value Continuum). If
both exist in a FOB the level of satisfaction will likely be low.

The findings from this study are in line with previous research on FOBs and
cohesion. Previous studies have consistently found that owning families with higher
levels of cohesion have less conflict throughout the FOB (Zody et al.. 2006). work
together more effectively (Lee. 2007). and have better strategic planning skills (Lansberg
& Astrachan, 1994). Unfortunately. previous research has been unable to find interacting
effects with cohesion, or anything resembling the curvilinear hypothesis of Olson et al..
(1979a; 1979b). One possible explanation is the disregard to cautions within family
systems research suggesting a multi-ratcr methodology over a single rater method in
studying the curvilinear effects ofcohesion (Thomas & Ozechowski. 2000). This study
is the closest representation of the hypothesized negative aspect of the upper end of the

cohesion scale with FOBs as the study population. In this study. families with higher

200

levels of cohesion magnified the negative relationship between value orientation and
satisfaction. While this study does not offer a definitive causal relationship between
cohesion and value orientation. the results from this study offer a similar caution as Olson
et al.. ( l979a: l979b) for family systems on the upper end of the cohesion continuum.
Olson (2000) cautioned that maladaptive behaviors develop when family systems are too
close.
Value Orientation

Value orientation is a complex variable and its effects change depending on the
level of analysis within the system. For example. when we look at value orientation as
the average value orientation across a FOB (i.e.. the mean value orientation for all
individuals in a particular FOB). we are measuring the actual value orientation ofa FOB.
When we take this approach we see a negative relationship between value orientation and
satisfaction. FOBs closer to the family side of the Value Continuum have. on average.
lower levels of satisfaction. Both Specific Aims 2 and 3 showed this relationship. Model
5 from Specific Aim 3 is the strongest evidence of this relationship and shows that the
overall value orientation ofa FOB accounts for approximately 41% of the differences
between businesses for satisfaction. Similarly. Specific Aim 2 (in Phase 1: Step 2)
showed that the r-squared from the Value Orientation-Satisfaction slope in figure 4.2.34
is 0.3888 (or 38.8% variance explained). Therefore. we can generalize from these
findings that approximately 40% ofthe between FOB difference in satisfaction is due to
the overall value orientation of each FOB. This is a negative relationship where the

greater the value orientation (closer to the family side) the lower the level of satisfaction.

201

When looking only at this level of analysis one could conclude that F 085 that are
closer to the family side ofthe Value Continuum are less successful (defining success as
the level of satisfaction throughout the business). In this case Dyer‘s (2006) argument to
professionalize the F OB, or take strides to make the FOB more business-like and reduce
family characteristics, would seem logical; however the relationship between satisfaction
and value orientation is slightly more complex.

Although many have assumed that perceptions such as value are unified across
owners, family members. and employees (Dyer, 2006; Fleming. 2000; Galvin et al..
2007) this study challenges this assumption and shows that there is not a great deal of
unity in value orientation within FOBs. This is illustrated by the unconditional model in
Specific Aim 4 (where 75.4% ofthe variance in value orientation is within businesses
and only 24.6% is between businesses). This brings to light two limitations with the
professionalizing hypothesis. First, the professionalizing hypothesis (Dyer. 2006)
assumes that the owners know that the FOB is not professional already. Often, as found
in this study, the owners perceive the FOB as closer to the business side of the Value
Continuum. in relationship to their employees. Since owners tend to see their FOBs
closer to the business side of the Value Continuum already. the suggestion to
professionalize would seem like more of the same. This may be a missed opportunity to
help owners who. rather than being too close to the family side of the value continuum.
are not in tune with the perceptions of their employees.

Secondly, this perception problem is not just a structural issue (where too many
resources are transferred into the family) but a systemic perception problem involving

owners, family members and employees. Phase 1: Step 2 found that an individual’s level

202

of satisfaction is negatively related to the distance he/she is from the average level of
value orientation within his/her business. For example. if business A has a total value
orientation of 14.1. and two individuals B and C, within business A have corresponding
value orientations of 14.3 and 15.7. it is likely that individual B (with a value orientation
score of 14.3) also will have a higher level of satisfaction than individual C. This
relationship was further supported in Phase 2: Step 1 where it was shown that after
controlling for the value position ofa FOB (at level-2). there was little effect from an
individual's value orientation (level-1 ). However there was an additional negative effect
for the difference between an individual‘s value orientation and the mean of his/her F OBs
value orientation. This relationship reduced an individual's satisfaction by an estimated -
2.08 level-1 coefficient. while the level-2 value orientation reduced individual
satisfaction by -1.19. In other words. the effect of having a value orientation that varies
significantly from the FOB mean is much greater than the negative level-2 relationship.
In summary. while the overall value orientation ofa FOB is important.
satisfaction is affected to a greater extent by unifying the values within a FOB. This
finding is supported with nearly three decades of theory and research on the positive
effects of unifying values and goals within FOBs (Davis & Stern, 1981; 1996; Galvin et
al.. 2007: Sharma. 2004). Furthermore. in many cases the problem is a perception
problem and not a family versus business structural problem. The latter can be addressed
with the structural resource transfer changes in the professionalizing hypothesis: the
former requires a more systemic solution that involves communication or boundary
evaluations and modifications that are addressed below in the family boundary

discussion.

203

Family Boundary

In the FOB literature there has been some debate about the role of boundaries
between the family and business systems. In some theories it has been suggested that
F OBS should maintain a somewhat rigid boundary between the family and business
(Blanco-Mazagato. de Quevedo-Puente, & Castrillo, 2007;. Dyer, 1986; 2006; Levinson,
1971: Fleming. 2000). Most often these theories encourage FOBs to strive to resemble
non-FOBs by limiting the amount of resource transfers from the business to the family.
and building in stronger boundaries between the family and business. While theories like
these gather support, empirical evidence continues to disprove the rigid boundary
hypothesis. For example Olson et al.. (2003). Zahra (2005). and Zody et al. (2006) all
have shown that when a rigid boundary is in place within FOBs. the business does do
better (in terms of revenue growth) but there is increased conflict within the family and
ownership subsystems. This study supports these findings. In this study a rigid boundary
was found to increase conflict throughout the business by increasing the value perception
differences between employees. owners and family members.

This study operationalized the boundary between the family and business systems
as communication interactions between individuals in each subsystem. More specifically
two types of communication were measured. total communication and communication
specific to the owning family. It was theorized. based on the assumptions of the Three
Circle Model, that if a rigid boundary existed within a FOB system there would be little
to no flow of communication across subsystems. Specific Aim 1 explored this hypothesis
by fitting the Three Circle Model across the total communication matrix in F083.

Conceptually. if the model fit well for a FOB, it was due to having rigid boundaries

204

between the subsystems which made communication greater within subgroups than
across subgroups. Since this model fit well only for FOBs with very low levels of
satisfaction. it was concluded that the rigid boundary hypothesis was incorrect and
actually decreases satisfaction within FOBs.

This study did not stop at this finding but also measured the patterns for
communication specific to the owning family. Two separate measures were used for this
communication pattern. The first was the block modeling analysis used in Phase 1: Step
2. In this exercise there was evidence ofa relationship between the strength of the family
communication boundary and individual value orientations. FOBs that exhibited a strong
family-business boundary (such as Companies 2. 7 and 10) had an associated decrease in
satisfaction and an increase in value orientation differences for individuals who had been
cut off from family communication. A closer examination of companies 4. 5. 6. and. 9
showed this same relationship between the boundary strength and value orientation. This
examination ofcompanies 4. 5. 6. and 9 is particularly interesting because this finding
compared employees within the same FOB. thereby limiting almost all possible unknown
variables. F urthermore, the measurement of Family Access in both Phase 1: Step 2 and
Phase 2: Step 2 showed that individuals who were cut off from family communication
had a greater disagreement in value orientation (from their FOB mean value orientation)
and Phase 2: Step 1 showed that this disagreement has a strong negative relationship with
satisfaction. Therefore a rigid boundary will decrease a F OB’s ability to unify
individuals around a shared value orientation. In cases where non-family employees were

cut off from family communication. the result was developing pockets of isolated

205

networks that tended to reduce satisfaction and increase the distance between the real
FOB value orientation and an individual’s perception of the FOB value orientation.

It should be noted that the results from this study do not suggest that
professionalizing a FOB is a bad thing. It is highly likely that FOBs should be able to
move freely across the value continuum as external and internal events may require
temporary moves (Distelberg & Sorensen, 2009). In other words. there are times where a
FOB should be closer to the family side of the Value Continuum. such as when the
family moves through a transition or encounters an environmental stressor. At other
times. a FOB should be closer to the business side of the Value Continuum (or in other
words professionalize). such as when there is an economic down downturn because the
business requires added resources to manage the additional stress. Overall a FOB should
be able to move along the continuum when external or internal stimuli require a move.
Therefore. this study does not suggest that one position on the Value Continuum is better
than another. but offers a strategy for moving along this continuum.

According to this study, the strategy for a FOB that required a move from the
family side of the value continuum to the business side (or professionalizing) would
include an assessment of the current value position of each subsystem and the strength of
the family-business boundary (it is also likely. but not supported in this study. that the
owning family’s level of adaptability would be important). In other words. if a FOB has a
value orientation closer to the family side of the Value Continuum. and the current
economic downturn required the FOB to shift closer to the business side of the value
continuum. the first step would be to encourage the family system to shift closer to the

business side of the value continuum. Unlike the assumptions in the professionalizing

206

hypothesis. this study does not assume that the employees of this FOB will make the
same shift. If a rigid boundary exists within this FOB, the family will make the move but
the employees will either maintain the same position or move even closer to the family
side ofthe Value Continuum. which would result in conflict. Ifa rigid boundary is in
place. the next step would involve creating lines of communication between the family
and employees. By taking this action the employees would decrease their value
orientation and increase in their level of satisfaction.

The findings from the boundary exploration showed that having access to family
communication is not only important for family members but also for non-family
employees. This finding is not completely new to the field of F OB, as “family meeting”
and “family council” theories have previously highlighted the importance of facilitating
communication about the owning family within FOBs (Arnoff& Ward. 2002:
Habbershon & Astrachan. 1997: Tower. Gudmundson. Schierstedt. & Hartman. 2007).
While the concept is not new. this study is one of the first empirical tests. and more
importantly this study describes the relationship between boundaries. value perception
and satisfaction. Therefore this study bridges the gap between the family meeting
literature and the boundary research (Haynes et al.. 1999; Kaye, 1991; Olson et al.. 2003;
Stafford et al.. 1999; Zuiker. et al. 1998).

Due to the importance and complexity of this finding it is important to illustrate it
in the following short summary of the unintended consequence of Company 103 rigid
boundary. Prior to the study. Company 10 was in the process ofa generational transfer of
ownership. At the end of 2008. it became apparent that Company 10 could no longer

financially sustain two owners (the father and the son). Therefore they began a plan to

207

buy out the father. This would be considered a shift towards the business side ofthe
value continuum as the goal was to reduce (in the long term) the amount ofbusiness
resources moving towards the family. Both the father and the son understood the long
term goals, and thought that the plan was in the best interest of the business even though
both would have rather had the business stay in the current ownership structure. The
problem with this move was that the father and son felt they should also increase the
strength of the family-business boundary. From their perspective they wanted their
employees to see the business as a real business and less like a family business. The
effect of the stronger boundary was that the employees saw the business even more like a
FOB with father and son having many talks outside of the business. and many structural
changes happening that “they weren’t privy to”. After this author spent some time with
the father. son and two key managers (post data collection). it was apparent that there was
a great deal ofmiscommunication and incorrect perceptions about the future of the
business. After only two meetings, the business built in lines of family-business
communication and the conflict and misunderstandings have been dramatically reduced.
Company 10‘s experience demonstrates the importance ofthese boundaries. It
also shows the interactions between subgroups, satisfaction. value orientations. and
boundaries. Furthermore it shows how easily these concepts can be overlooked in
practice and how with very little investment they can have a dramatic effect on the

satisfaction within a FOB.

208

Discussion of Methods: Limitations

While the findings from this study bring much needed insight into the role of
family dynamics. value orientations. and boundaries. there are a few limitations regarding
the methodology and generalizability of these findings.

Three issues can be considered a limitation of the methods used in this study.
First the sample size of businesses may be considered small by some. Also having only
1 1 businesses on level 2 is a limitation for HLM methodologies. It has been suggested
that HLMs should have at least 30 groups on level 2 with at least 30 individuals in each
group (Snijder & Bosker. 1999). While it is possible to have fewer level 2 groups when
there are more than 30 individuals in each group (this study had 73 individuals on
average in each group). we should still consider this sample population somewhat small
for HLM and therefore we need to interpret the cross level interactions with caution.

One cross level interaction was proposed in this study. In Phase 2: Step 1
cohesion was modeled as a cross level interaction. The danger with this model is a type I
error, because the small sample population may produce artificially low variance
components which would shrink the error term and create an artificially smaller
confidence interval (Raudenbsuh & Bryk, 2002). C onceptually the worry here would be
that even though cohesion increased the magnitude of the value difference-satisfaction
slope, this finding may not be accurate or even true. If this study proposed this model by
itself we would probably disregard the cross level effect of cohesion, but this study found
this same effect in Phase I: Step 2. therefore even though this is a statistical limitation.
support from other methods reduce the concern of a type I error, and we should have

confidence in the finding that value perceptions and cohesion interact.

209

3131

his

3113

“-1111

ft (1 n

A second limitation is that the sampling techniques in combination with the small
sample (at level 2) may have biased the F OBs that participated in the study. For
example, while close to 70 F083 were invited to participate in the study. only 1 1
businesses in this one mid-western state actually participated. Also, these FOBs may be
substantively different from the actual population of FOBs because of their interest in the
researcher and the study. The researcher built trust with these businesses prior to the
study through previous research with three Nonprofit membership groups in the area and
through his writings in Family Business publications. Similarly. the F085 that
participated were interested in learning about their FOB in comparison to the other F 08s
in the study. Also. most of these FOBs maintained a membership with a nonprofit group
that specialized in FOB issues. All of three ofthese issues likely influenced which FOBs
selected in and out of the study. For example. these FOBs were possibly more self aware
of the effects of family ownership. F OBS that are active in the FOB community and
aware of their FOB status may be different from FOBs that are not active in the FOB
community and do not understand that their status as an FOB has effects on family and
business functioning.

A similar limitation comes from the individual level sample size. While there
were close to 900 individuals associated with these 1 1 FOBs that could have been
studied. only 492 individuals actually took the survey. It is unclear whether the 400
individuals that did not take the survey would have significantly different experiences.

The third and largest limitation to the study was the economic environment in
which the study took place. It is largely agreed that January 2009 was a time of

economic depression. Since this study collected data from January to mid-April. it was

210

11111111

11351

be:

10111

1110'

tics

limited by this economic environment. The most significant effect of this environment
was the lack of participation by FOBs. In Chapter 111 it was noted that 12 additional
businesses were originally interested in participating but by mid January they declined
because they had numerous concerns about the economy. In two cases the FOB owners
told the researcher that they reduced their employee count by 80% and did not want to
know the level of satisfaction within the business right now. It is entirely possible that
these 12 businesses might have had the variance this study needed to examine different

levels of adaptability in the owning family.

Discussion of Methods: Strengths

There are four major strengths to the study that set it apart from other FOB
studies. The first strength is the holistic sampling process. This study is the first study in
FOB literature to attempt to sample all members of the FOB system from the employees
to non-employed family members. Some studies have sampled multiple members ofa
FOB. but no study to date has produced a sample population ofthis depth. The benefit of
this sample population is the ability to measure the actual value orientations and
boundary strengths within F OBs. Previous research that has explored these areas has
done so by sampling one representative from each FOB system. While these studies are
able to sample more businesses. their results are somewhat limited. As we saw with this
study. owners. family members and employees often have different experiences. For
example, the owners of company 10 saw their FOB as close to the business side ofthe
Value Continuum and they were very happy. If we only had sampled the owners of this

business. we would have held this business up as a model of health. But when we went

211

farther and sampled the family members and employees. we see real problems with this
business, giving it the lowest level of satisfaction across all 1 1 businesses in this study.

The second major strength of this study is the use of both qualitative (social
network case studies) and quantitative methods. The qualitative exploration offered
valuable insights in the functioning of FOBs. These insights were then developed into
testable hypotheses. These hypotheses were then tested quantitatively using HLM. Since
the level 2 and level 1 sample size were relatively small for IILM, the findings by
themselves could be subject to misspecification within the HLM models. But the
qualitative findings added support and verified the HLM findings. Taken together this
mixed methods approach added considerable insight, which allowed for a more complete
explanation of FOBs through the Three Circle Model.

The third strength of this study is its roots in empirical and theoretical FOB
literature. This study began by examining the Three Circle Model which is the most
referenced theory in FOB literature. The proposed expansion of this model is also
gaining support in the FOB field as a quality integration of Human Ecology. family.
organizational. and FOB theory (Distelberg & Sorensen. 2009). Therefore the hypotheses
and research questions have already been proposed in the literature of FOB and thought
to be important aspects of FOBs. Furthermore. the findings ofthis study are in line with
current trends in the literature. For example. other empirical studies have found the same
relationship between cohesion, adaptability, and success. Other studies have found that
subgroup membership affects an individual’s experiences with his/her FOB. This study
strengthens these previous findings by using a more in depth sampling process. Also. this

study presents the first integration of family dynamics. boundary strength. and

212

satisfaction in the literature. This integration is very important given the overall
agreement ofthe validity of General Systems Theory in FOB literature.

The fourth strength of this study is that the sampling procedures created a sample
population of businesses with a good representation of the demographic issues known to
influence FOB research. It was noted in Chapter II that the gender ofthe owner. the
generation ofowner. the size ofthe FOB, and the industry ofthe FOB affect outcome
variables. This study represented all of these areas (female and male owners. ownership

nd 3rd

in founder, 2 and 4‘h generations, revenues ranging from $200.000-90 million,

employee size from 8—500. and multiple industries).

Implications for Family Owned Businesses

This study offers an in depth and complex discussion of functioning and health
within FOBs. Four points from this study are important to FOBs and should be
highlighted. First. FOBs should be aware ofthe effects of their value orientation. The
greatest awareness should be given to the overall value orientation of the FOB. as this has
the greatest effect on satisfaction across the FOB. In addition, owners of FOB may
incorrectly assume that since they believe their FOB is closer to the business side of the
Value Continuum that others within the FOB may not have the same perception. For
example. Companies 1. 2. 5. 6 and 10 were extremely surprised to learn that their
employees believed that the FOB was closer to the family side of the Value Continuum.
These business owners believed that they had done an effective job of convincing to their
employees that they were working in a FOB that valued the business system over the

family system. There is strong evidence in this study that the strength of the family-

213

business boundary is an important predictor of unifying the value perception across the
FOB.

While there is an effect of subgroup membership (employees have lower levels of
satisfaction and see the FOB closer to the family side of the Value Continuum) this effect
can be mediated by the boundary between the family and business system. For example,
in F OBs where there was a permeable boundary for family communication there was a
much greater level of agreement on the FOB’s value orientation and satisfaction. This
creates an interesting and somewhat counterintuitive situation for FOB owners who
believe that their FOB is closer to the business side ofthe Value Continuum. More
specifically. if an owner believes that their FOB is closer to the business side of the Value
Continuum they will probably attempt to limit the amount of “family communication"
throughout the business. For example an owner may try to produce an FOB where the
value orientation is close to the business side of the continuum. This owner may
discourage conversations about the owning family at work to achieve this end. In other
words. creating a rigid boundary between the family and business system. While on the
surface this makes sense. it may have a very negative effect. In this example when the
owner employs a rigid boundary. individuals on the cut offside ofthis boundary will
increase in their value orientation which is the opposite of what the owner was attempting
to do. The better option for this owner would have been to maintain his or her value
orientation but also encourage more communication between the family and business.
This would reduce the non-family member‘s value orientation and create a unified value

orientation which is closer to the owner's.

214

As seen with other studies of family dynamics. this study shows that the family
dynamics of the owning family have an effect on the FOB. There is some evidence from
this study that families that are not adaptable will not succeed in the F OBs. Also for the
first time the level of cohesion was seen to affect the FOB. While in general families that
are close do better than families that are distant. but this closeness has an interaction
effect with the value orientation of the FOB. Families that are very close have a danger of
producing a value orientation that is high and thereby reducing the level of satisfaction
throughout the FOB. These two findings together support the idea that owning families
have an effect on their FOB. It would be wise for owning families to work on their level
of closeness and work together to create a permeable boundary between the family and
business. These are difficult tasks and would be best addressed through methods
previous discussed in the FOB literature regarding unity of the Owning family. For
example family councils and family meetings may be very helpful in this venture ((Arnoff
& Ward. 2002: Habbershon & Astrachan. 1997: Tower. Gudmundson. Schierstedt. &
Hartman, 2007).

The most significant implication for FOBs from this study is the exploration of
non-family employee experiences. In this study it was shown that employees in general
have a lower level of satisfaction and a higher value orientation than family members.
These differences between the employees and family members become very problematic
when there is a strong boundary between the family and business. Many theories have
suggested that employees benefit from a “professionalized” FOB (Dyer, 2006; F leming,
2000). Theories like these tend to suggest that employees would rather not be involved in

the family’s business. But this does not seem to be true for this study. When employees

215

are connected to the family communication they have a similar value orientation to their
FOB owners, and they have a higher level of satisfaction. Also. ifa FOB develops a rigid
boundary between the family and business, this boundary does not limit the
communication regarding the family. Rather. it produces two separate networks of
communication, one within the family systems, and one within the cut off nonfamily
employees. This cut off network seems to reduce satisfaction and increase the value
orientation of the non-family employees. While this study did not explore the content of
communication in these cut off networks. it is likely that the information being circulated
is not accurate as it is not connected to a source of accurate information (the family

network).

Implications for Future Research

There are three important implications for future research that should be
highlighted. The first comes from the methodology of this study. This study used a
sampling procedure that allowed for the inclusion of family. owners, and employees. It
was clear from this study that these three groups have varying experiences, perceptions,
and levels of satisfaction. The measured differences between these groups suggests that
other studies that measure only one individual from each FOB will not produce reliable
findings for the entire FOB system. Since the vast majority of FOB empirical research
samples only the owners. we should view the findings within these studies with caution.
This situation is particularly problematic when the outcome variables of interest involve
the effects ofthe owning family on the FOB system.

Secondly. this study shows support for the interactions between family dynamics.

value orientation. and boundary strength. Therefore. future studies that explore these

216

areas need to consider the interaction effect ofthese issues. For example. when exploring
the relationship between adaptability and FOB functioning. we need to consider the
interaction effect of the FOB value orientation, the level ofcohesion within the owning
family. and the strength of the boundary between the family and the business system.
There are a number of findings in this study that should be explored in more detail
in future studies. The first would be the effects of owning family adaptability. Other
studies have explored adaptability and found evidence of a relationship between
adaptability and success. but the current study did not find this same relationship. This is
more than likely due to sampling issues (the small N on level 2. and the 2009 economy).
Future studies may be able to explore this relationship more directly. or with other
outcome variables. Secondly the family-business boundary should be explored for
varying effects of strength. In this study the boundary was conceptualized as an
individual having or not having access. This allowed for the finding that having access is
better than not having access. But this conceptualization of the boundary did not account
for different levels of connectedness for individuals. For example do highly connected
individuals vary in perception and satisfaction compared to individuals with less ofa
connection? Or it may be possible as in Burke (2007) and Hatum and Pettigrew (2004).
that a connection to family communication is a curvilinear relationship where having too
much access has a negative effects on the individual as well as the FOB system. This
situation would create a similar continuum for connectedness as we have for adaptability
and cohesion where no connection and being too connected is problematic but having a
medium amount of connection is good. This exploration should build on the findings

from this study and consider the interaction between connectedness and subgroup

217

membership and family dynamics. For example the optimal level of connectedness may
vary by subgroup. and different family dynamics may create different levels of

connectedness.

Implications for Systemic Clinical Interventions

This research points to one of the foundational assumptions of general system
theory and that is that systems. while unique, follow basic rules of functioning by which
both big and small issues within a system can affect individuals within the system and the
system as a whole. Systemically trained clinicians who are effective in working with
families should be able to transition seamlessly into working with F OBS by relying on
their knowledge of general systems theory. For example problems that develop in family
system due to ineffective functioning ofa family system will develop in quite the same
way in a FOB. The following is an illustration of how family system concepts of
functionality relate directly to FOB functionality.

First. the most direct comparison of family and FOB functionality was seen in this
study with the exploration ofthe role of cohesion in owning families. Family systems
practitioners are aware that families who are lower on the cohesion scale tend to produce
cut offs within the family system. This same pattern was seen in FOBs in this study.
Company 2 had a lower score for cohesion and this family had a child that was cut off
from the family system. The family even asked the researcher to not contact that
individual for this study. A very similar pattern was seen in Company 7. Ilow this
develops at the FOB level is that a rigid boundary between the family and business

systems leads to a cut off between important components of the FOB often leading to

218

difficulties within the FOB. Either there is a rigid boundary between the two systems or
sections/subgroups within the FOB are cutoff from other systems within the FOB.

On the other side of the cohesion continuum the comparison between the family
and FOB systems is not as direct. Looking at Company 10, the owning family is very
close with the highest cohesion score in this sample population. For family systems
theorists, enmeshment means that there are diffuse boundaries between subsystems.
While this family is very close. it is cutoff from the business system. While this cut off
functioned in its intent to protect the employees from the family communication. it
created low levels of satisfaction in employees. This cut off is similar to family systems
that “protect“ their children by not letting them interact with external systems. In the
case of company 10. the boundary around the family system is strong. The problem is
that this strong boundary around the family system prevents them from forming a
permeable boundary between the family and business system. This is not to say that the
family to business boundaries will always follow these two examples of cohesion. but
rather to explain that family cohesion does influence the family to business boundary.

Similarly. adaptability has been seen to effect the functioning ofthe FOB. This
study does not provide definitive results but does provide limited findings suggesting that
FOB families need to be adaptable. Future studies may show that this is the most crucial
element of family systems in FOBs. It may be that without a high level of adaptability in
the owning family. the FOB will not survive long. especially if external economic
stresses develop.

Secondly, many family system clinicians are already working with F OBs. We

have to assume that if62"/o ofthe North American population is employed by a FOB. that

219

nearly 60% of all individuals (and possibly more) seeking family therapy. are directly
influenced by FOBs. While it may not be practical to enter into a FOB when the client is
an hourly 3rd shift worker. it makes good systemic sense to work at the FOB level when
the client family is also the owner ofa FOB. Systems theory tells clinicians that
interventions are more effective when they involve more components ofa system. For
example adolescent substance abuse treatments are beginning to focus more and more on
the adolescent‘s surrounding family and community context. This is also why family
therapists strive to work with families rather than individuals alone. In this same fashion
family therapists should seek to understand how their client’s family system influences
their FOB and vise a versa.

Third. effective systemic interventions with FOBs will come directly out of good
systemic theory just like good family based interventions are solidly rooted in systems
theory. In this study it was found that the best option for FOB functioning was to have a
family system that was: 1) close. 2) that had a lower level of value orientation at the FOB
level. and 3) achieved a high level of unity for value orientation at the individual level. If
one of these areas is not optimal for an individual FOB. the intervention would closely
mirror family system interventions. For example if there was not unity in a value
orientation in a particular FOB one should look first at how communication flows
through the system; if there were cut offs within the FOB. the goal or intervention would
be to build communication bridges. This mirrors family therapy. When a problematic
behavior develops in a child, the systemically oriented therapist would evaluate how

communication is used to perpetuate the problematic behavior.

220

In summary. what is known in family systems theory and practice regarding
systems and function will hold true also in FOB systems. Therefore a good systemic

therapist will be able to understand and work with FOBs.

Concluding Remarks

In conclusion. this study is a step forward for the field of FOB. It was shown that
the original Three Circle Model has some merit in explaining differences across FOBS.
but there are noticeable limitations. especially in identifying functioning within FOBs.
This study found that the Three Circle Model can be strengthened by integrating
boundary strengths. value orientations. and family dynamics within the Three Circle
Model. While the field has begun to recognize the importance of boundaries within FOB.
it is often theorized that these boundaries should be strong, or prevent business to family
interactions. This study. along with other empirical research caution against this rigid
boundary concept. and suggest that a permeable boundary is the most beneficial for
FOBs. Furthermore. similar to previous research. this study found that the F 083 that are
close have FOBs that are happier. Finally this study is the first empirical attempt to
understand how values affect F OBs. This study found that values are a complex concept
and interact with other variables such as the owning family's level of closeness and the
boundary strength within the FOB. In summary. the findings from this study suggest that
F OBs should have close owning families and work towards a boundary that is permeable.

One of the greatest contributions of this study is the methodology used. This
study shows that the typical one rater methodology used in the majority of FOB research
has severe limitations. The multi-rater sampling along with the inclusion of statistical
methodologies suitable for interdependent systems used in this study provided a great

221

 

depth of information. Future researchers can learn from this process and develop similar
methods which will either challenge or strengthen many ofthe previous findings within
the existing research.

In conclusion, FOBs are a foundation to the US. economy. They also influence
many individuals as the majority of workers in the world are employed by FOBs.
Understanding how these systems function, as well as understanding how to strengthen

them will have a global effect.

222

 

APPENDICES

223

APPENDIX A: Gate Keeper Interview Guide
Questions for family business owners
To be administered verbally
Name
Company name
Year business was founded
Primary Industry

Gross Profit for 2006 2007
Projection for 2008

 

 

 

Number of employees working for the business

Name of family members employed in business full time

Name of family members employed in business part time

Construct three generation genogram of family and include their relationship to
business

10. List names and contacts to employees.

224

 

APPENDIX B: FACES III

'1'

 

 

 

 

 

 

 

 

 

 

 

225

_ Items for FACES lll
Cohesion Items (1 = .77
X = 39.8
SD = 5.4
Emotional Bonding Factor
Loading
1
2 I. F amil_ 1' members feel very close .60
to each other
3.3. Family togetherness is very .47
important
Supportiveness
23. Family members ask each other .51
“for help
24. Family members consult other .48
family members on their
decisions
Family Boundaries
25. Family members/eel closer to .49
other/amily members than to
people outside the family
.36. We like to do things with/us! our .39
immediatefamilI'
Time and Friends
27 Family members like to spend .69
free time with each other
28. We approve ofeach other 's .43
friends
Interests and Recreation
29. When our family gets together for .54
activities, everybody is present
30. We can easily think of'things to .43
do together as afamiQ¥__ “fl __ _
Adaptability a = .62
X = 24.1
'___________ so = 4.7
Leadership Factor
Loading
2
3 l . Different people act as leaders in our .35
familt‘
32. It is hard to identify the leader(s) in our .38
family
Control
33. The children make the c ecisions in our .34
family
3 4. In solving problems, the children 's .37
suggestions are/allowed
Discipline
35. Children have a say in their discipline .48

 

 

36.

37.

3 (‘1’.

39.

4 0.

Children and parents discuss
punishment together

Roles and Rules

Our/amily changes its ' way of
handling tasks

We shifl household l‘cspoiIsibilllic
from person to person

Its hard to tell who does which
household chores

Rules change in our_finnilv

S

.37

.45

.38

.34

.36

 

226

 

 

10.

11.

12.

13.

14.

15.

16.

17.

18.

APPENDIX C: Family Member Survey

Please rate the following items using the scale below. Please rate your

experience of your current family

1 2 3
Strongly Neutral
Disagree

Family members feel very close to each other
Family togetherness is very important
Family members ask each other for help

Family members consult other family members on
their decisions

Family members feel closer to other family members
than to people outside the family

We like to do things with just our immediate family

Family members like to spend free time with each
other

We approve of each other's friends

When our family gets together for activities,
everybody is present

We can easily think of things to do together as a
family

Different people act as leaders in our family
It is hard to identify the leader(s) in our family
The children make the decisions in our family

In solving problems, the children's suggestions are
followed

Children have a say in their discipline
Children and parents discuss punishment together
Our family changes its’ way of handling tasks

shift household responsibilities from person to person

227

5
Agree
Strongly

 

 

19. Its hard to tell who does which household chores

20. Rules change in our family

228

APPENDIX D: Participant Survey

Please answer the following questions thinking about your family and [INSERT
COMPANY NAME]

1. Your Name

 

2. Age

 

3. Circle one Male Female

4. Circle all that apply in regards to your relationship to [INSERT COMPANY NAME]
a- Owner
b. Employee
c. Manager
d. Family member of owning family

e. Board of directors member

f. Other

5. For the following questions, please assign a score, which positions [INSERT COMPANY NAME]
between the paired statements. (Select one for each pair of statements)

A manager’s qualifications (education, 1 2 3 4 5 6 7 Family members are given
experience. etc.) are the only preference in hiring and promotion
characteristics considered in hiring and decisions.

promotion decisions.

All employees are compensated Family members are paid more than
(excepting dividends) based solely on 1 2 3 4 5 6 7 non-family members in comparable
their position and performance. positions.

This company is a business. which This is a family, which happens to
happens to employ people from the 1 2 3 4 5 6 7 be in business together.

same family.

The owner(s) primarily get financial and The owner(s) primarily get
professional satisfaction from this satisfaction from working with family
business: working with family is a bonus. 1 2 3 4 5 6 7 members; the financial rewards from

the firm are a bonus.

1 Please rate the following items using the scale below

 

1 2 3 4 5 6 7 8 9 10
Very Somewhat Very
Dissatisfied Satisfied Satisfied

8. Your level of satisfaction with your involvement with the business [ J

229

9.

10.

11.

12.

13.

14.

Your level of satisfaction with the ownership/management of the
business

 

Your level of satisfaction with the employees within the business

_—_____ .._._ z_._‘

Your level of satisfaction with members of the owning family

 

Your level of satisfaction with the amount of conflict throughout the
business

 

Your level of satisfaction with the future direction of the business

 

Your level of satisfaction with how problems are solved within the
business

 

 

For the following questions you will be asked to identify individuals associated with
[INSERT COMPANY NAME]. You may list up to five names. If you cannot think of a
person who fits one or more of the items below please leave the item blank. Please
also identify your relationship to the individual you identified (e.g. mother, father,
owner, manager, co-worker)

In the last three week who have you had a meaningful conversation with regarding
[INSERT OWNING FAMILY NAME] family, or issues specifically related to the [INSERT
OWNIGN FAMILY NAME]?

 

 

 

 

 

 

 

 

Name Relation to you
Name Relation to you
Name Relation to you
Name Relation to you
Name Relation to you

 

 

In the last three week who have you had a meaningful conversation with regarding the
day to day functions of the business (e.g. job responsibilities, problems with coworkers,
production changes. time off)

 

 

 

 

 

 

 

 

Name Relation to you
Name Relation to you
Name Relation to you
Name Relation to you
Name Relation to you

 

 

In the last three week who have you had a meaningful conversation with regarding the
overall strategy and future of the business (e.g. strategic planning, succession planning,
initiating or changing governance boards)

Name Relation to you

 

230

Name

Name

 

 

Name

 

Name

 

Relation to you
Relation to you
Relation to you

Relation to you

231

 

 

 

 

 

APPENDIX E: Informed Consent

Exploration of Families in Family Owned Businesses

CONSENT TO ACT AS A HUMAN RESEARCH SUBJECT

RESEARCH TEAM

Lead Researcher:

Brian Distelberg Michigan State University Intern
Family and Child Ecology Department

(616) 481 -3524

distelbe@msu.edu

Faculty Sponsor:

Adrian Blow Ph.D

Family and Child Ecology

(517) 432-7092

38 Human Ecology, East Lansing, MI 48824,
blowa@msu.edu

PURPOSE OF STUDY

The purpose of this study is to explore the interaction between families and family owned
businesses. This research will explore the influence of the owning family on the family
business and vise versa. You are being asked to participate in a research study of family
owned businesses. You have been selected to participate in this study because of your
relationship to a family owned business through either employment in a family owned
business or blood or legal relationship to the owning family of a family business.

In the entire study, you will be asked to complete a short (10-15 minute) survey which
focuses in on your experience with a family owned business. Specifically you will be
asked about your level of satisfaction with your family business and other specific
questions about the family business.

If you are under 18 you cannot be in the study.

WHAT YOU WILL DO

There are two separate phases to this study. First the researchers will conduct an
interview with the identified owner of the family business. Then the researchers with the
permission of the owner will contact all employees and family members of the family
business. The following outlines these two phases.

If you are the Owner or an indentified key person to the business

Prior to collecting information from the family members or the employees of a family
owned business, the researcher will conduct a short interview with the identified owner
of the business. In this interview, you will be asked to allow access to employees and
family members and to collaborate with the researcher in obtaining demographic

232

information (e.g. number of employees, industry of operation, 2006, 2007 and 2008
revenue) as well as help construct a list of employees and family members who are
eligible to participate in the following two phases. Additionally, in businesses where
employees computers are subject to company supervision, or oversight, you will agree
to not access individual employee or ownership survey responses.

If you are a family or business member

You will participate in a short survey (10-15 minutes) by a means of your choosing
(internet, telephone, or pen and paper). The survey will ask about your experiences with
working in the business. Family members will be asked to complete a similar survey, but
also to complete a survey asking for their experience with being a family member of the
owning family. This survey is somewhat longer and should take no more than 15 minute
to complete.

RISKS AND DISCOMFORTS

This study involves no more than minimal risk. There are no known harms or discomforts
associated with this study beyond those encountered in normal daily life. The researcher
will also make every effort to respect you right to privacy and when results of the study
are made public all indentifying information will be removed which could indentify the
individual and the family owned business. For individuals using an internet based
survey, you should be aware that in some businesses other individuals within your
business may have access to your survey responses. The owner of your business has
agreed to not access your survey responses for the purposes of this study. But you
should be aware of the potential for others to access your information if you use a
company owned computer to take the survey through the internet. If you are not willing
to take the internet survey, you may take a pen and paper survey, or a telephone survey.

POTENTIAL BENEFITS

The benefits of participation include the knowledge gained from the three assessments,
taking part in study will educate practitioners and service provider of family businesses,
and other family businesses. Knowledge gained from the three assessments will be
presented to each organization and when possible suggestions based on the
assessments will be given to the business. These three assessments include:

1. The communication map illustrates how information flows through the system. Often
times there are ineffective communication blocks, and more time than not there is a key
person that all or most communications flow through. Interesting to this study is that it is
rarely the CEO/President.

2. We also look at value orientation. In other words is the family business a “family
business" with a big “F” or big “".B This assessment has been scaled through the
standardization of a national sample (2007 American Family Business Survey). What we
found is that this value orientation is a continuum. And where the business falls on that
continuum has implications for desired future goals. For example Family businesses like
to keep resources in the family and prefer to use succession strategies that promote
equal (not necessarily equitable) sale of the business to the next generation. Conversely
family Businesses, prefer to keep resources in the business. (Pay family less and prefer
to sell the business outside the family). While this assessment is interesting and gives a
business an opportunity to examine their value orientation and associated resource
transfers and future goals, this does not predict success in future goals. What does
predict success is how aligned everyone in the business is with the value orientation.

233

That is why we have the majority of individuals in the business report their perception of
the value orientation. For example a family Business may want to sell the business
outside of the family. This is successful when key individuals are aligned, but extremely
difficult when only the CEO holds this value and the rest of the system sees the business
as a Family business. So this along with the communication map provides a lot of
information that can be used to build strategies and align individuals with a common
vision and value orientation. Basically avoid a lot of frustration and failure in strategic
planning.

1.

We also administer an assessment for family dynamics within the family system.
This is a well known and thoroughly tested assessment (FACES IV). The purpose
is to look at the how family systems with different dynamics employ different
communication patterns. For example others have stated that varying family
dynamics employ varying level of boundaries between the family system and the
business system. The hypothesis stated in the literature points out that certain
typologies are better than others. This is new, and we are unsure of the direct
benefit, that is why we are doing the study. We do believe there will be important
information gained for the business but don’t feel comfortable stating what that is
yet, because this is the first study to look at this issue in depth. For the family
system there is benefit. Many Marriage and Family Therapists use this assessment.
To do this assessment for a family in therapy would cost the family upwards of
$1,000. There is a plethora of information available from this assessment for the
family.

After collecting the data the research will come back to the business and discuss
their results. We will collaborate with each business to find the best medium for
disseminating results.

Finally, we offer a lottery system for every business. Right now we have funding to
have one $50 gift card for every business (which is given out through a lottery).

The actual process is:

1. First meeting (over phone or in person with a key individual. We collect
some demographic information about the business (year founded, revenue
for three years, number of employees)

2. Discuss the most effective way to administer the two surveys (above). We
are looking for an 80% response rate or better. This includes employees,
owners and family members (may be employed or not employed by the
business). In many cases email surveys have worked, but we have options
for paper and telephone surveys in cases where email and internet are not
effective.

3. Discuss any additional information that might be valuable to collect at this
time.

4. Administer the surveys

5. Discuss results with key individuals in the business

ALTERNATIVES TO PARTICIPATION

The only alternative to participation in this study is not to participate. You are invited to
participate in two phases of this research, but you may choose to participate in one phase
or not at all. You are also free to terminate your participation at anytime.

There is potential to modify the procedures and surveys when certain aspects of the
process interfere with business Operation or individual confidentiality.

234

 

COMPENSATION, COSTS AND REIMBURSEMENT

Your participation is strictly voluntary and you will not be paid for your participation in this
research study. All participants connected to your business will be eligible to receive a $50
gift card determined by a random drawing of names of participants. There are no known
costs to you for participation in this study.

CONFIDENTIALITY

Your confidentiality will be protected to the maximum extent allowable by law. All
identifiable information that will be collected about you will be removed at the end of data
collection. All other information will be stored and only the researchers will have access to
this data. All research data will be maintained in a secure location. Only the researchers will
be allowed access to it. All research data that is stored on a laptop computer is password
protected and stored in a locked facility. The research team, (Brian Distelberg and Adrian
Blow), are the only individuals with access to your study records to protect your safety and
welfare. Any information derived from this research project that personally identifies you will
not be voluntarily released or disclosed by these entities without your separate consent,
except as specifically required by law. Publications and/or presentations that result from this
study will not include identifiable information about you. The researchers will keep the
research data for 7 years.

YOUR RIGHTS TO PARTICIPATE, SAY NO, OR WITHDRAW

Your participation in this research project is completely voluntary. You have the right to say
no. You may also change your mind or withdraw from the study at any time during the
course of the study. You also have the right to choose not to answer specific questions or to
stop participating at any time.

IF YOU HAVE QUESTIONS
If you have any comments, concerns, or questions regarding the conduct of this research
please contact Brian Distelberg at (616) 481-3524 or email: distelbe@msu.edu

If you have questions or concerns about your role and rights as a research participant,
would like to obtain information or offer input, or would like to register a complaint about this
study, you may contact, anonymously if you wish, the Michigan State University’s Human
Research Protection Program at 517-355-2180, Fax 517-432-4503, or e-mail irb@msu.edu
or regular mail at 202 Olds Hall, MSU, East Lansing, MI 48824.

 

235

 

VOLUNTARY PARTICIPATION STATEMENT

You should not sign this form unless you have been given a copy of this document for your
records.

Participation in this study is voluntary. You may refuse to answer any question or discontinue
your involvement at any time without penalty or loss of benefits to which you might otherwise be
entitled. Your decision will not affect your future relationship with the Family Business Alliance or
your Employer. Your signature below indicates that you have been given a copy of the
information in this consent form, have had a chance to ask any questions about the study, and
agree to participate.

I agree to participate in the study

 

Subject Signature Date

 

Printed Name of Subject

 

Researcher Signature Date

 

Printed Name of Researcher

236

APPENDIX F: Additional Sociograms

Figure 6.1 Company 1: Employee Communication

31016

 

 

 

- Employee
Not in business as employee or owner

0 Family

[:1 Not family

 

 

 

 

237

 

Figure 6.2: Company 1: Ownership Communication

M1006
/

3
i
l

.1013

 

- Employee
Not in business as employee or owner

0 Family

[:I Not family

 

 

 

 

238

Figure 6.4: Company 2: Owner Communication

 

[2010 F2004
\\'| [/1
r/
‘I\\ /i
/
.\ /
\ y /2003
\ "/1 11//
.00
i , ’/ \‘\
,4/ \\\_
y // \.\
1% ' \\\‘
i
i / \\
i. "\2005
[meow—H2002 \\
i' \
. \\~.
i \
l \
g \..,,
V2006
{f
l
"l
izoos

 

- Employee
Not in business as employee or owner

0 Family

[I Not family

 

 

 

 

240

Figure 6.5: Company 3: Employee Communication

’3198
$3204 /

[3196 /.l3197

3205 \ . I3210

N32 13 /|3234
\ /
\‘3206

1 3209
\ f

[3173 / 3172

 

 

/

[3119 43102 ‘3142

 

- Employee
— Not in business as employee or owner
Family

D Not family

 

 

 

241

Figure 6.6: Company 3: Owner Communication

|I31‘l8

Q3131 l T3162

[ ! p3231

‘3130
.\

pszoe

i320?

3226

    

l3119
1-3112

25 1 . “
31 13120 l
1 03219

 

- Employee
:-- Not in business as employee or owner
0 Family

D Not family

 

 

 

 

242

Figure 6.7: Company 4: Employee Communication

.4129 [’4195

 

  
 

    

 

 

I417? 14115
\1. ‘0.
I4122 - l. 4
I413 , 4172 i 131
\x? 4134 4141
I4180——~ 4130 4157 1 \‘\>‘4127 /
fiflwa—ans 0 .
, ‘1 4171—P--—**’I4149
[4146
4176
4179 11 \
J i083 \\ 0. 4178 ‘4145
l 7’
r 1’ "\
/ , 0 4170
l4113 l ” i. “'0 ,7".
,’ 4104 44196 mg; V4191 /
.0/ [4175
34190
- Employee
0:“:- Not in business as employee or owner
0 Family
D Not family

 

 

 

243

 

Figure 6.8: Company 4: Owner Communication

04191
i

‘4131
\

|4115

 

 

 

1’ iii“
[4128 K1171 ,1’
yr \‘4141
i4139

- Employee
"ii"? Not in business as employee or owner
0 Family
U Not family

 

 

 

244

Figure 6.9: Company 5: Employee Communication

Il ’5117

5119

[15128

 

1&5107

 

- Employee
Not in business as employee or owner
0 Family

[I Not family

 

 

 

245

Figure 6.10: Company 5: Owner Communication

'5121

5135

\105 5106

[5123 / \ 1‘

 

[5108
5115

5113

 

- Employee

- Not in business as employee or owner
0 Family

U Notfamily

 

 

 

246

Figure 6.11: Company 6: Employee Communication

‘, 2
V

  

/
7}
9 a4
u ’
($045“
439.7741
4‘.

i 1
V!
‘ ‘fl
.- \’:.4‘0§‘:: -7"
\ \ ..
4..
\
.an.

-‘. A

ill

l
l.
i

Ill
1

1

All

‘4

is.» .
v7, ~ ‘1
ill

7%
. l”

, .1171
4%

h
v '
,4
*4

ill

0

Mr.“ . A.
#7355“; '1 \ {“‘I‘iIa>"<*Q \ '7‘
4h...» — " 7...:
/ [Elfilg'fiba‘
'1

1"!

'14
l

J
g
3’4
i-
534; .
A“
S
A
1
/

 

- Employee

- Not in business as employee or owner
Family

I] Notfamily

 

 

 

247

Figure 6.12: Company 6: Owner Communication

6521

6150

 

6528 A 0247
I6512
74141 .5462 " 6192 253130
‘6404
117 r .5459
l6538 6537
I6525 ‘
6302 [6285 6527
i 6255
6242
'55" 6178
- Employee
- Not in business as employee or owner
0 Family
[I Not family

 

 

 

248

 

62m

6229

\.,

6143

 

Figure 6.13: Company 7: Employee Communication

7121

 

.7123
7118

7113

7119

 

 

 

- Employee

- Not in business as employee or owner
0 Family

['_'| Not family

 

 

 

249

Figure 6.14: Company 7: Owner Communication

 

 

7113
7104
7120
‘4 .
.7106 2 7102 i 7105 ‘ . 7112
7103
i 7111
7109 \‘7110
/'7116
7108
- Employee
- Not in business as employee or owner
0 Family
D Notfamily

 

 

 

250

 

Figure 6.15: Company 8: Employee Communication

 

121
I8129 :‘8120 8102

l8131
514 ' /.....
4‘ i ‘r' 8105
1 \‘18119
I8133 ———I8115 31°“ “3
8104
0110
8114
8109
8106
8107

 

l8116
- Employee
- Not in business as employee or owner
0 Family
D Not family

 

 

 

251

 

Figure 6.16: Company 8: Owner Communication

 

 

 

8129
8121
.8124
I! 8120
x 8108
‘1
/' 8101
8109
102 8119
- Employee
- Not in business as employee or owner
0 Family
[3 Notfamily

 

 

 

252

 

Figure 6.17: Company 9: Employee Communication

 

 

M3
9007
[9006
' ’.' 9101
09005 9002
- Employee
- Not in business as employee or owner
Family
D Not family

 

 

 

253

 

Figure 6.18: Company 9: Owner Communication

.0... w

 

9002
9003
09005
- Employee
- Not in business as employee or owner
0 Family
D Not family

 

 

 

254

Figure 6.19: Company 10: Employee Communication

010008
010009
[110010

 

‘10007

 

 

1 10004
1 10005
10003
- Employee
- Not in business as employee or owner
0 Family
D Not family

 

 

 

255

 

 

Figure 6.20: Company 10: Owner Communication

 

 

 

010108
010009
[110010
010007
; 10006
1005
- Employee
- Not in business as employee or owner
0 Family
[:1 Not family

 

 

 

256

Figure 6.21: Company 11: Employee Communication

‘9"“” u0u
<>uaw

011007

11002

 

3 11009
11004
[11005
- Employee
- Not in business as employee or owner
Family
D Not family

 

 

 

257

 

Figure 6.22: Company 11: Owner Communication

 

 

 

 

I11004 11002
I11w5
011017
011008
011010
11009
11006
011011
- Employee
- Not in business as employee or owner
0 Family
[I Notfamily
258

REFERENCES

259

 

 

References

Ackoff, R. L. (1977). Towards flexible organizations: A multidimensional design.
OMEGA. 5, 649-662.

Aiken, M.. & Hage. .l. (1971). The organic organization and innovation. Sociology, 5. ()3-
93

Amarapurkar, S.S., Danes, SM. (2005). Farm business-owning couples:
Interrelationships among business tensions, relationship conflict quality. and
spousal satisfaction. Journal of Family and Economic Issues, 26, 419-441

Anderson, R.C., Mansi, S.A & Reeb, D.M. (2003) Founding family ownership and the
agency cost of debt. Journal of Financial Economics, 68. 263-285.

Anderson, R.C., Reeb, D.M. (2003). Founding-family ownership and firm performance:
Evidence from the S&P 500. Journal of Finance. 58, 1301-1328.

Anderson, S.A. & Gazazzi, SM. (1990). A test of Olson Circumplex Model: Examing its
curvilinear assumption and the presence of extreme family types. Family Process
29, 309-324.

Anderson, T., Carlson. J., & Getz, D. (2002). Family business goals in the tourism and
hospitality sector: Case studies and cross-case analysis from Australia, Canada.
and Sweden. Family Business Review, 15, 89-108.

Amerikaner, M. Monks, G. Wolfe, P. & Thomas, S. (1994). Family interactions and
individual psychological health. Journal ofCounseling and Development. 72.
614-620

Arthur Andersen/Mass Mutual (1997). Arthur Andersen/Mass Mutual American Family
Business Survey ‘97. Arthur Andersen Center for Family Business and Mass
Mutual: The Blue Chip Company.

Aronoff, C.: 2004, ‘Self-Perpetuation Family Organization Built on Values: Necessary
Condition for Long-Term Family Business Survival’, Family Business Review

17, 55—59.

Aronoff, C. E., Ward, J.L. & Astrachan, J.H. (2002). Family Business Source/wok. 3rd ed.
Family Enterprise Publishers

Astrachan, J .H.(2003).Commentary on the special issue: The emergence of a field.
Journal ofBusiness lr'enluring. 18(5), 567- 572.

260

 

Astrachan, J.H., Allen, I.E., Spinelli, S., & Whittmeyer, CB. (2003). Americanfamily
business survey. Boston, MA: MassMutual Financial Group/Raymond Institute
Survey.

Astrachan, .l. H., Klein, S. B., & Smyrnios, K. X. (2002). The F-PEC Scale of Family
Influence: A proposal for solving the family business definition problem. Family
Business Review, 15(1), 45-58.

Astrachan, J. H., & Shanker, M. C. (2003). Family businesses‘ contribution to the US.
economy: A closer look. Family Business Review, 16(3), 21 1—219.

Bahrami, H. (1992). The emerging flexible organization: Perspectives from Silicon
Valley. California Management Review, 34(4), pp. 33.

Beehr, T., Drexler, J.A., & Faulkner, S. (1997). Working in small family businesses:
Empirical comparisons to non-family businesses. Journal of Organizational
Behavior, 18(3), 297-312.

Bertalanffy. L. von. (1969). General System Theory. New York: George Braziller

Blanco-Mazagato, V., de Quevedo-Puente, E., Castrillo, LA. (2007). The trade-off
between financial resources and agency costs in the family business: An
exploratory study. Family Business Review, 20(3), 199-213.

Borgatti, SP. (2005). NetDraw: Graph visualization software. Harvard: Analytic
Technologies

Borgatti, S.P., Everett M.G., & Freeman. LC. (2002). UCINET 6 for Windows.
Harvard: Analytic Technologies.

Bork. D., Jaffe, D. T.. Lane. S. H., Dashew, L., & Heisler, Q. G. (1996). Working with
family businesses. San Francisco: Jossey-Bass.

Boss, P. (1987). Family stress. In M.B. Sussman & S.K. Steinmetz (Eds). Handbook of
marriage andfamily (pp. 695-723). New York: Plenum.

Bronfenbrenner, U. (1979). The Ecology ofHuman Development: Brim/intents by nature
and design. Cambridge, MA: Harvard University Press.

Bubolz, M.M. & Sontag, MS. (1993). Human ecology theory. In P.G. Boss, WJ.
Doherty, R. LaRossa. W.R. Schumm, & S.K. Steinmetz (Eds). Sourcebook of
family theories and methods.“ A contextual approach (pp. 419-448). New York:
Plenum Press.

261

Burke, B]. (2007). Working for the family: A study of family and nonfamily manager
subgroups in family business management teams. Dissertation. Rutgers
University

Carter, 8., & McGoIdrick, M. (Ed.). (1998). The Expanded Family Life Cycle:
Individual, Family and Social Perspectives , 3rd Edition. Boston: Allyn & Bacon.

Castillo, J., & Wakefield, M.W. (2007). An exploration of firm performance factors in
family businesses: Do families value only the “bottom line”? Journal of Small
Business Strategy, 17(2), 37-51.

Chrisman, J .J ., Chua, J.H., & Litz, R. (2003). A unified systems perspective of family
firm performance: An extension and integration. Journal of Business Venturing,
18(4). 467-472.

Chrisman, J. J ., Chua, .I. H., & Sharma, P. (1998). Important attributes of successors in
family businesses: An exploratory study. Family Business Review, 21, 35-47.

Cole, PM. (2000). Understanding family business relationships: Preserving the family in
the business. The Family Journal, 8, 351-361.

Coleman, 8., & Carsky, M. (1999). Sources of capital for small family-owned businesses:
Evidence from the National Survey of Small Business Finances. Family Business
Review, 12(1), 73—85.

Corwin, R. G. (1972). Strategies for organizational innovation: An empirical comparison.
American Sociological Review. 3 7, 441-454.

Cronbach, L. .l. (1951). Coefficient alpha and the internal structure of tests.
Psychometrika, 16(3), pp.297-334.

Daily, C.M., & Dollimger, MJ. (1992). An empirical examination of ownership structure
in family and professionally managed firms. Family Business Review, 5(2). 1 17—

136.

Damanpour, F. (1991). Organizational innovation: A meta-analysis of effects of
determinants and moderators. Academy of Management Journal, 3 4(3), 555-590.

Danes, S. M., & McTavish, D. (1997). Role involvement of farm women. Journal of
Family and Economic Issues 18(1), 69—89.

Danes, S. M., Rueter, M. A., Kwon, H. K.. & Doherty, W. D. (2002). Family FIRO
model: An application to family business. Family Business Review 15(1), 31-43.

262

Danes, S.M., Zuiker, V.S., Kean, R., & Arbuthnot, J. (1999). Predictors of family
business tensions and goal achievement. Family Business Review, 12, 241-252.

Davis. R. & Stern. D. (1981). Adaptation. survival and growth ofthe family business: An
integrated systems perspective. Human Relations, 34 (3), 207-224.

Davis, R, & Stern, D. (1996). Adaptation, survival, and growth of the family business:
An integrated systems perspective. In C. E. Aronoff, J. H. Astrachan & J. L. Ward
(Eds), Family business sourcebook II (pp. 278—309). Marietta, Georgia: Business
Owner Resources.

Dayley, J.G., Sowers-Hoag, K.M. & Thyer, BA. (1991). Construct validity ofthe
Circumplex Model of family functioning. Journal of Social Service Research, 15.
131-147.

Dean, SM. (1992). Characteristics of African American family-owned businesses in Los
Angeles. Family Business Review, 5(4), 373-395.

Denison, D.. L. Colleen and J. Ward: 2004, ‘Culture in Family-Owned Enterprises:
Recognizing and Leveraging Unique Strengths’, Family Business Review 1 7. 6] —
70.

Dess, G.G., & Robinson, R.B. Jr. (1984). Measuring organizational performance in the

absence of objective measures: The case of the privately-held firm and
conglomerate business unit. Strategic Management Journal. 5. 265-273.

Distelberg, B. & Sorenson, R. (2009). A systemic examination of family businesses: A
focus on values, resource flows and adaptation. Family Business Review,22(1),
65-81.

Distelberg, B. (2008, Working paper). The role of values and cohesion in family
businesses.

Donckel, R., & Frohlick. E. (1991). Are family businesses really different? European
experiences from STRATOS. Family Business Review, 4. 149-160.

Dyer Jr., W.G. (2006). Examining the “family effect” on firm performance. Family
Business Review, 19(4), 253-273.

Dyer, Jr., W. G. (1986). Cultural Change in Family Firms: Anticipating and Managing
Business and Family Transitions. San Francisco: Jossey-Bass

Eppink, J. (1978). Planning for strategic flexibility. Long Range Planning. 11, 9-15.

263

Farrell, M.P. & Barnes, GM. (1993). Family systems and social support: A test of the
effects of cohesion and adaptability on the functioning of parents and adolescents.

Journal of Marriage and the Family, 55, 1 19-132

Feltham, T.S., Feltham, G., & Barnett, J .J . (2005). The dependence of family businesses
on a single decision-making. Journal of Small Business Management, 43, 1-15.

Fetch, R.J. & Zimmerman, T.S. ( 1999). Marriage and family consultation with ranch and
farm owning families: An empirical family case study. Journal ofMarital and
Family Therapy, 25(4). 485-501.

Fleming, Q..I. (2000). Keep the/amily baggage out of the Family Business: Avoid the
seven deadly sins that destroy the family business. Simon & Schuster Adult
Publishing Group.

Fristad. M. (1989). A comparison of the MnMaster and the Circumplex family
assessment instruments. Journal of Marital and Family Therapy, 15, 259-269

Gallo, G.A., Tapies. J.. & Cappuyns, K. (2000). Comparison of family and non-family
business: Financial logic and personal preference. “Chair of Family Business“
IESE Research Paper No. 406 BIS. University of Navarra.

Galvin, B., Astrachan, J .. & Green, J. (2007). A mericanfamily business survey. Boston.
MA: MassMutual Financial Group.

Gersick. K. E.; Davis. J. A.; Hampton. M.; Lansberg, I. (1997). Generation to
generation: Life cycles ofthe/amily business. Boston, MA: Harvard Business
School Press.

Gomez-Mejia, L.R.,Nufiez-Nickel, M.,& Gutierrez, I. (2002). The role of family ties in
agency contracts. Academy of Management Journal, 44(1), 81—95.

Green. R.G., Harris, R.N.. Forte. J.A. & Robinson. M. (1991). Evaluating FACES III and
the Circumplex Model: 2,440 families. Family Process, 30, 55-73.

Habbershon, T. G., & Astrachan, J. H. (1997). Perceptions are reality: How family
meetings lead to collective action. Family Business Review, 10(1), 37-52.

Habbershon, T.G., & Williams, ML. (1999). A resource-based framework for assessing
the strategic advantages of family firms. Family Business Review, 12(1). 1-25.

Hamilton, E. (2006). Whose story is it anyways?: Narrative accounts of the role of

women in founding and establishing family businesses. International Small
Business Journal, 24, 253-271.

264

Hampson, R.B., Hulgus, Y.F. & Beavers, R.W. (1991). Comparisons of self-report
measures of the Beavers Systems Model and Olson’s Circumplex Model. Journal
ofFamily Psychology. 4, 326-340.

Hanneman, R. A. and Riddle, M. (2005). Introduction to social network methods.
Riverside, CA: University of California, Riverside ( published in digital form at
http:I/faculty.ucr.edu/~hanneman/ )

Hatum, A., & Pettigrew, A. (2004). Adaptation under environmental turmoil:
Organizational flexibility in family-owned firms. Family Business ReVlL’H‘, 17(3).
23 7-259.

Haynes, G.W., Onochie, J .I., & Muske, G. (2007). Is what’s good for the business, good
for the family: A financial assessment. Journal of Family and Economic Issues,
28, 395-410.

Haynes, G.W, Walker, R., Rowe, B.R., & Hong, G. (1999). The intermingling of
business and family finanaces in family-owned businesses. Family Business

Review, 12(3), 225-239

Hienerth, C ., & Kessler, A. (2006). Measuring success in family businesses: The concept
of configuraltional fit. Family Business Review, 19, 1 15-133.

Jorissen, A., Laveren, E., Martens, R., Reheul, A. (2005), "Real versus sample-based
differences in comparative family business research", Family Business Review.

Vol. 18 No.3, pp.229-46.

Kaye, K. (1991). Penetrating the cycle of sustained conflict. Family Business Review,
4(1), 21-44.

Kalleberg, A.L., & Leicht, K.T. (1991). Gender and organizational performance:
Determinates of small business survival and success. Academy o/‘Management

Journal, 34(1), 136-161.

Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006). Dyadic Data Analysis. Guilford
Press: NY

Kerr, M., & Bowen, M. (1988). Family evaluation. New York: Norton.
Kreft, I. & de Leeuw. J. (1998): Introducing Multilevel Modeling. London: Sage.

Krijnen. H. C. (1979). The flexible firm. Long Range Planning, 12, 63-75.

265

Lansberg, 1., & Astrachan, J. H. (1994). Influence of family relationships on succession
planning and training: The importance of mediating factors. Family Business
Review, 7(1), 39-59.

Lee, J. (2006). Impact of family relationships on attitudes of the second generation in
family business, Family Business Review, 19(3), 175—191.

Lee, Y.G., Danes, S.M. & Shelley, MC. (2006). Work roles, management and perceived
well-being for married women within family businesses. Journal of Family and
Economic issues, 27, 523-541.

Lee, Y.G., Rowe, B.R. & Hong, GS. (2006). Third shift women in business-owning
families. Journal of Family and Economic Issues, 27, 72-91.

Levinson, H. (1971, March/April). Conflicts that plague family business. Harvard
Review, pp. 90-98.

Lussier, R.N., Sonfield, MC. (2006). The effect of family business size as firms grow: a
USA-France comparison. Journal of Small Business and Enterprise Development,
13, 314-325.

Marsden, P. (2005). Recent developments in social network measurement. In Models and
Methods in Social Network Analysis, P.J. Carrington, J. Scott, & S. Wasserman.
Cambridge Press NY

Masuo, D., Fong. G., Yanagida, J.. & Cabel, C. (2001). Factors associated with business
and family success: A comparison of single manager and dual manager family
business households. Journal of Family and Economic Issues, 22, 55-73

McConaughy, D.L. (2000). Family CEOs vs. nonfamily CEOs in the family-controlled
firm: An examination of the level and sensitivity of pay to performance. Family
Business Review, 18, 121-132.

Miner, J .B. (1997). The expanded horizon for achieving entrepreneurial success.
Organizational Dynamics. 25(3), 54-67.

Minuchin, S. (1974). Families andfamily therapy. Cambridge, Massachusetts: Harvard
University Press.

Nichols, M.P. & Schwartz, RC. (2004). Family Therapy Concepts and Methods. 6‘h ed.
New York: Pearson.

Olson, DH. (2000). Circumplex model of marital and family systems. Journal ofFamily
Therapy, 22(2), 144-167.

266

Olson. DH. (1994). Commentary: Curvilinearity survives: The world is not flat. Family
Process, 3 3, 471 -478.

Olson, DH. (1985). FACES III (Family Adaptation and Cohesion Scales). St. Paul. MN:
University of Minnesota.

Olson, D.H., Sprenkle, D.H., & Russell, CS. (197%). Circumplex model of marital and
family systems: I. Cohesion and adaptability dimensions, family types, and
clinical applications. Family Process, 18, 3-28.

Olson, D.H., Sprenkle, D.H., & Russell, C.S. (1979b). Circumplex model of marital and
family systems: IV. Theoretical update. Family Process, 22, 69-83.

Olson, P. D., Zuiker, V. S... Danes, S. M., Stafford, K., Heck, R. K. Z., & Duncan. K. A.
(2003). The impact of the family and business on family business sustainability.
Journal of Business Venturing, 18(5), 639—666.

Overholt, M. H. (1997). Flexible organizations: Using organizational design as a
competitive advantage. Human Resources Planning, 20(1), 22-32.

Perosa, L.M., & Perosa. S.L. (1990). Convergent and discriminant validity for family
self-report measures. Educational and Psychological Measurement. 50, 855-868.

Pratt. D.M., & Hanson, JG (1987). A test of the curvilinearity hypothesis with FACES II
and 111. Journal of Marital and Family Therapy. 13. 387-392.

Schulze, W. S., Lubatkin, M. H., Dino, R. N., & Buchholtz, A. K. (2001). Agency
relationships in family firms: Theory and evidence. Organizational Science,
12(2), 99—116.

Sharma. P. (2004). An overview ofthe field of family business studies: Current status
and directions for the future. Family Business Review, 27, 1-36.

Sharma, R, Chrisman, J.J., & Chua. J.H. (1997). Strategic management ofthe family
business: Past research and future challenges. Family Business Review, 10(1). 1-
35.

Shanna, P. & Nordqvist, M. (2008). A classification scheme for family firms: From
family values to effective governance to firm performance. In Tapies, J. & Ward.
J .L (Ed.) Family values and value creation: How dofamily-owned businesses
foster enduring values. Palgrave Macmillan Publishers. pp.71-101.

Sharma, P. & Rao, AS. (2000). Successor attributes in Indian and Canadian family firms:
A comparative study. Family Business Review, 23, 313-330.

267

Sonfield, M.C. & Lussier, RN. (2005). Family business ownership and management: A
gender comparison. Journal of Small Business Strategy, I 5, 52-75

Sonfield, M.C., Lussier, R.N., Pfeifer, S., Manikutty, S., Maherault, L., & Verdier, L.
(2005). A cross-National investigation of first-generation. second-generation. and
third-generation family businesses: A four country anova comparison. Journal of
Small Business Strategy, 16, 9-26

Sonfield, M.C., & Lussier, RC. (2004). First second and third generation family firms: A
comparison. Family Business Review, I 7, 189-202.

Sorenson, R.L. (2000). The contribution of leadership style and practice to family and
business success. Family Business Review, 18, 183-200

Stafford, K., Duncan, K.A., Danes. S.M., Winter, M., 1999. A research model of
sustainable family business. Family Business Review 12, 197—208.

Stewart, C.C., Danes, SM. (2001). Inclusion and control in resort family businesses: A
developmental approach to conflict. Journal of Family and Economic Issues, 22,
293-320

Raudenbush, S.W., Bryk, AS, (2002). Hierarchical Linear Models: Applications and
Data Analysis Methods. Sage Publications, Thousand Oaks.

Raykov, T., Marcoulides, G. A. (2006). Afirst course in structural equation modeling
(Second Edition). Mahwah. NJ: Erlbaum.

Rutherford, M.W., Muse, L.A. & Oswald, S.L (2006). A new perspective on the
developmental model for family business. Family Business Review, 19(4). 317-
333.

Taguiri, R., & Davis, J.A., (1982). Bivalent attributes of the family firm. Working paper.
Harvard Business School, Cambridge Mass Reprinted 1996, Family Business
Review, 9(2): 199-208.

Teal. E. J., Upton, N., & Seaman, S. L. (2003). A comparative analysis of strategic
marketing practices of high-growth U.S. family and non-family firms. Journal of
Developmental Entrepreneurship, 8(2), 177—195

Thomas. V. & Ozechowski. T. J. (2000). A test of the C ircumplex model of marital and

family systems using the clinical rating scale. Journal o/‘Marital and Family
Therapy, 26(4), 523-534.

268

Tower. C. B., Gudmundson, D., Schierstedt. S., & Hartman. E. A. (2007) Do family
meetings really matter? Their relationship to planning and performance outcomes
in small family businesses. Journal omea/l Business Strategy, 18(1). 85-94.

US. Census Bureau (2007). North American Classification System (NAIC S). Retrieved
November, 2008 from http://www.census.gov/cgi-
bin/sssd/naics/naicsrch‘2chart=2007

Vayda. AP. (1983). Progressive contextualization: Methods for research in human
ecology. Human Ecology, 11(3), 265-281.

Vera. C .F., Dean. MA. (2005). An examination of the challenges daughters face in
family business succession. Family Business Review, 18, 321-345

Ward. J. L. (1987). Keeping the/amily business healthy: How to plan/or continuing
growth, profitability. andfamily leadership. San Francisco: Jossey-Bass.

Wasserman, S. & Faust. K. (1994). Social network analysis: 117/[ethods and
applications. Cambridge: Cambridge University Press.

Westhead, P., Cowling, M. (1997). Performance contrasts between family and
non-family unquoted companies in the UK. Internationsl Journal of
Entrepreneurial Behavior and Research, 3, 30-52.

Whitchurch, G.G. & Constantine.L.L. (1993). Systems theory. In P.G. Boss W.J.
Doherty, R. LaRossa, W.R. Schumm, & S.K. Steinmetz (Eds),
Sourcebook of family theories and methods: A contextual approach (pp.
325-355). New York: Plenum Press.

Wong. 8., McReynolds, S., & Wong. W. (1992). Chinese family firms in the San
Francisco Bay areas. Family Business Review, 5(4). 355-3 72.

Zahra, SA. (2005). Entrepreneurial risk taking in family firms. Family Business
Review, 18, 23-40.

Zody, Z, Sprenkle, D., MacDermid. S., & Schrank, H. (2006). Boundaries and the
functioning of family and business systems. Journal of Family and

Economic Issues, 2 7(2). 185-206.

Zuiker, V.S., 1998. Hispanic Self-Employment in the Southwest Rising Above
the Threshold of Poverty. Garland Publishing, New York.

269