Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Organizational mimicry in American social movement communities: an analysis of form communication effects on the evolution of crisis pregnancy centers, 1989-2009
(USC Thesis Other)
Organizational mimicry in American social movement communities: an analysis of form communication effects on the evolution of crisis pregnancy centers, 1989-2009
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
ORGANIZATIONAL MIMICRY IN AMERICAN SOCIAL MOVEMENT
COMMUNITIES: AN ANALYSIS OF FORM COMMUNICATION EFFECTS ON
THE EVOLUTION OF CRISIS PREGNANCY CENTERS, 1989-2009
by
Bettina Maria Richards Heiss
_____________________________________________________________________
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)
December 2010
Copyright 2010 Bettina Maria Richards Heiss
ii
DEDICATION
To my husband, who encouraged me to embark on my journey through graduate school,
and our tiny child, who carried me over the finish line.
iii
ACKNOWLEDGEMENTS
Many individuals supported the completion of this research project. I am
especially grateful for the assistance provided by my advisor Peter Monge, who
introduced me to the study of organizational evolution and has been fostering my interest
in the topic ever since. He is an extraordinary role model, both as a communication
scholar and a human being, and I am deeply thankful that I got to work with him
throughout my years in graduate school. I am also indebted to my committee members
Janet Fulk and Mark Kennedy for providing both inspiration and support along the way.
While my encounters with these and many other professors and graduate students at USC
greatly influenced my development as a researcher, I could not have survived graduate
school without the help of the knowledgeable and helpful Annenberg School staff.
Among others, Anne Marie Campian, Christine Lloreda, and Imre Meszaros have been
kind to me throughout the years and always made me feel welcome at Annenberg.
A special thanks goes to Erik Meusel, who donated the extensive programming
work needed to extract and process the data in order to prepare them for analysis. Ingmar
Rapp gave patient advice on all my questions about statistical methods and Florestan
Ballstaedt produced the wonderful geographical maps. Thomas Kern was most
supportive in speeding up the process by giving me time off for completing the draft. In
addition, I want to thank Larry Finer and Suzette Audam from the Alan Guttmacher
Institute as well as Emily Gardner from NARAL Pro-Choice America. Due to their
support, I was able to obtain data from both organizations that proved invaluable for this
research project. Last, but not least, my husband David deserves highest praise for
unwaveringly championing this project and providing assistance in a million ways.
iv
TABLE OF CONTENTS
Dedication ii
Acknowledgements iii
List of Tables vi
List of Figures vii
List of Abbreviations viii
CHAPTER 1: INTRODUCTION 1
Purpose of the Study 1
Chapter Summaries 7
CHAPTER 2: ORGANIZATIONAL COMMUNITY EVOLUTION 11
Environments, Resources, Niches, and Forms 14
Variation, Selection, and Retention 23
Competitive Processes 27
Legitimation Dynamics 35
CHAPTER 3: THE EVOLUTION OF ORGANIZATIONAL FORM CONCEPTS 39
Organizational Forms as Complex Cultural Artifacts 41
Feature or Trait-Based Form Definitions 41
Structural Form Definitions 46
Identity-Based Form Definitions 49
The Communicative Aspects of Form Evolution 53
Form Stabilization as a Discursive Process 54
The Role of Organizational Labeling Strategies for Form Evolution 58
Some Implications for Organizational Ecology Research 63
CHAPTER 4: AGGRESSIVE MIMICRY AS A COEVOLUTIONARY STRATEGY 67
The Concept of Aggressive Mimicry 67
Aggressive Mimicry in Organizational Communities 71
The Spectrum of Mimicry Behaviors 74
Conditions for Aggressive Mimicry 81
Evolutionary Implications of Aggressive Mimicry 86
CHAPTER 5: AGGRESSIVE MIMICRY WITHIN THE POPULATION OF CRISIS
PREGNANCY CENTERS 91
Crisis Pregnancy Centers and the Pro-Life Movement 94
The Model Population: Reproductive Health Care Providers 98
The Mimic Population: Crisis Pregnancy Centers 104
CPC Subforms Based on Varying Degrees of Mimicry 106
Hypotheses about CPCs 114
v
Founding Rates of CPCs 114
Founding Rates of CPC Mimics 118
Aggressive Mimicry Label Transformations 122
Organizational Failures 126
CHAPTER 6: METHOD 132
Data Sources 132
Tax-Exempt Charities Databases 132
The Alan Guttmacher Institute RHP Database 135
Online Databases Listing CPCs 135
LexisNexis News Database 137
NARAL Policy Index 137
Supplementary Data Sources 139
Data Preparation Procedures 140
Construction of CPC Database 140
Organizational Vital Rates and Attributes 145
Organizational Label Transformations 148
Constructing a Measure of Cognitive Legitimacy 153
Sociopolitical Legitimacy Operationalization 167
Measures 170
Analyses 178
Founding Rate Analyses 178
Transformation and Mortality Rate Analyses 182
Additional Aspects of Statistical Modeling 185
CHAPTER 7: RESULTS 190
Descriptive Analyses 190
Hypothesis Tests 202
Founding Rates of CPCs 202
Founding Rates of CPC Mimics 206
Aggressive Mimicry Label Transformations 210
Organizational Failures 215
CHAPTER 8: CONCLUSION 222
Discussion and Implications 222
Limitations 231
Suggestions for Future Research 238
BIBLIOGRAPHY 246
APPENDICES 283
Appendix A: CPCs Included in the Analysis Grouped by NTEE Major Group 283
Appendix B: Search Terms for Newspaper Article Research 284
Appendix C: Generalization Thesaurus Used for Textual Analysis 285
Appendix D: Thesaurus Translating Concepts and Phrases Into Six Meta Concepts 290
vi
LIST OF TABLES
Table 1: Viability Regions for Aggressive Mimicry Strategies........................................ 89
Table 2: Examples for Common CPC Labels Covered in Newspapers From 1980 to
2009..................................................................................................................... 113
Table 3: Summary of Hypotheses................................................................................... 131
Table 4: Summary of Searches Conducted in NCCS and GuideStar Databases ............ 142
Table 5: Examples for Typical CPC Labels Classified Along a Mimicry Continuum... 150
Table 6: Descriptive Statistics of Meta Networks Derived From News Coverage
About CPCs ........................................................................................................ 164
Table 7: Descriptive Statistics and Correlations for Founding Rate Analyses............... 199
Table 8: Descriptive Statistics and Correlations for Label Transformation Rate
Analyses.............................................................................................................. 200
Table 9: Descriptive Statistics and Correlations for CPC Failure Rate Analyses .......... 201
Table 10: Negative Binomial Regression Estimates of CPC Foundings, 1989-2009..... 203
Table 11: Negative Binomial Regression Estimates of CPC Mimic Foundings,
1989-2009 ........................................................................................................... 207
Table 12: Piecewise Constant Exponential Hazard Rate Estimates of CPC Mimicry
Label Transformations, 1989-2009..................................................................... 211
Table 13: Cox Regression Estimates of CPC Failures, 1989-2009 ................................ 216
Table 14: Cox Regression Estimates of CPC Mimic Failures, 1989-2009..................... 219
Table 15: Summary of Hypothesized Effects and Corresponding Results From
Hypothesis Tests ................................................................................................. 221
vii
LIST OF FIGURES
Figure 1: Key Events in a Genealogical Lineage Over Time ........................................... 21
Figure 2: Homogenizing, Segregating, and Lateral Niche Transformations.................... 32
Figure 3: Construction of a Genealogical Diagram Based on the Distribution of
Organization Size in a Population......................................................................... 44
Figure 4: Niche Adjustment Dynamics Between Mimic and Model Populations............ 74
Figure 5: Changes in the Population Numbers of RHPs in the U.S. Since 1973.............. 99
Figure 6: State Abortion Policy Climate in the Years 1989 and 2009............................ 103
Figure 7: Newspaper Coverage of CPC Subform Labels From 1980 to 2009 ............... 112
Figure 8: Meta Network Representations of Form Discourse About CPCs Over
Time .................................................................................................................... 166
Figure 9: Total CPC Population Vital Rates From 1970 to 2009................................... 190
Figure 10: Number of CPCs by State per 100,000 Women of Age 14 to 44.................. 191
Figure 11: CPC Population Count Stratified by Subform From 1970 to 2009............... 192
Figure 12: Subpopulation Entries of CPCs From 1970 to 2009 ..................................... 193
Figure 13: Subpopulation Exits of CPCs From 1989 to 2009 ........................................ 194
Figure 14: Label Transformations Among CPCs From 1990 to 2009............................ 195
Figure 15: Mortality Rate Comparison Between CPC Mimics and Non-Mimics.......... 196
Figure 16: Mortality Rate Comparison Between Large and Small CPCs ...................... 197
viii
LIST OF ABBREVIATIONS
AGI The Alan Guttmacher Institute
ARDA Association of Religion Data Archives
ARIS American Religious Identification Survey
BMF Business Master File
CPPR Coalition to Protect Patients' Rights
CPC Crisis pregnancy center
DBF Dedicated biotechnology firm
EIN Employer Identification Number
FIPS Federal Information Processing Standard
HMO Health management organization
IRS Internal Revenue Service
MSA Metropolitan statistical area
NAICS North American Industrial Classification System
NAF National Abortion Federation
NARAL National Abortion Rights Action League
NCCS National Center for Charitable Statistics
NGO Nongovernmental organization
NIFLA National Institute of Family and Life Advocates
NRLC National Right to Life Committee
NTEE National Taxonomy of Exempt Entities
RHP Reproductive health care provider
VSR Variation, selection, and retention
ix
ABSTRACT
Communication scholarship has significantly improved organizational ecology
models by emphasizing that the members of organizational communities are participating
actively in the enactment of their environment. Organizational forms are viewed as
contested and multi-faceted action frames which emerge from community interactions
and simultaneously shape them. Form negotiations mediate the evolutionary fate of
population members adhering to a form. Existing institutional ecology approaches focus
on discursive strategies foregrounding the distinctive and unique organizational identity
underlying a certain form. The assumption prevails that organizational populations can
only thrive if their form is recognized clearly by diverse audiences as this entails that it
has achieved cognitive legitimacy. But not all organizational populations benefit from
clearly communicating who they are, particularly if they are operating in an environment
that attaches low cognitive legitimacy to their form. The biological concept of aggressive
mimicry, which refers to a competitive evolutionary strategy based on form ambiguity,
illuminates the circumstances when members of a population exploit the lack of clarity
about their form to their benefits. Among other tactics, aggressive mimics in
organizational populations engage in purposive labeling of their names to increase their
resemblance with a population of models.
For the purpose of this investigation, aggressive mimicry is introduced and the
conditions under which organizational populations might employ it as a strategy are
examined. The short-term and long-term consequences of its use are presented along with
suggestions about how to revise existing ecological models to account for the
evolutionary benefits of ambiguous form communication. A longitudinal investigation
x
of the changing population composition of crisis pregnancy centers (CPCs) in the United
States provides an opportunity for testing a revised set of expectations in an empirical
setting. Some CPCs routinely engage in aggressive mimicry when they closely imitate
the organizational forms of reproductive health care providers (RHPs) in an effort to
compete with them for the same "clients," women facing unwanted pregnancies. There
are two ways to enter the subpopulation of mimics: (1) CPCs select organizational labels
reminiscent of the organization names typically chosen by RHPs at the time of founding,
and (2) CPCs change their existing organizational names so that they resemble those of
RHPs at some point during their existence.
Drawing on multiple archival data sources, the effects of differing cognitive and
sociopolitical legitimacy levels on the organizational vital rates of CPCs are investigated.
As a measure of cognitive legitimacy in the sense of communicative form stabilization,
the study relies on a content analysis of media discourse about CPCs. Based on the
analysis of newspaper coverage of CPCs spanning a period of over two decades, textual
network approaches are employed to develop a novel longitudinal measure of cognitive
form legitimacy. The influence of legitimacy effects and other factors such as density-
dependent effects, the availability of environmental resources, and cohort effects are
examined in a variety of ecological models. Negative binomial regression models are
used to examine the dynamics of CPC founding events. The CPC label transformation
events, which represent their employment of a mimicry strategy, are estimated with
piecewise constant exponential hazard models. The effects of the covariates on CPC
failure rates are examined in a series of Cox proportional hazard models.
xi
The results of the analyses suggest that as cognitive legitimacy levels rise, CPCs
engage in mimicry transformations at a higher rate. At the same time, their mortality rates
increase under conditions of increasing levels of cognitive legitimacy. Thus, mimicry
strategies seemingly bear penalties for both impostors and non-mimicry subpopulations.
However, mimics among CPCs are found to have superior survival chances when
compared to non-mimics. Additional findings indicate that organizations entering the
mimic subpopulation through adaptation disband at a lower rate than CPCs that are
"born" into mimicry via differential selection.
Keywords: cognitive legitimacy, community ecology, Cox regression, event history
analysis, evolution, institutional ecology, labeling, mimicry, negative binomial
regression, nonprofit organizations, organizational forms, organizational identities,
piecewise constant exponential models, population ecology, pro-life, meta network
analysis, social movement organizations
1
CHAPTER 1: INTRODUCTION
Purpose of the Study
In recent years, scholars have utilized social movement theories to explore the
role of interorganizational communication in the evolution of organizational communities
(Aldrich & Ruef, 2006; Clemens, 1996; Lounsbury, Ventresca, & Hirsch, 2003; Rao,
1998). This research has contributed greatly to our understanding of the importance of
organizational forms as frames that coordinate community interactions. Essentially, the
new perspectives have reconceptualized forms as evolving social codes that are
discursively negotiated by diverse audiences (Hannan, Pólos, & Carroll, 2007). Rather
than being based on objective features that allow for their systematic classification,
organizational forms are now believed to emerge and stabilize through discursive
processes and interorganizational enactment (Kennedy, 2008; Monge & Poole, 2008;
Moore & Hala, 2002; Rao, Monin, & Durand, 2003; Ruef, 1999; Weick, 1979). In a
mutually constitutive fashion, organizational communities enact "fields that are knitted
together by collective beliefs about desirable product attributes, market structures,
appropriate ways of doing business, and the relative quality of member firms" (Porac,
Ventresca, & Mishina, 2002, p. 593).
Many scholars acknowledge that multiple audiences consisting of producers and
consumers actively participate in the collective process of form invention, transformation,
and contestation (Baron, 2004; Hannan, Baron, Hsu, & Koçak, 2006). Institutional
ecologists emphasize that in addition to its sociopolitical legitimacy, the recognizability
of a form plays an important role for the sustainability of populations adhering to it
(Suchman, 1995). A new type of organization must attain "acceptance . . . as a taken for
2
granted feature of the environment. The highest degree of cognitive legitimacy exists
when a new product, process, or service is accepted as part of the sociocultural and
organizational landscape" (Aldrich & Ruef, 2006, p. 186). This way, clearly developed
and relatively stable organizational forms function as helpful cognitive templates that
regulate the embeddedness of organizations in networks of "exchange, competition and
production" (Fligstein & Dauter, 2007, as qutd. in K. Weber, Heinze, & DeSoucey,
2008).
The ability to construct an identifiable organizational form is often seen as a
critical prerequisite for the successful establishment and subsequent proliferation of a
population (Hannan, et al., 2007). Emphasizing how they differ from other populations
affords its members a measure of shelter against competitive pressures originating from
within their community. The attainment of cognitive legitimacy is deemed so vital by
scholars that members of an emergent population are assumed to readily collude in the
collective project of carving out a distinct niche for themselves (Lounsbury, Ventresca, &
Hirsch, 2003). Organizational labeling has been studied as an important survival strategy
for organizations to claim form membership in categories that are recognized by diverse
audiences (Glynn & Abzug, 2002; Glynn & Marquis, 2004).
While this new line of theorizing attests to the importance of communication and
acknowledges the complexity of organizational forms as templates for organizing, it fails
to articulate the potential strategic benefits for a population if discourse about its form
remains ambiguous. The concept of aggressive mimicry (Lloyd, 1965) challenges the
emphasis that has been placed on form recognizability as it highlights organizational
strategies that seem to rely on vagueness about a form rather than on its distinctiveness.
3
Generally, all types of mimicry revolve around maintaining form ambiguity (Starrett,
1993) rather than clarity. Biologists use the concept of aggressive mimicry to describe the
behavior of members of a species that benefit from the very ability to avoid being
recognized as predators by their prey (Ruxton, Sherratt, & Speed, 2004). In particular,
aggressive mimicry is defined to occur when members of a predatory population deceive
their prey by closely imitating a model species that is perceived as harmless by their
targets (Wickler, 1968). The classical mimicry system therefore always involves three
parties: Model, mimic, and dupe (Pasteur, 1982).
Mimics engage in deception towards a population of dupes by displaying signals
that cause the dupes to mistake them for members of a population of models (Wickler,
1968). In other words, predatory mimics exploit the friendly or, at a minimum, neutral
relationship between model and dupe populations as their resemblance to the models
allows them to better approach the unsuspecting dupes. According to a mimicry
framework, the fabled wolves in sheep's clothing can be considered as mimics as they are
sending out ambiguous cues commonly associated with the appearance of sheep.
Accordingly, members of a population of deer do not perceive their true identity, but
mistake them for sheep upon encountering them. As the relationship between deer and
sheep is neutral, members of the deer population may be duped into approaching the
impostors instead of fleeing from them, which likely results in their early demise.
Biologists have identified mimicry as a coevolutionary strategy that can be very
successful under some, but not all circumstances (Huheey, 1988; Mallet & Joron, 2007;
Turner, 1988). The occurrence of aggressive mimicry in the context of organizational
populations requires a thorough investigation of the conditions enabling it.
4
Traditionally, organizational ecology has concentrated on investigating
institutional form diversity and organizational forms in the private sector, including
studies of American phone companies (M. L. Barnett & Carroll, 1987) and Dutch
newspapers (Boone, van Witteloostuijn, & Carroll, 2002). Exploring population
composition changes in voluntary organizational communities takes on new importance
in light of the increasing growth of the diverse field of nongovernmental organizations
(NGOs; as documented by V. A. Hodgkinson, Weitzman, Abrahams, Crutchfield, &
Stevenson, 1997). The current analysis adds to this field of research as it examines the
applicability of postulates about organizational ecology and resource partitioning (Carroll
& Swaminathan, 2000; Hannan & Freeman, 1977) within the context of social
movements composed of populations of voluntary organizations. NGOs in social
movement communities are influenced particularly strongly by legitimacy concerns.
These and the competitive dynamics governing their interactions invite further study.
An examination of the foundings, label transitions, and failings of crisis
pregnancy centers (CPCs), which constitute a specialized nongovernmental
organizational form within the pro-life movement, represents a showcase for the pursuit
of aggressive mimicry within an organizational social movement community. The
evolution of the CPC population in the United States has not yet captured the attention of
organization ecologists. CPCs originally emerged from a pro-life grassroots mobilization
against the legalization of abortion in the United States in 1973. The changing
organizational labeling strategies of some CPC subforms over time suggest that it might
be exactly their close resemblance to reproductive health care providers (RHPs) that has
contributed to the proliferation of the population. Their imitation of RHPs helps CPCs
5
when they compete with them for the same client group, which is composed of women
potentially facing unwanted pregnancies. Their basic organizational mission to dissuade
abortion-seeking women from obtaining the procedure has remained the same, but the
organizational naming strategies of CPCs have transformed over time and in close
response to the population fluctuations of RHPs. Viewed from a biological perspective,
CPCs with organizational labels reminiscent of RHPs constitute the mimics in a three-
party mimicry system. This system further includes RHPs as the models mimicked by
CPCs and their clients as the dupes responding to the ambiguous signals communicated
by the mimics.
The application of the concept of aggressive mimicry from the life sciences to
organizational evolution allows for theorizing that poses intriguing challenges to the
traditional sets of assumptions informing both organizational ecology perspectives
(Hannan & Freeman, 1977; Hunt & Aldrich, 1998) and institutionalism (DiMaggio &
Powell, 1983). Communication-centered approaches (Monge & Poole, 2008) are relevant
for illuminating some of these emerging peculiarities as they acknowledge the role of
discursive processes as important mediators of form evolution within organizational
communities. These principles aid in making sense of mimicry phenomena that can be
observed in a variety of industries, from the emerging industrial field of cosmetic
dentistry to the ailing industry of hedge funds. In the cosmetic dentistry sector, which is a
growing industry that revolves around aesthetic procedures such as teeth-whitening,
organizational entrepreneurs are mimicking the appearance of dentists' offices (Witcher,
2008) even though their staff members are not professionally trained. Thus, they are able
to draw on the legitimacy of professional dentists towards their patients when offering
6
their services to clients. Up until the occurrence of the recent financial crisis, the inner
workings of hedge funds have "stay[ed] opaque to the general investing public" (Fung &
Hsieh, 1999, p. 309) and they were able to appropriate the legitimacy associated with
mutual funds. Increased media coverage detailing the high risks typical for hedge funds,
however, informed investors about their deceptive similarity with traditional investment
funds, and the dupes began to steer clear of such high-volatility investment options
(Bookstaber, 2007).
The main purpose of this research was to introduce aggressive mimicry as a
concept that may capture a unique, but not uncommon set of evolutionary strategies
employed by organizational populations which communicate to external audiences their
adherence to a specific organizational for strategic purposes. This endeavor requires some
familiarity with community ecology and recent advances in the definition and
operationalization of organizational form concepts. Accordingly, an extended review of
current literature on form evolution in organizational communities is included, which
summarizes the contributions of communication perspectives as they infuse institutional
ecology (Baum & Powell, 1995; Suchman, Steward, & Westfall, 2001). This theoretical
framework focuses on community ecology in combination with insights from new
institutionalism. Community ecologists study the coevolutionary dynamics that govern
the interactions between organizational populations, so they typically take a higher-level
analytical perspective than population ecologists. Several postulates based on institutional
ecology approaches are revised to accommodate the theoretical contributions that arise
from adopting the concept of aggressive mimicry.
7
A second purpose of the project was to draw on insights from communication
studies to advance the development of textual approaches to measuring the cognitive
legitimacy (Johnson, 2004; Rao, 2001; Suchman, 1995) of the CPC organizational form
over time. The third objective of undertaking this research was to shed light on an
overlooked population of social movement organizations. Feminist scholars and social
movement theorists have explored societal struggles revolving around reproductive rights
from a variety of lenses. They have looked at the activities of social movement
organizations in both pro-choice and pro-life movements (Maxwell, 2002; Staggenborg,
1991). But the role of CPCs within the field of reproductive health care has not been
examined previously. Therefore, this study employs longitudinal ecological modeling to
elucidate how this population has evolved. While observers have noted that CPCs have
not only grown in numbers (Hartshorn, 2003; Lin & Dailard, 2002), but also considerably
changed their appearance since the first CPCs opened their doors, their proliferation and
large-scale transformation have not been analyzed systematically before. By examining
an understudied, but important social movement sector, bringing in new perspectives
from the life sciences in order to advance institutional ecology, and providing an
alternative operationalization of cognitive legitimacy based on meta network analysis, the
project contributes to our understanding of large-scale organizational processes.
Chapter Summaries
Chapter 2 presents a variety of evolutionary concepts useful for understanding
how organizational forms have been customarily conceptualized in population and
community ecology models. After a review of ecological approaches to organization
studies, a brief description of environments, resources, organizational niches, and forms
8
follows and genealogies are introduced as conceptual tools for mapping form evolution.
Together with the principles of variation, selection and retention, the long-standing
tradition in ecology to differentiate between Lamarckian and Darwinian models of
evolution is presented and discussed. Density-dependent competition and resource-
partitioning processes are reviewed in terms of their relevance for niche and form
transformations. The chapter concludes with a first look at the typical ways in which
legitimation processes have entered into ecological models.
Chapter 3 focuses on a variety of impulses from new institutionalist theorizing
and communication studies, which have spurred efforts to reconceptualize organizational
forms. Altogether, three approaches to form definitions are differentiated. All of them
flow into current perspectives on organizational forms as complex and evolving cultural
artifacts. The presentation of communicative aspects of organizational form emergence
and transformation entails envisioning form stabilization as a discursive process, which is
described next. The section is followed by a brief discussion of the role of organizational
labeling strategies in the evolution of forms. At the end of the chapter, legitimacy
concepts from organizational ecology are revisited as novel form concepts necessitate the
reexamination of existing ecological models.
In chapter 4, the concept of aggressive mimicry is introduced, first in its original
formulation in the life sciences, then as it may apply in organizational communities. The
characteristics of aggressive mimicry as an evolutionary strategy are presented and
examples for populations employing it are provided. As aggressive mimicry is a
comparatively specialized competitive ecological behavior, its boundary conditions and
possible short-term as well as long-term effects on organizational communities are
9
thoroughly examined. At the end of the chapter, the implications of the occurrence of
aggressive mimicry for current ecological models are explored.
Chapter 5 showcases the population of crisis pregnancy centers CPCs in the
United States as a potential case of aggressive mimicry. In a brief summary of the
development of the CPC population within the pro-life movement, the role of CPCs
within a larger social movement community is outlined. Several subpopulations of CPCs
are presented to familiarize the reader with the study population. The changing
population composition of CPCs is described in its historical context. Next, the
coevolutionary relationship between CPCs and RHPs is examined, with particular
emphasis on the regulatory environment within which the competitive struggle between
the two populations has been unfolding. The chapter concludes with a number of
hypotheses about factors affecting the birth, mimicry transformation, and demise of CPCs
during the last twenty years.
Chapter 6 gives an overview over the archival data sources employed, the
procedures used to prepare the data, the measures created for the analysis, and the
analytical models selected to test the hypotheses. While most information about the
organizational vital rates of CPCs was extracted from the nonprofit database of the
National Center for Charitable Statistics (NCCS), many additional sources were used to
validate the information and construct the overall sample. The data preparation is
presented in detail as many decisions had to be made along the way about how to best
deal with gaps in the data record and inconsistent information about the fate of
organizations in the sample. For constructing the measure of cognitive legitimacy as well
10
as the operationalization of organizational label changes, textual analysis methods were
employed, which are also described in more detail in this chapter.
Chapter 7 focuses on the results of the study. In the first part, the vital rates of
CPCs are presented and entry as well as exit rates of CPC subpopulations over time are
displayed in a variety of graphs. Preliminary nonparametric analyses of the differential
survivor functions characterizing mimics versus non-mimics as well as large versus small
organizations are provided next. In addition, descriptive statistics for all variables
included in the statistical models are listed. In the second part of the chapter, the results
of a variety of hypothesis tests are presented. In order to evaluate hypotheses about the
founding rates of CPCs and CPC mimics, count models were estimated. Event history
modeling approaches were used to test hypotheses concentrating on organizational label
transformations and failures as outcome events.
In chapter 8, the study results are reviewed and the influence of form legitimacy,
environmental resource availability, population density, cohort membership, and
organizational size on the vital rates of CPCs is discussed. While the results confirm the
occurrence of aggressive mimicry among CPCs and some of the suggested revisions of
conventional ecological modeling appear justified, there were also unexpected findings,
which are examined in more detail. Several kinds of study limitations are outlined, which
pertain to the data sources and measures used as well as paradigmatic constraints
stemming from the theoretical frameworks employed. The chapter concludes with
recommendations for future research, which include further scrutiny of legitimation
processes, model-mimic interaction dynamics, and an examination of the ethical
implications of aggressive mimicry strategies employed by organizations.
11
CHAPTER 2: ORGANIZATIONAL COMMUNITY EVOLUTION
The following paragraphs provide a brief overview of the content covered in this
chapter. An ecological perspective on organizational change requires a shift from
examining individual organizations to studying the changing makeup of groups of similar
organizations called populations (Hannan & Carroll, 1992). As suggested by Ruef (2000),
the evolution of organizational populations is best understood if their transformation is
viewed from a community ecology perspective. It is their social environment, comprised
of other populations with which their fate is linked, that shapes, but also constrains their
evolutionary trajectory. Typically, community ecologists study how a community evolves
and forms appear and vanish in it by examining the population dynamics of foundings,
disbandings, mergers, and acquisitions (Carroll & Hannan, 2000).
Expanding on the scope of Hunt and Aldrich’s (1998) technology-focused
community definition, Aldrich (1999) maintains that "an organizational community is a
set of coevolving organizational populations joined by ties of commensalism and
symbiosis through their orientation to a common technology, normative order, or legal-
regulatory regime" (p. 301). A community thus conceptualized includes members of a
variety of organizational forms that coexist interdependently within a shared ecological
space. Boundaries around communities can be drawn based on geography, functional
interdependence (Ruef, 2000), or on the distribution of resources (Aldrich & Ruef, 2006).
The emerging community of biotechnology organizations (Oliver, 2009), e.g., includes
functionally interdependent populations such as biotech firms, venture capitalists,
regulatory agencies, pharmaceutical companies, and research institutions (Powell, White,
Koput, & Owen-Smith, 2005).
12
Community ecology as a theoretical lens elaborates Hannan and Freeman’s
(1977) original formulation of population ecology, which focused on a Darwinian-style
selection process between competing organizations populating a scarce resource space
(Astley, 1985). It constitutes a major shift in thinking about organizations as it
acknowledges their evolution within a larger environmental context. Additionally, the
community ecology framework not only includes an emphasis on community
relationships within a surrounding environment, but it also acknowledges that these
population linkages may be collaborative as well as competitive. Thus, is has proven to
be particularly valuable for studying social movements and communities composed of
nongovernmental organizations, which are often characterized by both competition and
mutualism (Simons & Ingram, 2004).
Hunt and Aldrich (1998) note that "the coevolution of an organizational
community depends on the simultaneous processes of variation, selection, retention
[VSR], and struggle at the population level. What emerges is a new phenomenon that has
subsequent consequences for the populations within the community" (p. 293). Variation,
selection, and retention lie at the heart of evolutionary change and will be explained in
more detail later in this chapter. Shumate, Fulk, and Monge (2005) summarize them as
follows:
Variation is defined as changes in routines, competencies, resources, or forms.
Variations can occur as the result of random events or as the result of planned
human actions (Romanelli, 1999). The source of variations can be either within or
among organizational populations as well as their individual or collective
environments (Astley & Van de Ven, 1983). Selection is defined as the
elimination of variation in organizational routines, the choice of one alternative
over others (Nelson & Winter, 1982). Retention occurs when organizations
choose to reaffirm past selections and maintain past routines by re-enacting them
over time. (p. 484)
13
In addition to population-level evolutionary processes, community ecology
approaches are sensitive to VSR processes on multiple levels of analysis (Van de Ven &
Grazman, 1999, p. 189). Thus, they highlight that evolutionary processes on the
intraorganizational, organizational, population, and community levels coevolve with each
other. According to this perspective, organizations have to simultaneously negotiate
between micro- and macro-organizational processes: They are subjected to upward as
well as downward causation while affecting other evolutionary levels through their
interactions.
A population of organizations is grouped conceptually based on the assumption
that all of its members share the same organizational form (Aldrich & Ruef, 2006). In his
book on organizational taxonomy, McKelvey (1982) lays out how form evolution occurs
within a social community of organizations. Organizational communities typically exhibit
functional differentiation, with distinct populations assuming specific roles associated
with their form (Freeman & Audia, 2006; Hawley, 1986). McKelvey (1982) argues that
due to this functional interdependence, "the course of evolution of one form is
determined by ecological pressures created in part by the evolution of other species in the
same community" (p. 249). Therefore, population dynamics can be observed in
concordance with the dynamics of organizational form evolution.
Ecological models emphasize the role of resources for the emergence, survival,
growth, and demise of organizations. Due to the similarities in their characteristics,
organizations in a population tend to exhibit comparable survival patterns, or, in other
words, a "shared fate [sic]" (Freeman, 1995, p. 223). The vital rates of populations in the
community fluctuate while they struggle for achieving external legitimacy and
14
compete with each other for scarce resources (Hannan & Carroll, 1992). Within a
communal ecological system, the size of the resource space is dynamic rather than finite.
The resources available to its members are moderated through community relationships
and due to their interorganizational linkages, some populations are buffered from direct
adverse effects stemming from reductions in the overall resource environment (Miner,
Amburgey, & Stearns, 1990).
Community ecology approaches are influenced by neoinstitutional theories
(Powell & DiMaggio, 1991) not only in the sense that they are sensitive to organizational
fields composed of functionally diverse organization types (J. W. Meyer & Scott, 1983).
They also tend to emphasize the need for longitudinal examinations of population
dynamics in which the varying evolutionary rates are analyzed separately, sociocultural
influences are taken into account, and technical environments are modeled.
Organizational ecology has been a well-established and thriving field, which has much to
offer when it comes to providing explanations for the evolutionary interplay between
populations and changes in the organizational forms chosen for entrepreneurial activity.
In the following, some basic ecological principles that are particularly important for
studying form evolution are reviewed in more detail.
Environments, Resources, Niches, and Forms
As ecological approaches tend to view organizations as members of various
species cohabitating in a shared biosphere, they emphasize the importance of studying
environmental conditions when it comes to tracking the vital rates of organizational
populations (Hannan & Carroll, 1992). The task of determining how exactly such an
environment can be conceptualized is not trivial as it requires a systematic approach to
15
determining the boundaries, analysis levels, and units of analysis in an organization's
environment (J. W. Meyer & Scott, 1983). As mentioned previously, "every organization
is embedded in a network of external influences and relationships which can be labeled as
its environment" (Miles & Snow, 1978, p. 18). Such social environmental processes,
which are mainly regulated through the multiplex relationships within a community, are
considered endogenous to its populations (Carroll & Hannan, 2000). In contrast,
exogenous environmental aspects are large-scale external factors that cannot be easily
influenced by organizational actors within the community. Major types relevant for
organizations include resource availability, governmental support and political forces,
macro-economic regulatory environments, technology regimes, and culture (Aldrich &
Ruef, 2006; Carroll & Hannan, 2000).
Organizational ecologists draw on the notion of environmental carrying capacity
(Hannan & Freeman, 1977; Lomi, Larsen, & Freeman, 2005) in order to appraise how
many organizations can thrive within a specific constellation of environmental
conditions. This concept is defined as "the maximum number of organizations in a
population . . . that can be supported by the social environment at a given point in time
[and it] can inform the emergence of forms in a community ecology" (Ruef, 2000, p.
678). The term munificence describes the degree to which "resources available to firms
are plentiful or scarce, controlling for the number of firms competing for those resources"
(Anderson & Tushman, 2001, p. 689). Accordingly, resource-rich environments exhibit
high levels of munificence and allow for the proliferation of many organizations. In
tandem with fluctuating resource environments, the associated carrying capacities are not
finite quantities, but can expand or decline due to cultural transformations and
16
changing policy environments (Aldrich, 1999). Prevailing societal attitudes and values
may, for example, enable or prevent the development of certain organizational forms.
Zelizer's (1978, 1981) studies on the market for life insurances and insurance policies for
children illustrate this observation. She found that certain organizational forms could not
thrive until the overall moral and social environment became more accepting of practices
involving the financial evaluation of human life. Similarly, the emergence and growth of
industrial communities focusing on biotechnology (Powell, et al., 2005) or organic foods
manufacturing (K. Weber, et al., 2008) were accelerated due to changing technological
and cultural environments.
Organizations differ greatly in their resource utilization patterns within a resource
space that is complex and multidimensional (Carroll, Dobrev, & Swaminathan, 2002).
They have differential needs for resources such as "funding, raw materials, labor,
information, and access to political and social support" (Freeman, 1995, p. 226). For
NGOs, human resources represent a particularly important dimension of their resource
space as they depend on mobilizing their constituents to get actively involved in
furthering their mission (J. D. McCarthy, Wolfson, Baker, & Mosakowski, 1988). As
Schoenherr and Young (1990) showed, the vitality rates of Catholic congregations are
significantly related to the size fluctuations of the population of clergy, which supplies
leadership for them. Consequently, competition between congregations is commonly
modeled by examining the demographic characteristics of their client and volunteer bases
(Popielarz & McPherson, 1995). Following a similar logic, Baum and Oliver (1991)
grouped daycare centers in Toronto into distinct resource niches by stratifying them
according to the age range of the children they were licensed to serve.
17
The ecological niche concept has been imported successfully into organizational
evolution as it aids in examining resource-based competition between populations
drawing on similar environmental resource combinations. As noted by Popielarz and
Neal (2007), "the niche of a species is the set of environmental states in which it thrives"
(p. 68). It is especially useful to consider niches from a functional perspective as they
describe "the role of a population (or species) in a community, a population's 'way of
earning a living'" (Hannan & Freeman, 1989, p. 95). A fundamental niche includes all
those positions in a limited resource space that allow a population to access the resource
mix necessary for its survival (Hsu, 2006b; Popielarz & Neal, 2007). Due to the scarcity
of resources and the ongoing competition between populations for accessing them,
fundamental niches are broader than the "realized" niches actually occupied by
populations.
The differences between populations in terms of their ability to extract resources
from their environment are captured with the concept of niche width (Freeman &
Hannan, 1983; Rhee, Kim, & Han, 2006). A typical distinction is drawn between
generalist and specialist populations, which "take opposite positions concerning their
niche width and fitness" (Péli, 1997, p. 5). Generalist populations are able to exploit a
broad range of environmental resources, but there is no point along the continuum where
their survival chances are particularly good. Correspondingly, their fitness distribution,
which plots different positions in the resource space against a population's survival
probability, appears flat and broad. In contrast, the niches of specialists are narrow, which
means that they can only exploit a very restricted range of environmental resource
combinations. However, their survival chances are excellent throughout the whole
18
width of their niche, which is revealed in the shape of their fitness function: It is slender
and peaked. Carroll, Dobrev, and Swaminathan (2002) suggest that "a specialist designed
well for a particular environmental state will always outperform a generalist in that same
state" (p. 5). In summary, both a generalist and a specialist niche position are associated
with distinct benefits and costs: The former is more advantageous for long-term survival
in environments that are highly variable whereas the latter is better suited for
organizational persistence in stable environments (Popielarz & Neal, 2007).
Organizational niches and forms constitute concepts that are tightly linked.
Carroll and Hannan (2000) maintain that populations are groups of organizations that
depend on the same organizational resource combinations and are "characterized by a
particular organizational form" (p. 65). Therefore, organizational form definitions for a
population of organizations frequently draw upon examining the resource niche its
members occupy. They are vital for ecological studies as population transformations can
only be evaluated once it is specified what exactly it is that is changing. Form concepts
"clarify the point that evolution is the study of changes in types, not changes in individual
entities" (Romanelli, 1991, p. 84). According to organizational ecologists, forms can be
thought of as meaningful abstract concepts that classify organizational populations based
on their distinct characteristics. Not surprisingly, scientists have repeatedly tried to
develop taxonomies and typologies for organizational populations based on a variety of
different criteria (Romanelli, 1991; Scott, 1981).
Niches and forms must be kept conceptually distinct even though they define each
other (Hannan & Freeman, 1989). Whereas the former specifies a resource combination,
the latter describes a set of organizational characteristics that are salient for the
19
exploitation of this resource combination. C. Alexander’s (1964) discussion of form as a
design problem that is context-specific illustrates this relationship between form and
niche. He notes that if someone were to invent "a simple kettle . . . [s/he] has to invent a
kettle which fits the context of its use" (p. 60). The form "kettle" specifies a whole range
of design requirements which have to be met by objects in order to be classified as
kettles. Due to these characteristics, objects that qualify as kettles can be put to a range of
uses typically mapped out for kettles. Accordingly, an organizational form always entails
various specifications of how its adherents fit into their resource space. The dynamics
between populations encroaching upon each other's niche space also affect the boundaries
of forms (Hannan & Freeman, 1989).
Astley (1985) points out that community ecology emphasizes how new lineages
emerge rather than restricting itself to studying only established organizational forms.
However, it is very challenging to predict the emergence of a new form (Romanelli,
1991) and determine when it has turned into a distinct recognizable blueprint for
organizing that differs from its predecessors. A major difficulty mentioned by Aldrich
and Baker (2001) is that usually only relatively common populations are able to attract
scholarly attention as those that fail to proliferate do not appear in sufficient densities
over extended periods of time to attain notice. Secondly, organizational forms must have
become somewhat stable over time in order to be able to function as classificatory objects
useful for the study of change.
Organizational genealogies are conceptual visual tools that describe the extent
and direction of form change as they constitute "a record of descent or lineage of a group
from its ancestors until the present day" (Van de Ven & Grazman, 1999, p. 187).
20
Genealogical analyses such as Van de Ven and Grazman’s (1999) health care industry
study are able to represent longitudinal shifts in organization forms in ways similar to the
construction of phylogenetic trees in species biology (Weitzman, 1992). By inspecting
the frequency and types of branching occasions in a genealogy, researchers learn about
the rate and extent of form change and are prompted to look for coevolutionary
population dynamics that explain this change. Biologists have developed cladistic
classification principles, which group species based on their common descent from the
same progenitor. These principles are essential for constructing a genealogical tree as it
requires tracing change back in time to a shared ancestor (I. P. McCarthy & Gillies,
2003). Genealogies are valuable supplements to organizational taxonomies based on the
similarity of characteristics alone (Richards, 2009) as they contextualize forms as
occurring within specific historical environments and provide explanations for factors
that may have precipitated their branching into subforms.
Due to creative hybridization events and managerial recombination (Haveman &
Rao, 2006; Oliver & Montgomery, 2000; Rao & Singh, 1999), organizational forms may
have evolved from unusual ancestors, and genealogies are able to illuminate these
unexpected ancestral entanglements. As described by Van de Ven and Grazman (1999),
some of the modern health management organizations (HMOs) they studied evolved
from insurance companies, others from nursing schools and poorhouses. Genealogies are
only approximate and conceptual mappings of form diversification, so it is not possible to
quantify from a genealogy how strongly a new subform differs from its ancestral
predecessors (Nguyen, 2005). However, it provides all the elements needed to illustrate
the full range of possible form transformation events, as can be seen in Figure 1.
21
Figure 1: Key events in a genealogical lineage over time.
Note. Adapted from Van de Ven and Grazman (1999, p. 188).
Items B and E in Figure 1 show branching and convergence events. These refer to
form transformations in which a single form subdivides or multiple forms combine into a
single one. It is noteworthy that despite having existed in earlier times, forms can
disappear entirely (see item C in Figure 1), particularly if new entrepreneurial templates
replace them. The resource recovery industry, for example, has practically vanished,
along with its procedures and technical processes, after it was substituted by recycling as
an industrial category in the late 1990s (Lounsbury, et al., 2003). During the 1980s,
resource recovery was mainly associated with the incineration of waste, which was then
considered the most promising solution for municipal solid waste management
(Lounsbury, et al., 2003). Some forms are no longer represented by actual organizations,
as was the case with plantation agriculture in the South of the U.S. (Ruef, 2004b), but
they persist as cultural templates as they are still recognizable organizational blueprints
pursued in other regions of the world.
22
Organizational forms may also disappear and appear again at a later point in time.
This means that there may be periods when an organizing blueprint exists, but no
entrepreneurs are actually employing it, as was the case with pubs and liquor shops
during Prohibition (Carroll & Hannan, 2000). Carroll and Hannan (2000) argue that "the
notion of form provides a natural way of defining the continuity of organizational
populations over gaps in existence" (p. 61). It may be quite consequential for the study of
populations to have thorough knowledge about the organizing templates employed by
them. Populations adhering to a previously extinct, but once well-known form, which
may be called "de antiquo" (Dobrev, 2001, p. 422), will usually grow at a faster pace than
new populations that first have to construct novel niches for themselves. Organizations
are said to enter a population "de novo" if they employ its form template at the time of
founding. In contrast, organizations that switch into an existing population by exiting
from a different one are described as "de alio" population entrants (Carroll, Bigelow,
Seidel, & Tsai, 1996). Another type of entry mechanism is observed among organizations
that "arise by merger or fission" (Carroll & Hannan, 2000, p. 42) of organizations within
a population, which is sometimes called a "de ipso" entry.
As Usher and Evans (1996) argue, it is demanding to define transformation events
as one has to evaluate at which point a form has changed to the degree that the resulting
template must be considered as novel. Rao and Singh (1999) offer recommendations that
are based on differentiating between weak and strong types of speciation. If a new form
differs from a previous one in regard to all its central features, it exemplifies strong
speciation. Analogously, "weak speciation occurs when the new form differs from
existing ones on only one or two less core dimensions" (Rao & Singh, 1999, p. 66).
23
Eventually, genealogies constitute conceptual drawings aimed at depicting transformation
events in ways that capture change over time best, but they cannot remedy the
definitional difficulties that are bound to emerge (M. S. Weber, 2010).
It is important to keep in mind that genealogies do not track the transfer of actual
material resources such as warehouses and personnel from parent to progeny (for a
noticeable exception, see D. J. Phillips, 2002). Instead, they summarize compilation
processes (Suchman, Steward, & Westfall, 2001), which entail the cultural transmission
of organizing blueprints. However, it can be shown how organizational forms as a kind of
genotype are in fact reproduced via tangible resource transfers happening within and
between existing organizations, which has been demonstrated empirically in Phillips's
(2002) study on law firms in Silicon Valley.
Variation, Selection, and Retention
The identification of relevant evolutionary units for the study of change is a first
step towards developing an ecological model of organizational community dynamics. A
brief review of essential theories describing how change actually unfolds may help to
illuminate a core set of theoretical mechanisms governing the evolutionary process in
general. Sociocultural evolution can be decomposed into three interlinked and partially
incompatible elements. As noted by Campbell (1965), evolution entails the processes of
variation, selection, and retention. Variations stem from organizational diversity.
Organizations may vary in terms of one or several attributes, for example characteristics
such as the sets of routines they employ or the organizational forms they pursue. The
emergence of variety can be due to "whatever reason: planned, unplanned, haphazard,
systematic, random, predictable, or heterogeneous variations in some activity,
24
behavior, or structure (Aldrich, 2008, p. 34). The occurrence of variation is a vital aspect
of the evolutionary process as new variants constitute new evolutionary possibilities. Due
to selection pressures, some variations are preferred over others and proliferate
differentially while others disappear (Campbell, 1965). In other words, environments
differ in terms of the organizational attributes they favor and those organizations with a
feature mix that best fits their selection criteria are considered fittest. Thus, selection
always arises from specific environmental conditions and at any given time, those entities
that fulfill their requirements best are most likely to thrive. As a third evolutionary step,
"retention occurs when selected variations are preserved, duplicated, or otherwise
reproduced so that the selected behavior is repeated on future occasions or the selected
structure appears again in future generations" (Aldrich, 2008, p. 34). Environmental
conditions are conducive to such a persistence of selected variants if they are relatively
enduring as this means that the selection criteria also remain stable.
While theorists generally agree on the three steps encapsulated in the VSR model,
they tend to foreground either selective or adaptive explanatory mechanisms when it
comes to the study of change observed in the composition of organizational populations
(Usher & Evans, 1996). Population ecology scholars usually conceptualize compositional
transformations as a result of the differential entry and exit of organizations (McKelvey,
1994), which means that selective or Darwinian change is emphasized in their models.
They argue that on the level of populations, composition changes happen when novel
organizational variants replace older cohorts of organizations and thereby alter the
existing mix of population characteristics. Viewed from this perspective, entry-selection
mechanisms, which govern the ease of entry for novices, are thought to differ from
25
exit-selection mechanisms, which determine the dropout of incumbents (W. P. Barnett,
Swanson, & Sorenson, 2003).
Entrepreneurs may, for example, be highly motivated to enter a certain industry,
but if the operation of an organization in this industry requires sophisticated resources
such as professionally trained specialists, entry-selection pressures are high and only few
foundings can occur. Changing business regulations governing industries exemplify exit-
selection mechanisms. If additional taxation is levied upon the members of an existing
population, those without the necessary endowment of financial resources are likely to
exit the industry first. The dynamic interplay of both entry and exit rates shapes the
trajectory and speed of the population's overall compositional change. Kim and Lee
(2003), for example, applied a Darwinian framework to explain how differential selection
processes in the computer memory industry led to the overall demise of small, specialized
organizations and a proliferation of large, diversified latecomers. They found that while
the specialists dominated the market early on, their market share was taken over via
replacement by larger firms that entered the industry at a later stage. The authors
attributed this to the technological regime of the industry (Kim & Lee, 2003).
Impulses from institutionalist perspectives, which traditionally focus on the
isomorphic adaptation of organizations to their sociopolitical and cultural environment (J.
W. Meyer & Scott, 1983; Scott, 1987), have pressed for an emphasis on adaptation as the
main evolutionary motor. Adaptive or Lamarckian processes take place during the
lifetime of an organization, which is believed to adapt intentionally to its environment.
As Baum and Singh (1996) note, "for adaptionists, variability among organizations over
time reflects the cumulative changes of individual organizations. In contrast, selection
26
theories depict organizations as relatively inert entities for which adaptive response is not
only difficult and infrequent, but hazardous as well" (pp. 1261-1262). According to
Nelson (2007), Lamarckian processes differ from Darwinian processes as the variations
created through adaptation are purposive instead of blind and tend to be imitated by
others if seen to be effective.
In their study on the changing composition of gas stations, Usher and Evans
(1996) show that over a period of 30 years, novel forms of gasoline outlets emerged
which eventually began to replace existing forms. In the 1950s, gasoline retail stations
mostly consisted of service stations, which offer gas pumps and a repair shop. In later
decades, gasoline entrepreneurs experimented with offering a variety of additional
services as well as with shedding their repair stations. Organizational adaptation occurred
when existing service stations added or deleted organizing elements such as car wash
facilities, repair shops, and convenience retail shops, as long as they retained the core
competency of offering gasoline. Selection events on the level of organizations included
an organization's shutdown of operations as well as its total abandonment of gas retail as
a core business. In addition to the service stations, three main subpopulations were firmly
established by the end of the 1980s. These included gas-pumps-only retailers, gas stations
with car washes, and businesses with convenience stores. Based on their examination of
the relative contributions of adaptive as well as selective processes to the compositional
transformation of the population as a whole, the authors concluded that both adaptation
and selection mechanisms offered powerful explanatory contributions (Usher & Evans,
1996).
27
Given the mounting empirical evidence for the joint occurrence of selection and
adaptation, it can be concluded that both Lamarckian and Darwinian processes play an
important evolutionary role (McKelvey, 1994). Bryce and Singh (2001) regard them as
complementary mechanisms for studying population transformations. In a similar vein,
many organization theorists have argued that both of them lie at the heart of population
change in terms of its composition (Usher & Evans, 1996). Dimmick (2003) suggested
that "clearly, both forms of population change occur; the theoretical task is to identify
the conditions that produce these forms of evolution" (p. 5). A joint examination of entry-
selection through births, exit-selection through deaths, and adaptive transformations is
necessary to adequately model compositional changes within and between populations in
organizational communities.
Competitive Processes
Population ecology has been criticized for its preferential focus on competition
over symbiosis (Clarke, 1991; Young, 1988). First, as summarized by Young (1988), it is
difficult to translate a biological competition model into the study of organizations as it
cannot capture the full breadth of dynamics between organizations. Second, concepts
such as species, niches, and resources pose definitional challenges when used in the
context of the sociology of organizations. Third, population ecologists usually infer rather
than directly observe that "such competition for resources either takes place or accounts
for the survival of organizational species" (Young, 1988, p. 22). However, the process
models developed from the population ecology perspective have proven very influential
for the theoretical explication and the empirical validation of population
interdependencies and the resulting social stratification of organizational communities.
28
So even though community ecology approaches explicitly include the potential for
variation due to form emergence and collective action (Bryant & Monge, 2008), it is
fruitful to review the basic tenets of competition among populations.
Populations are usually embedded in partially mutualistic and partially
competitive relationships within their communities (M. L. Barnett & Carroll, 1987;
Hawley, 1986), contending with some and rallying with others. Baum and Singh (1996)
assert that "the potential for competition between any two organizations is directly
proportional to the overlap or intersection of their organizational niches" (p. 1263). It also
depends on overall resource availability as organizations tend to cooperate only as long
as resources abound (Astley, 1985; Hawley, 1986). Expressed more generally, groups of
organizations can be thought to compete with each other if they rely on the same scarce
resources. Depending upon the extent to which the niches of populations overlap,
relationships between populations within a community can range from full-blown
competition to symbiosis and even display a mix of both (Aldrich & Ruef, 2006).
Resource niches are multidimensional, which makes it possible that competition between
populations is fierce in terms of one research dimension, but relaxed in terms of others.
Sørensen (2004) illustrates this differential resource reliance in his study on the foundings
of Danish companies. He discovered that the founding rates were reduced in industries
characterized by heavy reliance on human resources if their labor markets were already
depleted.
Two types of competition can be differentiated: Diffuse competition ensues when
organizations compete with each other in large numbers whereas direct competition
describes situations in which individual dyads of organizations identify their respective
29
alter as a competitor (M. L. Barnett & Carroll, 1987). Research on population
competition often takes organizational proximity into account to gauge the relative
contributions of direct versus diffuse competition (Carroll, 1997). Regardless of the type
of competition, organizational densities on various geographical scales have been shown
to greatly influence the strength of intra- as well as interpopulation competition. This
preoccupation with population counts goes back to the concept of density dependence,
which constitutes a cornerstone of population ecology modeling (Hannan & Carroll,
1992).
The notion of density dependence is based on the assumption that the level of
competition between organizations utilizing the same resource space is contingent upon
the number of organizations in these populations. Density-dependent processes are able
to account for patterned expansions and contractions of populations in terms of their size,
which is routinely observed across a variety of organizational communities. According to
Hannan and Carroll’s (1992) density-dependence model, young and immature industries
that are sparsely populated offer abundant resources to emerging populations. Each
population in the community comes to occupy a distinct niche in a finite resource space.
These resource niches become more densely populated as time passes. With increasing
population density, legitimacy increases, which precipitates further population growth.
However, the marginal benefits of rising legitimation are finite as "legitimation of an
organizational population increases with density at a decreasing rate and approaches a
ceiling at high levels" (Carroll, 1997, p. 126).
Ultimately, competitive forces intensify and the community reaches its carrying
capacity, i.e., its maximum density (Ruef, 2000). If the density of organizations per
30
niche exceeds its carrying capacity, diffuse competition pressures set in (Dobrev, 2001)
and competition operates at full force. In other words, at this point, the niche becomes so
crowded that organizations are starting to compete with large numbers of peers that they
can no longer distinguish individually (Dobrev, 2001). Ultimately, founding rates drop
and the struggle for scarce resources heightens the risk of mortality.
Graphically, the initial rise and subsequent decrease in the density of a population
can be represented by a ∩-shaped graph, with the number of organizations on the vertical
axis and time measured on the horizontal axis. Density-dependent processes have been
found to apply to a diverse set of industries ranging from manufacturing, financial
services, health care, social services, and telecommunications, to the nonprofit sector (as
listed by Carroll & Hannan, 2000, pp. 218-219). Studies have shown that not only
overall population densities play a role for the establishment of new organizations, but
that the densities of similar populations as well as the specific densities of individual
subforms within a population may drive population entry. Marrett's (1980) study on the
growth of women's medical societies in the 19th Century illustrates this tendency: She
discovered that it was mostly fueled by social network connections of female physicians
with members of established male medical societies.
If a population reaches a high level of density due to its confinement to a specific
niche, resource partitioning may set in as a "late-stage process" (Romanelli, 1991, p.
138). In addition to affecting the number of organizations in a population, the increasing
level of competition also influences the degree of organizational form diversity. After the
carrying capacity of a niche occupied by a population has reached a certain point,
speciation events occur, because populations will try to reduce competition by
31
decreasing their amount of niche overlap with rival populations (Baum & Singh, 1994a;
Carroll, et al., 2002). Resource partitioning is also observed as a population-level
outcome of intrapopulation competition as organizations tend to react to competitive
pressures by seeking to differentiate themselves from similar competitors (Baum &
Singh, 1996). Scholars have used the terms ecological crowding and niche crowding
(Usher & Evans, 1996) to explain the increasing tendency of new community entrants to
carve out small specialist niches for themselves instead of joining a highly competitive
market.
Mature populations are commonly found to partition the niche space available to
them between generalist and specialist subforms in order to alleviate competitive
pressures on all population members. Empirical examples for density-dependent resource
partitioning abound and cover both the not-for-profit and commercial worlds. For
example, Boone et al. (2002) studied Dutch news companies, Mezias and Mezias (2000)
examined the feature film industry, Swaminathan and Carroll (2001) looked at the wine
industry, Greve, Pozner, and Rao (2006) researched the microradio movement, and Soule
and King (2008) investigated the environmental, peace, and women's movements. In
Rindova and Fombrun's (2001) study on specialty coffee retailers, the process by which
the new entrepreneurs entered the community is described as a collective effort sustained
among a subset of organizations. It proved successful in carving out a stable niche
sheltered from the competition by traditional commercial roasters.
Commonly, three major patterns of competition-related niche transformations can
be observed (Amburgey, Dacin, & Kelly, 1994), which are depicted in Figure 2. If
evolutionary pressures, often within restrictive regulatory climates, force a population
32
to concentrate on a smaller niche space, the result is an overall homogenization in terms
of its member characteristics (see Panel A). A differentiation into generalists and
specialists occurs due to segregating pressures, which produce a bimodal distribution of
the population along the resource space (see Panel B). Additionally, competitive
pressures may result in a lateral displacement of a niche further down the resource
continuum (see Panel C).
Figure 2: Homogenizing, segregating, and lateral niche transformations.
Note. Adapted from Amburgey, Dacin, and Kelly (1994, p. 242).
A range of characteristics affecting the ability of organizations to cope with
competition has been identified. Age, size, and structural complexity are among the most
important of these factors. They affect the nimbleness with which organizations respond
to competition as well as their resilience when it comes to surviving the organizational
change efforts they initiated in order to escape it. Older and smaller organizations, for
example, are more likely to fail after initiating an organizational transformation event
(Baum & Singh, 1996). In general, organizations differ in their proneness to structural
inertia, which denotes their reluctance or inability to adapt to shifting environmental
conditions. Larger, older, and more complex organizations are generally believed to be
more resistant to change (Hannan & Freeman, 1984). Structural inertia makes it
A B Original population
Moved population
C
33
difficult for very established organizations to engage in transformations necessitated by
novel environmental conditions. Interestingly, their very inability to adapt may be
advantageous for these organizations as survival benefits accrue to them based on their
resource endowments and reputation alone. Their large size has been found to buffer
organizations from demise even though they might be comparatively "weak competitors"
(W. P. Barnett, 1997, p. 128). With structural inertia being routinely favored by heavily
regulated organizational selection environments, it is therefore possible that inertial
pressures dominate the evolution of populations.
The term environmental imprinting (Stinchcombe, 1965) is a concept that
describes how environmental conditions at the time of founding leave a lasting
impression on organizations and limit the degree to which they can transform themselves.
As Stinchcombe (1965) argues, "organizations formed at any given time must obtain the
resources essential to their purposes by the devices developed at the time. Since these
devices differ, the structures of organizations differ" (p. 212). According to Brüderl and
Schüssler (1990), new organizations have occasionally been found to suffer from the
"liability of newness" (p. 530), but empirical evidence for this phenomenon remains
inconclusive. It encompasses a heightened mortality risk for newly founded organizations
as they are still trying to determine how to best exploit their new niche and such search
processes are costly. New entrants may also lack the social ties necessary to compete at
full force. Additionally, entrepreneurial ventures that assume a novel form face the
liability of engaging in activities still lacking recognizability and legitimation in their
larger social environment (Aldrich & Fiol, 1994). While the liability of newness has been
confirmed frequently for commercial enterprises, research suggests that the
34
organizational world of nonprofit organizations appears to be more benign towards
newcomers (Brüderl & Schüssler, 1990).
The conditions experienced by organizational cohorts at the time of founding may
also play a role in determining if their newness becomes a liability. The concept of
density delay, for example, describes how population entrants "born" into crowded,
densely populated niches face higher mortality risks throughout their existence (Carroll &
Hannan, 1989a, 1989b; Delacroix, Swaminathan, & Solt, 1989). Conversely, in his
comparison study of survival among breweries in the U.S. and newspapers in Argentine,
Swaminathan (1996) demonstrated that organizations founded in adverse, resource-scarce
periods did suffer from the liability of newness at the beginning, but showed higher
resilience than their peers during later life stages. He called this phenomenon "trial-by
fire" (Swaminathan, 1996, p. 1352) and raised intriguing questions about its implications
for the protection of nascent industries through governmental intervention.
In summary, organizational ecologists have discovered many important processes
through which competitive densities influence the survival of organizations. However,
density effects are not easy to capture as they operate on multiple levels. Thus, the
determination if they should be evaluated on the local, metropolitan, state, or national
levels can be challenging (Lomi & Larsen, 1996). Theoretically, the density dependence
perspective has to be also sensitive to the issue of a population's mass (Carroll, 1997),
"which can be thought of as density weighted by the sizes of all organizations in the
population" (Hannan & Carroll, 1992, p. 191). The introduction of population mass is an
attempt to take the size of organizations into account when they populate a resource
35
space, but this concept has not diffused as widely as the basic density model, which
enjoys broad acceptance within the field of organizational ecology.
Legitimation Dynamics
Community ecology has long adopted the neoinstitutionalist concept of
legitimacy as an important cultural resource for the sustainability of new ways of
organizing. Carroll (1997) notes that "social legitimation and diffuse competition" (p.
125) are the two main propellers behind the long-term evolution of organizational forms.
Legitimation processes are associated with organizational forms, not with the populations
adhering to them (Delacroix, et al., 1989), and members of new populations are believed
to work together to establish recognition for their form. The establishment of form
legitimacy is considered necessary for the successful exploitation of resources
(Delacroix, et al., 1989). Legitimacy can be differentiated along two major dimensions.
Aldrich and Baker (2001) describe cognitive legitimacy as "the acceptance of knowledge
about a kind of venture as a taken for granted feature of the environment" and contrast it
with sociopolitical legitimacy, "the acceptance by key stakeholders, the general public,
key opinion leaders, and government officials of a new venture as appropriate and right"
(p. 211). Thus, cognitive legitimacy describes the ease with which an organizational form
is identifiable and recognizable by members of other populations in a community.
Corresponding to this distinction, Hannan (1997) refers to prescriptive versus constitutive
form legitimation (p. 197):
An organizational form is legitimated in the prescriptive sense when its ostensible
purposes, structures, and observable routines conform to a social system's formal
rules. A form is legitimated (or institutionalized) in a constitutive sense when it
has the status of a taken-for-granted social fact. (p. 198)
36
Sociopolitical legitimacy can be thought of as a precursor for extracting resources
from an organizational environment as it mobilizes stakeholders to act more favorably
towards forms considered legitimate. As has been mentioned before, there is empirical
evidence that the founding and survival rates of populations are boosted if they adhere to
forms that are accepted by their social environments. Tucker, Singh, and Meinhard
(1988), for example, were able to demonstrate this relationship in their study on
voluntary social service organizations. The relative importance of sociopolitical
legitimacy varies between industries. In some organizational communities, regulatory
environments are very stable and predictable. Under these circumstances, shifts in
population composition and size over time are captured best with models that are
attentive to density fluctuations. In other industries, discontinuities in the "policy regime"
(Dobbin & Dowd, 1997, p. 507) may override density effects, as has been shown by
Dobbin and Dowd (1997) in their study on railroad foundings in the state of
Massachusetts.
The status of being taken for granted within a social environment is considered
advantageous for survival as it removes ambiguities and streamlines interactions. Thus,
populations adhering to a cognitively legitimized form are able to go about their business
with higher efficiency and reduced interaction costs. In the context of cognitive
legitimacy, Aldrich and Baker (2001) discuss the role of auxiliary populations in the
process of increasing form recognition between populations. By bestowing credentials
onto members of populations within the community, such external arbiters occupy
important structural positions in the struggle for attention and recognition.
37
Legitimacy considerations enter into ecological modeling in a very
straightforward fashion. Organizational population models are traditionally based on the
assumption that a form’s cognitive legitimacy automatically increases with the rising
density of organizations exhibiting that form. Density-dependent processes are based on
the assumption that population entries are "directly proportional to constitutive
legitimation, and inversely proportional to competition" (Dobrev, 2001, p. 423). Thus,
legitimation is built into density dependence processes given that "as the number of
organizations using a particular blueprint increases, the blueprint becomes a legitimated
organizational form" (McKendrick & Carroll, 2001, p. 664). Rather than measuring the
two types of legitimation directly, organizational ecologists have relied on density
fluctuations as evidence for legitimation processes at work. This practice has produced
criticism towards the density-dependence model as mere organization counts are
considered an insufficient indicator, they constitute an indirect legitimacy measure, and
they emphasize diffuse competition when measured on the population level (Carroll,
1997).
Despite of these weaknesses, ecological modeling has created interesting insights
in terms of the connections between form legitimation and population vital rates.
McKendrick et al. (2003), for example, have demonstrated that an emergent form gains
recognition more swiftly if its adherents are spatially concentrated and if they are
founded specifically for the purpose of the new enterprise rather than switching to it from
other industries. The authors concluded that the rate of legitimation of a novel form
accelerates with the number of organizations pioneering it "de novo" rather than "de alio"
(McKendrick, et al., 2003, p. 68). Further, legitimation dynamics have been found to
38
operate on a larger geographical scale than competitive processes (Bigelow, Carroll,
Seidel, & Tsai, 1997; Carroll, et al., 1996; Hannan, Carroll, Dundon, & Torres, 1995).
Scholars speculate that this extended reach stems from the fact that the legitimation of a
form diffuses more easily across geographical boundaries than competitive effects.
Aldrich and Baker (2001) discuss legitimation processes at the organizational,
population, and community levels and refer to "the media, educational institutions, and
certifying agencies" (p. 224) as populations in key positions to influence them.
Even though some organizations may occupy dominant roles when it comes to
shaping form legitimation processes, legitimacy remains conceptualized by ecologists as
a feature of a form that is not specific to individual organizations. Thus, interesting
tensions arise between organizational identities and reputation-seeking behaviors on the
one hand and the collective population outcome of legitimation-building on the other
hand. While legitimation has already found acceptance in ecological models,
opportunities remain for capturing its effects in more nuanced ways. As will be outlined
in the following chapter, new conceptualizations of organizational forms have spurred
theoretical developments that have shown much promise to address this task.
39
CHAPTER 3: THE EVOLUTION OF ORGANIZATIONAL FORM CONCEPTS
Both ecological and institutional approaches to organizational form
transformation have made valuable contributions to the scholarly discourse on form
definition. They have also spawned longstanding and vibrant empirical research
traditions. The next paragraphs provide a brief overview of the content covered in this
chapter. Romanelli (1991) notes that "the concept of organizational form refers to those
characteristics of an organization that identify it as a distinct entity and, at the same time,
classify it as a member of a group of similar organizations" (pp. 81-82). While it appears
immediately intuitive that the concept of an organizational form is useful, if not
necessary, for students of organizational change, it is anything but trivial to determine
what elements should flow into defining a form.
Throughout the years, several approaches to form definitions have emerged that
ultimately gave rise to new theories about how to conceptualize forms. Differing form
definitions are also associated with different mechanisms governing the classification of
organizations into groups. Additionally, form definitions vary in their implications for
conceptualizing form change and in which manner this change is connected with
organizational evolution. It is worthwhile to examine the major definitional approaches to
forms in some detail as all of them have contributed to new insights about the
relationship between the evolution of forms, legitimacy dynamics, and the developmental
trajectory of organizational populations.
As will be explained in detail later, feature- (McKelvey, 1982) and structure-
based (Burt & Talmud, 1993) form definitions emerged first. In recent years, identity-
based form concepts (Hsu & Hannan, 2005) gained in popularity due to a rising
40
scholarly emphasis on cultural aspects of institutions and organizations. In addition to
stimulating many other areas of study, this cultural turn sparked heightened interest in
studying the mental models and action frames of entrepreneurs (Schoonhoven &
Romanelli, 2001), the stratification of industries based on cognitions as "industry
mindsets" (M. E. Phillips, 1994, p. 384), the phenomena of cognitive inertia and path
dependence among organization leaders (Witt, 2000), and the ecology of "cultural forms
occupy[ing] niches in social space" (Popielarz & Neal, 2007, p. 76). The emerging
perspective on organizational forms as entrepreneurial blueprints gave rise to renewed
efforts to gather insights about the ways in which such collectively shared cognitive
frames emerge, stabilize, and change. An identity-based definition of forms also stressed
the socially constructed aspects of organizational forms as created through social
interactions on several levels within organizational communities, which prompted
theorizing about its linkage with managerial strategizing.
Ultimately, the new approaches to forms also led scholars to reconsider existing
definitions of cognitive legitimacy as the "taken-for-grantedness" of a form as a widely
recognized mental model. As a result, a variety of fruitful modifications to existing
ecological models were generated. The purpose of this chapter is to retrace the major
steps through which this process unfolded. It begins with introducing feature- und
structure-based form definitions and lays out the theoretical implications of identity-
based form definitions. The significance of communicative strategies in the context of
form evolution is discussed next, and some of the resulting implications for
organizational legitimation as well as the modeling of ecological dynamics are
summarized.
41
Organizational Forms as Complex Cultural Artifacts
Feature or Trait-Based Form Definitions
In an attempt to systematically account for the form diversity observable in a
variety of communities, early attempts sought to develop an overarching theoretical
framework for a general organizational classification system (McKelvey, 1982; Pólos,
Hannan, & Carroll, 2002). Usually, these taxonomical systems depended upon classifying
organizational forms according to their features and traits (Carroll & Hannan, 2000).
Exploring organizational systematics as a classificatory problem resembling the
determination of categories in biology, McKelvey (1982) notes that "two key questions
for taxonomists are: How many attributes should be considered, and how similar do all
organizations have to be in these attributes?" (p. 25). Once these questions have been
answered, a form definition may be derived to serve as a reference point for gauging
change.
Even if only applied within a specific empirical research context, feature- or trait-
based form definitions are common as they are useful in drawing boundaries around
populations for the purpose of tracking their vital rates. Organizational demographers
purposively stratify populations according to a priori categories that they develop
themselves or borrow from a variety of data sources, such as industry directories,
governmental registries, and encyclopedias (Carroll & Hannan, 2000; Romanelli, 1991).
Often, these categories focus on discernable traits that differentiate organizational
populations in terms of a set of core features. Almost all ecological studies employ
feature-based classification schemes at least when it comes to selecting organizations for
inclusion in research.
42
Drawing on an early classification of organizational characteristics by Hannan
and Freeman (Hannan & Freeman, 1984), Rao and Singh (1999) identify "goals,
authority relations, technologies, and markets" (p. 66) as the core features of an
organization. An investigation of changes in the core versus the peripheral features of an
organizational form allows for an evaluation of the extent of form change. Usher and
Evans (1996), for example, classified gas stations into different forms according to the
services they provided, and then studied how the density of organizations adhering to
these forms in a certain geographical region fluctuated over time. Thus, by treating the
populations as sets of real-world instances of diverse forms they could compare their
relative frequencies over time. They concluded that the overall form composition of gas
stations was undergoing considerable change (Usher & Evans, 1996). Arruñada,
González-Diaz, and Fernández (2004) compared the regulatory environment in Spain
with the one in the United States to explain differences in organizational sizes and
contracting conventions among trucking companies.
Feature-based form speciation events can occur on the basis of selective
replacement on the population level as well as through adaptive modification within the
organization. In this sense, one can think of "speciation as a creative act on the part of
innovating actors in the society" (Lumsden & Singh, 1990, p. 146). Organizational
entrepreneurs tend to borrow liberally from various kinds of existing organizational
forms, thus following a recombinatorial logic (Rao & Singh, 1999) when they create
novel ways of organizing, even if these are genealogically distant from each other. Rao
and Singh’s (1999) typology of form change is somewhat vague, but draws attention to
the fact that almost all new organizational forms are creatively hybridized from those
43
that already exist. It is based on examining to what degrees entrepreneurial organizations
add or delete "primitive entities called organizing elements" (Rao & Singh, 1999, p. 71),
so it seeks to determine if change affects the core of an organization or only its periphery.
If entrepreneurs neither add nor delete such organizational features, they are essentially
engaging in a reproduction of the form that leaves it unchanged. The more elements they
are adding and deleting, the more "radical" is the form divergence observed (Rao &
Singh, 1999, p. 72).
Viewed from a feature-based form perspective, the appearance of a form also
signals that the activities of a new group of similar organizations entering a community
have achieved cognitive legitimacy and are recognized as "a taken for granted feature of
the environment" (Aldrich, 1999, p. 230). Just like a rare mutation in biology may be
restricted to only a few animals or plants, an organizational form innovation can be so
exotic that it fails to be adopted by a critical mass of organizational actors. However, the
proportion of organizations exhibiting the distinct feature must be large enough to attain
discernment by their larger social environment in order for the new form to attain
cognitive legitimacy. This speciation determination in organizational classification differs
from biological phylogenesis in the sense that it essentially assumes a perception process
cumulating in the breaching of a recognition threshold that cannot be pinpointed with
accuracy. Van de Ven and Grazman (1999) point out that this inheritance process for
organizations differs from that of biological species because it is not due to "sexual
propagation of genes and DNA" (p. 186). As biological speciation principles do not hold
when it comes to sociocultural evolution, attempts to replace genes with routines
(Feldman, 2003) or "comps" (McKelvey, 1982, p. 454; Nelson & Winter, 1982) to
44
model inheritance in the organizational context have proven problematic (Baum & Singh,
1994b).
Feature-based approaches to form definitions take the changing prevalence of
various features within a population as a starting point for determining form change.
Figure 3 shows how a genealogical analysis based on size as a single continuous
organizational attribute could translate feature frequencies at three points in time into a
genealogical tree. It showcases – somewhat artificially – how ecological branching would
occur if it solely depended upon the differential distribution of size as a single observable
organizational characteristic in a population.
Figure 3: Construction of a genealogical diagram based on the distribution of
organization size in a population.
45
At time 1, the population of interest is slightly bimodal in its expression of the
feature, but the second mode remains below a certain recognition threshold, so only one
genealogical lineage is recorded. At time 2, an analyst would conclude that as the
distribution of the attribute has significantly shifted towards a larger proportion, a
branching event has occurred, spawning an organization form that is typified by a
distinctly larger size than the original form. At time 3, the distribution of size is
recognizably bimodal, which results in the emergence of two distinct organizational
clusters, so the genealogical diagram registers a speciation event that produces two
diverging organizational populations.
A diagram like this could, for example, be drawn based on examining the size
segregation of grocery stores over the years. They have grown from their "mom and pop"
form to larger retail venues and subsequently diverged into either even larger grocery
stores or small convenience stores. It is important to note that at most points in time, the
existing population of grocers exhibits a much larger variety of sizes, but these do not
give rise to form differentiation unless they are recognized as distinct. Once the feature-
based grouping according to subform has occurred, it is very easy to model the
competitive interrelationships between organizations. A feature overlap coupled with
spatial proximity invariably means that there is a niche overlap between populations that
translates into competitive pressures if the feature is relevant for extracting resources
from the niche. This niche overlap can be quantified for modeling competition, which has
been demonstrated in many settings, e.g. between charitable organizations (Galaskiewicz
& Bielefeld, 1998) and day-care centers (Baum & Singh, 1996) located in the same city.
46
Despite their effectiveness and continuing popularity, feature-based classification
methods have been criticized heavily (Romanelli, 1991). Among other things, scholars
have argued that such classification tends to be inductive, which means that it relies too
heavily on specific instances. Additionally, selecting observable features among
organizational entities draws attention to them as individual entities operating in isolation
and obscures their functional interdependence with others. In any event, the difficulty
remains when it comes to determining which attributes of the many possessed by
organizations should be used for the purpose of distinction (Cerulo, 2002; Karafillidis,
2009). As an alternative to the focus on decomposing organizations into their organizing
elements, structural definitions have emerged which take the linking patterns of
organizations as a point of departure for defining forms.
Structural Form Definitions
With advances in network theory and methodology, structural approaches to form
definition have been enjoying increasing attention. Theorizing on market networks (W. E.
Baker, 1984; White, 1981, 2002) establishes the conceptual linkage between a
population’s structural signature (Contractor, Wasserman, & Faust, 2006; Monge &
Contractor, 2003) and its organizational form. By focusing on similar structural
embeddedness patterns (Burt & Talmud, 1993; Mizruchi, Stearns, & Marquis, 2006;
Uzzi, 1996; Zukin & DiMaggio, 1990) as the basis for a shared organizational form, a
network-based approach examines the role similarities and local network configurations
common to populations of organizations. Structural equivalence in a community network
corresponds to a population’s ecological niche in the sense that "organizations provide
resources for other organizations and that organizational actors are identified through
47
the resources they provide for and consume from other organizations" (Audia, Freeman,
& Reynolds, 2006, p. 386). This sensitivity to the relations forged by organizational types
as indicative of their form emphasizes what organizations are "doing and not merely what
[they] look like" (Popielarz & Neal, 2007, p. 67).
Not all structure-based classification systems are relevant to community
ecologists. Pólos et al. (2002) point out that scientists have not only used the form
concept to identify and differentiate populations for the purpose of studying them.
According to these authors, the concept has also served, synonymously with the term
network architecture, to refer to the structural configuration of organizations without
taking their specific identity into account (Djelic & Ainamo, 1999; Greif, 1996).
Examples for this convention include references to the network form of organization
(Powell, 1990) and discussions that focus on global transformations of dominant
organizing architectures such as the bureaucratic organization (Lewin, Long, & Carroll,
1999; Victor & Stephens, 1994). Obviously, this is a marked departure from how
organization scientists are commonly using the term to study evolutionary processes of
niche formation (Carroll & Hannan, 2000; Hannan, Carroll, & Pólos, 2003). While it is
useful for network scholars to classify diverse types of organizations according to ideal-
type networking patterns, this conceptualization of form does not typically find
application within organizational ecology.
Structural approaches employed for purposes of population identification entail
that organizational forms are defined in terms of the relationships that populations
adhering to them engage in (DiMaggio, 1986). Thus, its "interaction with key
constituents helps define a population’s boundaries" (Aldrich & Baker, 2001, p. 226).
48
Carroll and Hannan (2000) note that "the processes that create and reproduce the
boundaries – social network ties, closed flows of personnel among a set of organizations,
technological discontinuities, social movements articulating the interests of a set of
organizations – are the key to understanding forms" (p. 63).
Based on a structural form definition, speciation events occur when population
members differentiate themselves from their peers in terms of networking patterns and
organizational boundaries. If their network signature changes, their niche changes, and
vice versa. Thus, structural approaches acknowledge that the networking behaviors of
organizational populations serve as recognizable features noted and acted upon by other
social actors. Populations pursuing similar linking strategies are considered structurally
equivalent within their community, which means that due to the overlap of their niche
structures, they will find themselves in competition with each other (Simons & Ingram,
2003). According to structural form definitions, organizations sharing a form compete
with each other as they "have identical suppliers and identical consumers" (Porac,
Thomas, Wilson, Paton, & Kanfer, 1995, p. 203). While these definitions foreground
direct competition between structurally equivalent actors within the community network,
they can also accommodate collaboration between organizations exhibiting similar
embeddedness patterns. However, if organizations occupying similar network positions
strongly depend on extracting resources from their environment through network
linkages, competition with others will rise with the rising number of organizations
sharing the same network signature.
49
Identity-Based Form Definitions
Critics have argued that the previously presented form definitions are unable to
link form theory conceptually with the notion of organizational identities, which has
gained importance among scholars of organizations (Baron, 2004; Hannan, et al., 2006).
More recent literature on organizational identity formation, which is largely informed by
an institutional perspective on organizations, has emphasized that distinguishing between
various types of organizations is an ongoing communicative process in which a variety of
observing audiences within organizational communities participate (Hsu & Hannan,
2005; Romanelli & Khessina, 2005). This line of inquiry is based on social constructivist
aspects of form differentiation and argues that cognitive taxonomies can be derived from
an examination of form communication, i.e., communication relevant to organizational
forms. According to this perspective, which has also informed theorizing about the
taxonomies of communicative genres (Yates & Orlikowski, 2002; Yoshioka, Hermann,
Yates, & Orlikowski, 2001), cognitive classification mechanisms are tied to discursive
interactions within the community about what a form entails (Greve, et al., 2006). Forms
are evolving cultural artifacts which describe whatever aspects of populations are
perceived as salient by their audiences (Hannan, et al., 2006). Further, form identities as
social codes are cognitive categories with which organizations are grouped into classes.
The emerging emphasis on studying organizational identities as tightly related
with forms also renewed connections between theorizing about organizational forms in
particular with conceptualizing the evolution of social forms in general. In the broadest
sense, the emergence of forms as embedded within systems of cultural codes requires the
study of organizational evolution, of which the meaning of organizational forms is an
50
integral part. Tracing the evolution of meaning structures has been a longstanding project
within sociology (Breiger & Mohr, 2004; Durkheim & Mauss, 1903/1963). Durkheim
and Mauss (1903/1963) succinctly summarize the thrust of this inquiry:
Far, then, from man classifying spontaneously and by a sort of natural necessity,
humanity in the beginning lacks the most indispensable conditions for the
classificatory function. Further, it is enough to examine the very idea of
classification to understand that man could not have found its essential elements
in himself. A class is a group of things; and things do not present themselves to
observation grouped in such a way. We may well perceive, more or less vaguely,
their resemblances. But the simple fact of these resemblances is not enough to
explain how we are led to group things which thus resemble each other, to bring
them together in a sort of ideal sphere, enclosed by definite limits, which we call a
class, a species, etc. (pp. 7-8)
Focusing on the cognitive differentiability of organizations that represent one
form from those that represent another, Pólos et al. (2002) propose "that forms be
considered as recognizable patterns that take on rule-like standing and get enforced by
social agents" (p. 89). As a cognitive schema, an organizational form both informs about
an organization’s identity and describes its relationship to other organizations (Clemens,
1996; Swaminathan & Wade, 2001). So the concept of organizational identity is not used
to describe an individual organization, but it denotes the social identity of types of
organizations as perceived by others. In this sense, one can "define identity in terms of
social codes (comprised of sets of social rules and signals) that specify the features that
an organization can legitimately possess. These codes can be enforced by members of the
organization (insiders) or by external actors on which the organization depends for
resources and support (outsiders)" (Carroll & Hannan, 2000, p. 68).
Organizational form identities only emerge relative to the identity positions of
others, which means that populations claim identity positions in a social space already
51
occupied by other populations (Glynn & Abzug, 2002; Ruef, 2000). Rao, Davis and
Ward (2000) observe that "organizations acquire a social identity from the industry to
which they belong, the organizational form they use, and through membership in
accrediting bodies" (p. 222). In the context of the emergence of a whole new
organizational community, it appears that none of these identity markers are available
yet, so entrepreneurs lack these anchors when constructing their organizations' identities.
As has been shown for many novel entrepreneurial ventures (Haveman & Rao, 1997;
Jong, 2006; Lounsbury, et al., 2003; Rindova & Fombrun, 2001), audiences have to
engage in the extended negotiation of meaning and much definitional work until an
industry and its organizational forms are institutionalized and their interrelatedness is
understood widely by relevant audiences. However, even novel organization forms in
previously unknown industries are constructed based on archetypal templates that
represent knowledge about the organization as a social form. In this sense, no
organizational identity has to be crafted anew and is always to a degree borrowing from
existing cultural blueprints for organizing.
Identity-based form definitions have been used to develop further insights about
the relationship between assumed form and organizational outcomes. Hsu (2006b)
suggests that there is a trade-off between assuming a specialist form identity versus a
generalist form identity. She notes that "producers who participate in a variety of market
categories are less likely to instill clear identities in the eyes of relevant audiences relative
to producers with 'perceptually focused' identities" (p. 2006). Both Hsu (2006b) and
McKendrick and Carroll (2001) argue that identity-based organizational forms cannot
52
congeal if entrepreneurs in an emerging market do not adhere to narrow, clearly
recognizable identity niches.
Form identities translate into cognitive templates that differ in their persistence
(Rosa & Porac, 2002), which has been shown empirically in the context of the emergence
of product market categories. They may also vary in terms of "permeability, salience,
durability, and visibility" (Lamont & Molnár, 2002, p. 186). According to Baron (2004),
there are an additional range of characteristics with which form identities can be
described. He argues that producers do best when they assume a strong identity, which
combines sharpness (McKendrick & Carroll, 2001), resonance, and authenticity (Baron,
2004). An identity that is sharp and resonates with others is recognizable and prompts
imitations. Frequently, highly visible organizations within an industry will take the lead
in defining the most important aspects of an enterprise of a certain kind and others will
follow its lead, which will result in a relatively homogeneous group of organizations. A
focused identity is narrow and distinct from neighboring identities. However, it may be
difficult to broaden it if it is defined in strong contrast to other identities (Baron, 2004).
An authentic identity "invoke[s] a non-economic logic for action" (Baron, 2004, p. 14),
which is advantageous for organizations as it is usually evaluated favorably by external
audiences.
Scholars of organizational identities emphasize that identity-based form
transformations are risky endeavors (Hannan, et al., 2006). As Hannan et al. (2006)
found, organizations which changed their identity experienced increased failure rates and
their overall performance dropped off markedly. The authors examined if these negative
consequences could be offset in situations when the identity transformation enabled
53
organizations to improve their fit with environmental conditions. However, they
concluded that the accumulation of negative effects associated with identity change
outweighed its potential benefits in all cases (Hannan, et al., 2006).
The Communicative Aspects of Form Evolution
As identity-based form concepts bring cognition, culture, and social interactions
into the foreground of the study of form evolution, communication research is able to
offer some insights as to how these processes may unfold. As "meaning is always under
construction and is created, affirmed, or refuted through communication between human
agents" (Brummans, et al., 2008, p. 26), form evolution can be studied as an ongoing
discursive construction process with methods developed by communication scholars.
Communication-centered approaches such as organizational rhetoric (Green, Li, &
Nohria, 2009; Sillince & Suddaby, 2008) provide valuable concepts for tracking form
stabilization. Communication research also offers many conceptual tools for dealing with
the notion of form communication as unfolding between various audiences. It provides
particularly well-developed concepts to examine the special role of media organizations,
which make up an important group among these audiences.
Neoinstitutional social movement theories point to the importance of movement
discourse for constructing the boundaries that demarcate legitimate forms of organizing
(Hensmans, 2003). Throughout the years, communication scholarship has synthesized
many ideas about framing from the study of social movements, and the notion of
organizational forms as discursive constructions resonates well with the literature on
framing (R. J. Alexander, 2009; Benford & Snow, 2000; Clemens, 1996; Gray, 2003;
Kennedy & Fiss, 2009; McAdam, McCarthy, & Zald, 1996; Perretti, Negro, & Lomi,
54
2008; Scheufele, 1999). According to the "cultural-frame institutional perspective [sic],"
which holds that "new organizational forms arise when actors with sufficient resources
see in them an opportunity to realize interests that they value highly" (Rao, 1998, p. 915),
the active role of participants in form negotiations deserves scholarly attention.
If new forms are linked with organizational identities and are viewed as actively
introduced and championed by form entrepreneurs (Rao, Monin, & Durand, 2003), the
communicative processes by which diverse groups of form activists compete for
promoting differing frames (Rao, 2009; Rao, et al., 2000) can be elucidated from a
communication perspective. Institutional entrepreneurs carve out niches for new forms
"by defining opportunity, identifying distinctive resources, and prying them away from
existing uses" (Rao & Kenney, 2008, p. 353). An investigation of the communicative
aspects of forms offers an opportunity of not only studying influential communicators
with the discursive power to produce entirely new forms (Abrahamson & Fairchild, 2001;
Aldrich & Ruef, 2006; Lounsbury & Glynn, 2001; Rindova & Fombrun, 2001), but it
also allows for detecting a variety of individual communication strategies with which
organizations navigate within a social space stratified by identity niches (W. P. Barnett &
Woywode, 2004).
Form Stabilization as a Discursive Process
The evolution of forms as cultural templates unfolds through ongoing negotiations
within and between audiences and can be appropriately examined within a discourse-
centered framework (Maguire & Hardy, 2009; N. Phillips, Lawrence, & Hardy, 2004).
Thus, scholars studying the discursive construction of institutions and organizations
recommend that the data used for examining form evolution include discourse in some
55
form (DiMaggio, 1991; Porac, et al., 2002; Rosa, Porac, Runser-Spanjol, & Saxon,
1999). As D. J. Phillips et al. (2004) note, "we argue that institutions are constituted by
the structured collections of texts that exist in a particular field and that produce the
social categories and norms that shape the understandings and behaviors of actors" (p.
638). The coherence of the resulting categories can also be derived from the text that
serves as evidence for discourse about the entities classified according to them.
Ideally, an analysis of form evolution attends to all audiences actively
participating in discourse about it. As "audience reception establishes identity"
(Zuckerman & Kim, 2003, p. 30) and external audiences are influential evaluators of
form emergence (Perretti, Negro, & Lomi, 2008), the first task an analyst has to
undertake is to determine which groups make up the audiences most prominently
involved in the process. It is noteworthy that external audiences will not always base their
differentiation of organizational forms upon core features considered important by trait-
based definitions of organizational form (Baron, 2004; Hannan, et al., 2006).
Identity-based approaches are so powerful as they do not specify a priori which
dimensions will prove most salient for audience when it comes to the task of classifying
organizations according to form. As they can accommodate both feature- and structure-
based form dimensions in addition to those related to identities, they are the most
extensive approach to the definition of forms. Communication about a form becomes
essential for audiences to derive information about what a group of organizations has in
common (Hsu & Hannan, 2005), and these commonalities may matter to some
constituencies, but not to others. While one type of audience may foreground the
organizing elements typical for a population, another may emphasize the way the
56
population is embedded in its social environment, and yet another may focus on its
identity. Therefore, examinations of organizational forms should be sensitive to
audiences negotiating all of the three form aspects.
Within an industrial community, typical audience groups are made up of
customers, suppliers, producers, rating agencies, and members of the media. External
evaluators occupy a particularly influential position within the process of form
negotiation as they take over the task of framing the activities of organizations for a wide
audience (Hsu & Podolny, 2005). The stabilization of categories that enable audiences to
classify organizations and their products is furthered by media discourse (Kennedy, 2008)
as media institutions function as intermediaries that synthesize form discourse for others
(Zuckerman, 1999). Journalistic accounts about organizational forms are a vital
constituent for audiences to develop perceptions about its salient characteristics
(Romanelli & Khessina, 2005). NGOs and other organizations particularly influenced by
social approval depend upon "public opinion, regulators, accrediting bodies, professional
associations, and legislatures" (Galaskiewicz & Bielefeld, 1998, p. 10) as resources for
garnering sociopolitical legitimacy.
Audiences come to expect certain features associated with types of organizations,
and if these expectations are violated, they react unfavorably (Hannan, et al., 2006). This
negative evaluation of norm transgression is a central sociological concept (Dahrendorf,
1964; Mead, 1934; Merton, 1957), which has also been examined in the context of
interpersonal communication (Afifi & Metts, 1998; Burgoon & Hale, 1988).
Organization theorists further assume that such form expectation violations bear negative
consequences for the performance of organizations adhering to the form as they will
57
result in their devaluation by audiences (Hannan, 2005). Referring to Zuckerman's (1999)
study on securities analysts, Popielarz and Neal (2007) explain that "violating
expectations can raise the likelihood of failure through either neglect or active withdrawal
of support (Zuckerman 1999)" (p. 71).
A communication-centered perspective on organizational forms highlights that as
discourse about a form changes, so will the cognitive classification systems that are
emerging from this discourse. As described by Hsu (2006a) and tested empirically in her
research, "the language an individual uses to interpret and evaluate an object reflects his
cognitive representation of the category within which the object resides" (p. 475). If the
cognitive schemata people hold about social forms differ from each other, firm categories
are unlikely to stabilize. In other words, forms stabilize as cultural entrepreneurs carve
out discursive niches for them within a symbolic space (Weyer, 1989). Abrahamson and
Fairchild (1999, 2001) studied discourse over time to demonstrate the somewhat transient
nature of semantic category evolution and were able to connect their findings to
organizational practices. They found a coevolutionary relationship between discourse
about quality circles in the management literature and the adoption of the practice in
organizations. Categories do not always have to stabilize; they can also remain
ambiguous and the boundaries between forms may remain blurred (Lounsbury & Rao,
2004). The emergence of consensus is difficult to predict and to measure as there might
be intermittent periods of dissensus preventing the stabilization process (Porac, et al.,
2002).
Contemporary textual analyses of form communication employ sophisticated
computerized tools to provide measures of cognitive coherence (2006a, 2006b; Hsu &
58
Podolny, 2005). These analyses are, for example, based on an examination of adjectives
and the synonym overlaps of adjective pairs that are compared. Employing such a
technique, Hsu (2006a) demonstrated that movie genres differ in terms of the
classificatory clarity and level of agreement with which critics review them. She argues
that as critics have had difficulties to establish clear mental schemas when evaluating
movies in genres such as a western, these products have suffered in terms of their market
valuation. Her findings support the assumption of form theorists that entities placed in
clearly demarcated categories perform better.
The Role of Organizational Labeling Strategies for Form Evolution
As mentioned previously, communicative approaches informed by social
movement theorizing emphasize that audience members play an active role in the form
negotiation process. Case studies abound in which influential form entrepreneurs
strategized to bring about the emergence of a new form (Rao, 1998, 2009; Rao & Giorgi,
2006; Rao, et al., 2003). Munir and Phillips (2005), for example, used a discourse
approach to study how Kodak assumed the role of an institutional entrepreneur to bring
about institutional change, fundamentally changing the role of photography as a
technology in people's lives. Recent studies suggest that whole industries can appear if
cultural entrepreneurs fueled by social movements use symbolic resources strategically to
carve out new niches for organizational populations and whole communities associated
with them. Examples include the slow-food-movement inspired market for grass-fed
meat (K. Weber, et al., 2008), the emergence of the wind energy market (Sine & Lee,
2009), and the U.S. recycling industry (Lounsbury, et al., 2003).
59
It should be noted that form entrepreneurship constitutes a notable and
exceptional feat among organizations adept at spurring collective action. In most
circumstances, form entrepreneurs are powerful and sizeable members of organizational
communities with privileged access to supporters and broad support from other
audiences. However, the strategies of individual organizations navigating the identity
space available to them to the best of their abilities differ from the methods employed by
such dominant cultural innovators. Rather than shaping the evolution of the discourse
through their own actions, these organizations are attempting to claim niche space within
a symbolic resource environment that is largely beyond their immediate influence. As
members of a population of organizations engaged in the same activities, their efforts can
be considered from the lens of community ecology. Before their strategic options are
discussed, it may be worthwhile to consider how individual strategy-making fits into the
organizational ecology framework.
Strategic management perspectives do not have to be at odds with ecological
models emphasizing natural selection pressures (Freeman, 1995). Drawing on Child’s
(1972) theorizing about strategic choice, many scholars (cf. Aldrich & Pfeffer, 1976;
Lewin, et al., 1999) have stressed the possibility for managers to engage in strategic
action despite downward constraints imposed by an individual population’s environment.
Astley and Fombrun (1983) define organizational strategizing as a "mechanism for
voluntaristic adaptation at the individual level" (p. 579). Lower-level adaptive moves do
not preclude higher-level selection as there is always a degree of mismatch between
managerial actions and changes resulting from them (Hannan & Freeman, 1984). Also,
the capacity for engaging in strategic realignment can be viewed as an organizational
60
characteristic "differentially distributed within a population of organizations"
(Burgelman, 1990, p. 173), which as such constitutes a feature that may be preferentially
selected in certain environments. Thus, an ecological framework can easily accommodate
organizational strategies as lower-level selection processes based on the highly
contextualized decision-making of managers (Fombrun, 1988).
Managerial strategies affect evolutionary processes on multiple levels. First,
managers are the ones that actively select from a variety of internal organizing elements
such as routines. Second, their actions also impact their organizations' selection
environments, for example when they engage in collective action to influence industry
policy. Third, their perceptions always mediate environmental conditions in the sense that
they modify their enterprise based on their subjective evaluation of environmental
selection pressures (Calori, Johnson, & Sarnin, 1992; de Chernatony, Daniels, &
Johnson, 1993; Scott, 1981). Research suggests that managers' cognitions about the
structure of their competitive environment (Howard, 1994; M. E. Phillips, 1994; Porac,
Thomas, & Baden-Fuller, 1989; Reger & Huff, 1993) are just as consequential for the
evolution of an industry as competitive pressures due to the actual resource niche overlap
between organizations.
The purposive managerial practice of recombining organizing elements as
described by Rao and Singh (1999) indicates that entrepreneurs may in fact play an
important role in the community negotiation of new organizational forms. In evolutionary
terms, the large pool of organizing practices available to managers produces the variation
necessary to create new forms. The creative combination of these elements gives rise to
novel forms (Rao & Singh, 1999). With the exception of well-connected and
61
resourceful large form entrepreneurs, individual organizations are only able to effect
significant transformations of their organizational form to a limited degree. However, all
of them are able to position themselves strategically within the symbolic environment
provided by interrelated social forms. One of the means of doing so revolves around the
manipulation of the organization name as an identity label, which constitutes the focus of
the current investigation.
Scholarly attention to labeling practices (Ashforth & Humphrey, 1997) is not a
recent phenomenon. Lamont and Molnár (2002) suggest that "there is a long tradition of
research, directly inspired by the Chicago School of community studies, that concerns the
internal symbolic boundaries of communities and largely emphasizes labeling and
categorization" (Lamont & Molnár, 2002, p. 181). Organization names function as
significant artifacts (Csikszentmihalyi & Rochberg-Halton, 1981) as "they signal
categories of meaning (Brown, 1958), sorting organizations into equivalent and
nonequivalent sets" (Glynn & Marquis, 2006, p. 226). Organizations can use their names
to communicate to external audiences which form identity they claim (Hartelius &
Browning, 2008; Kogut & Zander, 1996; Pratt & Rafaeli, 2006; Strauss, 1997), and their
strategies may prove consequential for their overall survival.
In a series of studies on organizational name choices, Glynn and her colleagues
(Glynn, 2002; Glynn & Abzug, 2002; Glynn & Marquis, 2004, 2006, 2007) were able to
show that audiences prefer recognizable organization names that are in line with current
institutional norms for naming. In general, ambiguous labels were found to be associated
with lower organizational performance. The authors also noted that individuals
evaluating organization labels vary both in their accuracy of determining industry
62
membership based on organization names as well as their preferences for legitimate
names (Glynn & Marquis, 2006). Specifically, Glynn and Marquis (2006) found that
"individuals who believe that an organization's name should reflect industry norms will
choose new organizational names that conform to institutionalized practices" (p. 133).
Glynn and Abzug (2002) also discovered that naming standards change over time and
vary across communities, which means that in some industries, labels are more
ambiguous than those ubiquitous in other markets. Firms tended to follow their peers and
adopted name types that appeared most legitimate within a given sociopolitical
environment, a preference the authors termed symbolic isomorphism (Glynn & Abzug,
2002). Most importantly, Glynn and Marquis (2004) showed with their research on
"dot.com" labels that under unfavorable sociopolitical legitimacy conditions,
organizations were quick to switch names in order to counteract the negative effects of an
illegitimate label. While signaling the organizational identity of an Internet company was
advantageous before the crash, their naming transitions suggest that it was perceived as a
symbolic liability afterwards.
All these findings demonstrate that identity strategies (Allaire & Wolf, 2004)
within the reach of individual organizations may have genuine survival implications.
Identity positioning in terms of labeling should be understood as a self-representational
external strategy of organizations rather than a fundamental change of their core
missions. Such changes, as was discussed previously, are located inside of organizations
and carry the risk of disrupting their operations so severely that recovery may become
difficult. In contrast, organizations can strategically exploit the fact that "identity gets
63
conferred by outsiders" (Hannan, 2005, p. 60) and may thus be able to preserve their
internal identity while catering to the (perceived) demands of external audiences.
Some Implications for Organizational Ecology Research
Novel specifications of organizational forms as evolving cognitive templates have
impacted ecological modeling in multiple ways, mostly as they necessitate making
revisions to the fashion in which such models capture cognitive legitimacy. The resulting
corpus of organizational ecology research examining the impact of cultural processes
shows great promise for a successful combination of insights from ecology and
institutionalism (Baum & Powell, 1995; Fombrun, 1988). Recent literature indicates that
there is substantial overlap and convergence between the two perspectives (cf. Amburgey
& Singh, 2002; Greve, 2002; Rao, 2002; Strang & Sine, 2002).
Originally, population ecology faced criticism from institutionalists for not
providing direct evidence that legitimation processes were at work (Carroll, 1997).
Ecological models rarely include the direct observation of legitimation as one of the
drivers of density-dependent processes. The sheer growth in the number of organizations
exhibiting a similar organizational form will, according to this perspective, lead to
increased cognitive (and sociocultural) legitimacy necessary for the survival of the
population as a whole (Aldrich, 1999; Dobrev, 2001). As argued by Suchman et al.
(2001), ecological approaches have tended to infer legitimation processes rather than
observing them "by examining the actual structure of cultural discourses or the actual
transmission of institutionalized models" (p. 358). Hannan et al. (2007) suggest "to place
social codes at the forefront of ecological analysis" (p. 21), thus calling for a systematic
inclusion of form legitimation processes into organizational ecology.
64
Novel form definitions highlight that perceptions about the density of entrepreneurs
in organizational communities may matter as much as the actual densities of their
organizational populations. Therefore, recent conceptualizations of cognitive legitimacy
include an emphasis on such perceptive processes. The explicit consideration of
communication about organizational forms as an important evolutionary trajectory
suggests that competitive intensity can be derived from perceived population densities in
community discourse. This idea is articulated and tested in Kennedy’s (2008)
investigation of the evolution of the computer workstation industry. His research
exemplifies how the analysis of industry discourse can aid in the determination of the
symbolic density of producers in a market.
In comparable fashion, McKendrick and Carroll (2001) "propose that a
legitimated organizational form emanates from the density of focused producers in a
market, rather than total density. The core idea is that the identity of a form derives from
the aggregated identities of individual organizations" (p. 676). Similar to W. P. Barnett
and Woywode (2004), Ruef (2000) notes that "within the community ecology approach,
then, a carrying capacity is defined as the maximum number of organizations having
some identity (potential or realized) that can be supported by the environment at a
particular point in time" (p. 678). As McKendrick and Carroll (2001) point out, the
temporal ordering of form stabilization as outlined in density-dependent processes may
not always hold. Their study demonstrates that the activities of organizations in the
market of disk arrays never congealed into a distinctly recognizable organizational form,
even after the market became saturated and competition between organizations increased.
65
Apart from revising their measurements for cognitive legitimacy, ecologists have
also been prompted to incorporate cultural aspects, cognitions, and communication
processes more explicitly into their models. Most importantly, they are starting to
conceptualize legitimation processes as products of an ongoing negotiation process
unfolding between collective observers (Barreto & Baden-Fuller, 2006). An increasing
number of studies include the analysis of linguistic constructions and discourse in order
to find evidence for institutionalization processes (Green, et al., 2009). In addition,
organizational ecologists are to be credited with research that established the connection
between strategic identity positioning and organizational survival. Code violations in the
sense that organizations defy easy classification or deviate from expectations associated
with their form have been consistently shown to have negative effects (Dobrev, Ozdemir,
& Teo, 2006; Hannan, et al., 2007).
New factors underlying the competition between populations have also been
found as the definition of friend and foe within organizational communities can be
conceptualized as a cognitive classificatory process. Actor-centered definitions of
competition take identity signaling in organizational communities into account, which
heightens the importance of research on rivalry construction (Porac & Thomas, 1990;
Porac, et al., 1989; Porac, et al., 1995; Porac, Wade, & Pollock, 1999) for the
development of sound ecological models.
Another contribution of novel form conceptualizations can be found in their
emphasis on the active strategic role of organizational entrepreneurs. Selected form
entrepreneurs instigate collective processes within communities to establish cognitive
legitimacy for a form. All organizational decision-makers exploit existing symbolic
66
resource niches in their communities by designing symbolic artifacts such as organization
labels in order to position themselves with form identities that exhibit adequate cognitive
legitimacy (A. J. Baker, 1991). Such strategic behaviors on the level of organizations can
be fruitfully included in ecological models examining the survival implications of these
strategic actions. Further, as studies on organizational labeling have shown, there is great
potential in considering the influence of sociopolitical and cognitive legitimacy aspects
separately, as "each concept reflects an important and distinctive phenomenon, and a
concern for theoretical precision dictates that the two should not be conflated"
(Jepperson, 1991, p. 198). The following chapters illustrate that there is yet some
potential for more theorizing about how the different dimensions of legitimacy interact
and create symbolic space that organizations can navigate in strategic fashion.
67
CHAPTER 4: AGGRESSIVE MIMICRY AS A COEVOLUTIONARY STRATEGY
In the life sciences, mimicry has been described as a successful coevolutionary
strategy for many different species as they take advantage of their very similarity to other
species in order to lure prey or evade predators (Wickler, 1968). An organization
engaging in aggressive mimicry is the proverbial wolf in sheep's clothing, a strategist
persuading an audience to accept its incorrect membership claims to a form category. The
concept of aggressive mimicry is useful for examining the conditions under which
deceptive identity strategies may be employed and it also aids in tracing possible effects
resulting from its usage.
As aggressive mimics thrive on form ambiguity and proliferate under conditions
of low cognitive legitimacy, their strategies test the limits of current ecological models,
which account for the evolutionary success of populations by pointing to their form
legitimation. An introduction of the notion of aggressive mimicry into theorizing about
organizations may engender insights about a range of innovative, but understudied form
identity strategies pursued by both commercial and nonprofit organizations. In the next
section, the concept is presented in more detail and examples from the world of
organizations are showcased. Conditions for its occurrence are outlined and the
implications of aggressive mimicry for organizational ecology modeling are discussed.
The Concept of Aggressive Mimicry
Originally, the notion of aggressive mimicry was introduced into biology by W.
G. and E. G. Peckham. They used it to describe cases in which a mimic population
employs its resemblance with a model population for predatory purposes (Wickler,
1968). These researchers describe the genera of the Attidae, which refer to tiny spiders
68
that look like ants and even move like them. This resemblance enables them to approach
their unsuspecting prey with ease (Peckham & Peckham, 1892). Mimicry behaviors in
general capture a range of phenomena that occur everywhere, as noted by Sundie et al.
(2006):
Humans are not the only species that fall prey to unscrupulous influence attempts,
such as false cues to similarity and familiarity. Mimicry and the false signaling of
group membership are employed by a wide variety of animals, insects, and plants
to further their own self-interests in survival and reproduction. (p. 311)
A lively scholarly debate about mimicry phenomena ensued after the British
biologist Henry Bates wrote in 1862 about the difficulties he encountered in
discriminating between separate species of butterflies. Noticing with surprise that they all
appeared to look alike, he remarked that he "was never able to distinguish the Leptalides
from the species they imitated, although they belong to a family totally different in
structure and metamorphosis" (Bates, 1862, p. 504). He concluded that there had to be
some evolutionary benefit for members of one species to approximate the appearance of
another (Wickler, 1968). Since then, many different types of mimicry have been
discovered, which usually can be distinguished according to two main functions: (1)
Protection from predators and (2) concealment of predatory activity (Moynihan, 1968;
Pasteur, 1982; Ruxton, et al., 2004). Both types will be examined briefly in the next
paragraphs.
Protective mimicry includes Batesian and Müllerian mimicry and occurs when the
signal-receivers within the mimicry system belong to a population of predators and the
dupes seek to evade becoming their prey. By approximating the appearance of the model
population, mimics gain protective benefits. In Batesian mimicry, the imitation of a
69
harmful model by a harmless mimic also affects the fitness of the model negatively
(Kloock, 2000). Many kinds of palatable insects, for example, engage in Batesian
mimicry when they appropriate the black and yellow warning colors of wasps, which
signal unpalatability to predators (Ruxton, et al., 2004). In Müllerian mimicry, two
harmful species mimic each other's warning signs symbiotically for protective purposes
(Wickler, 1968). Members of both populations derive advantages from using the same
warning patterns to inform predators that it is unprofitable to hunt them as their enemies
will learn more quickly to stay away when the signaling patterns are shared.
Aggressive mimicry constitutes behavior associated with catching prey or
outperforming a competitor, which differentiates it from protective types of mimicry. An
early definition of aggressive mimicry was forwarded by Poulton in 1890 (Brower,
Brower, & Westcott, 1960), who used the concept to describe cases "in which one species
resembles another unrelated one in order to approach it the better without exciting
suspicion for various detrimental purposes" (p. 343). The zone-tailed hawk is a predator
exhibiting aggressive mimicry as it tends to get confused by small animals when it
purposively hides in the midst of vultures (Wickler, 1968). Other examples described in
the literature include fish, plants, insects, and snakes (Jackson & Wilcox, 1990; Kloock,
2000; Lloyd, 1984; Pietsch & Grobecker, 1978; Sazima, 1977).
It is important to keep in mind that aggressive mimicry differs from camouflage,
which occurs when detection is avoided by imitating the appearance of inanimate objects
(Kloock, 2000). Camouflage can be easily differentiated from mimicry as it is not based
on mimicking a member of a recognizable other species, but it consists of evading
attention altogether through blending with the background. The purpose for
70
camouflage can be either defensive or aggressive (Pasteur, 1982). If a predator tried to
resemble a rock or a leaf in order to hide its presence, the behavior would be classified as
camouflage rather than as aggressive mimicry. Thus, the difference between camouflage
and aggressive mimicry is based on the distinction between eluding notice by blending
into the environment versus appearing as a species that elicits a specific response from
the signal-receiver.
In a classical aggressive mimicry system, three populations can be differentiated:
Models, mimics, and dupes. Both models as well as mimics display signals that are being
interpreted by a population of signal-receivers, the dupes (Wickler, 1968). Model and
mimic populations represent two species "which transmit the same signal and have at
least one signal-receiver . . . in common which reacts similarly to both" (Wickler, 1968,
p. 239). The evolutionary pressure underlying mimicry evolution is exerted by the signal-
receiver population, which compares the model and the mimic (Wickler, 1968).
Resemblance is assessed through sensory means and amounts, among other things, to
similarities in odor, coloring, behavior, or sounds (Pasteur, 1982, p. 176). Due to their
huge variability, documented mimicry phenomena are difficult to classify, and it is not
always possible to reduce the number of species in a mimicry model to only three
(Wickler, 1968). On the other hand, cases of intraspecies mimicry are known in which
model and dupe belong to the same species (Lloyd, 1965), which reduces the number of
populations in the system to a set of two. Some approaches classify only those cases as
aggressive mimicry in which actions of the aggressive mimic reduce the fitness of the
dupe (Kloock, 2000). However, for the purposes of the current study, a broader definition
of the term will be followed in which this requirement is relaxed. At a minimum, the
71
imitation of the model through the mimic has negative fitness effects for the model as its
"friendly" relationship with the dupe suffers.
Carpenter (1933) noted that "mimicry only affects external appearance" (p. 31)
and proposed three rules: a) Mimicry requires that mimics settle in close geographic
proximity to their models. b) The similarities between mimics and their models may not
be due to isomorphic evolutionary pressures exerted by that shared environment. c)
Mimicry patterns emerge from a coevolutionary relationship between mimics and their
models. Based on Brower et al. (1960), a fourth rule could be added: The number of
models must exceed the number of mimics. Underlying this notion is the assumption that
if the number of impostors surpasses the number of the imitated, the deceptive ploy is
detected and the mimicry system becomes unstable.
Aggressive Mimicry in Organizational Communities
Recently, renewed scholarly interest in human deception has emerged, both
among communication researchers as well as in organization studies (Burgoon, Blair, &
Strom, 2008; Goffman, 1986; Kubon-Gilke, 2005; Moss, 1996). Gerschlager (2005)
observes that deception occurs everywhere in markets, is very effective, and the dupes
may even profit from its occurrence. Aggressive mimicry employs deception in the sense
that it requires mimics to make incorrect membership claims towards an organizational
form. It exploits the fact that "the boundaries of markets are . . . ambiguous and must be
inferred" (Porac, et al., 1995, p. 205). Form deception is often judged negatively by at
least some audiences as it involves violating expectations about membership in social
codes. However, notions of what counts as deceptive are anchored in social conventions
and can be transient and vague (Thijs, 2002). For the purposes of determining the
72
evolutionary effectiveness of mimicry behaviors, the degree to which mimicry occurs
intentionally is not of primary importance to analysts even though it may be of ethical
relevance.
In the context of organizational ecology, resemblances between mimic and model
are mostly based on their behavioral similarities and their likeness in superficial
appearance, not on the physically observable phenotype of a species as in animal biology.
Moynihan (1968) emphasizes that a mimic’s similarities to the model "do not seem to be
adaptations to induce social reactions between the two species themselves, but rather to
facilitate similar reactions to both by other species which overlap both" (p. 321). This
means that surface resemblances are sufficient as long as they enable mimics to relate to
dupes in a similar fashion as models. An organizational mimic does not have to imitate
the inner organizational workings of its model in intricate detail to achieve the desired
effect.
As outlined previously, organizations are believed to routinely engage in identity
strategies with which they modify their symbolic representation (Rao, et al., 2003) in
order to elicit a positive evaluation from observing audiences (Baron, 2004). Usually,
these self-presentational efforts entail striving for clear and easily classifiable form
identities, which means that populations struggle to set themselves apart from those that
have established neighboring niches in identity space. W. P. Barnett and Woywode
(2004) illustrate this tendency in their study on Austrian newspapers by demonstrating
that the fiercest competition ensued between news organizations that had ideologically
similar positions, e.g. the newspapers at the left-to-centrist position were found to
compete most strongly with those located at the center-left.
73
In contrast, aggressive mimicry relies on the communicative value of ambiguity
(Davis, 1991). A population pursuing aggressive mimicry does not seek out an identity
niche that overlaps as little as possible with others’ niches (as it happens in saturated
industries where resource partitioning is starting to occur), but it attempts to resemble its
model so closely that interpopulation competition is maximized. So instead of trying to
distinguish one’s own population and appear different from others that might be
perceived as similar, organizational entrepreneurs select and articulate aspects
preferentially that resemble those of populations identified as close competitors. Over
time, interpopulation similarities increase, because selection pressures towards
convergence are beneficial for the mimics’ survival (Turner, 1988). According to
biological conceptualizations of aggressive mimicry, such purposive blending processes
are initiated by the aggressive mimicry strategists and may negatively affect the
livelihood of the model as well as the dupe (Wickler, 1968).
In this sense, aggressive mimicry results in a strategic adjustment of the mimics'
niche (Popielarz & Neal, 2007) that proves consequential for the survival chances of the
model population, which is why aggressive mimicry can be regarded as a form of
"predatory competition" (Aldrich & Ruef, 2006, p. 244). Overlaps in niches are usually
associated with different competitive effects, which means that the occupation of the
space by members of one population is affecting members of the other population more
strongly than vice versa (Baum & Singh, 1994c). Mimicry strategies result in a purposive
blending process initiated by the mimic population due to the lateral extension of its own
resource niche towards its model's niche. This niche expansion affects the overall level of
resources available to the models (Hannan & Freeman, 1989).
74
According to Delacroix et al. (1989), model populations can react to such niche
crowding by seeking to outperform their emerging competitors or they can "step aside"
and move their niche away from the heightened competition. If the mimics respond to
this coevolutionary move by yet another lateral niche extension, an evolutionary arms
race ensues, and a dynamic of consecutive niche-shifting moves (Carroll, et al., 2002)
develops. As illustrated in Figure 4, "when mimics negatively affect their models . . . the
models are expected to 'run away' from the resemblance, starting the classic evolutionary
chase seen in most Mathematical models of Batesian mimicry" (Kloock, 2000, p. 7).
Figure 4: Niche adjustment dynamics between mimic and model populations.
The Spectrum of Mimicry Behaviors
Aggressive mimicry is a strategy that may entail varying degrees of imitation,
ranging from a slight resemblance to familiar organizational forms to a meticulous
reconstruction of a target population’s appearance. In some industries, aggressive
mimicry occurs only within individual product markets. If organizations have taken great
care in constructing and communicating a specific market identity, they may resort to
aggressive mimicry in situations where this very identity curbs their ability to compete in
Mimic Model
Mimicry process
A
Mimic Model
Model flight
B
Mimic Model
Original population
Moved population
C
75
certain market segments. As maintaining their legitimate identity in their original market
is necessary for their long-term survival, organizations may decide to create transitional
pseudonyms for themselves adapted to other market segments (D. J. Phillips & Kim,
2008). Some large mainstream breweries obscure their generalist identities when
competing in the microbrewery niche market. Coors and Miller Brewing, for example,
are "two of three largest brewing firms [which] conceal their identities on specialty
products. Miller Brewing created the fictional name Plank Road Brewery to put on its
labels" (Carroll & Swaminathan, 2000, pp. 727-728). Similarly, contract brewers who
only market specialty beers do not actually produce them in their own facilities, which
delegitimizes them as craft brewers in the microbrewery market. Both contract brewers
and mainstream breweries concentrate their efforts on mimicking the same model
population, which consists of specialty brewers. In the case of breweries, the dupe
population consists of individual consumers who are comparing producer populations
when they make purchasing decisions.
Thijs (2002) documents an interesting early case of product-market mimicry in
his study of the Dutch textile industry in the 16th Century. He describes how textile
makers from other locations routinely engaged in the collective practice of falsely
identifying their products with labels stating that they were manufactured in Brussels,
which was known as a bustling center for tapestry-weaving artisans at that time. Over
time, the status of Brussels-manufactured textiles declined as customers caught on to the
deceptive practices and weavers' guilds in Brussels rallied for outlawing the practice.
Once the significance of Brussels as a location brand had been eroded, other production
places began to function as signals for superior craftsmanship, and the cycle began
76
anew. Tapestries made in Antwerp were sold as Brussels textiles, "narrow fabrics from
Holland and Zeeland were exported to Spain as 'Antwerpse smallekens' [narrow fabrics
from Antwerp]" (Thijs, 2002, p. 129), and "Bruges satin" was offered to customers that
had been woven somewhere else. As noted by Thijs (2002), "the original significance of
the place-name from which a textile came lasted only as long as the given production
center could maintain its monopoly over the manufacture of that product" (p. 129).
In its weak form, mimicry involves exploiting the legitimacy spillover (Kuilman
& Li, 2009) of other organizational forms. It can be thought to occur when corporate
booksellers seek to invoke the atmosphere of a library with the spatial arrangements of
their retail space. Reports about mall-based commercial teeth-whitening businesses that
have been mushrooming in the U.S. emphasize their peculiar resemblance to dentist’s
offices (Witcher, 2008). By mimicking the procedural structure of a visit at a dentist’s
office, these new types of entrepreneurial venues borrow liberally from the legitimation
of dentists as health professionals (Fiol & O'Connor, 2006). Rao and Singh (1999)
observe that it is certainly a legitimate and important aspect of organizational form
invention to borrow and recombine elements of other forms in the process.
Stronger varieties of aggressive mimicry depend on the successful imitation of a
known organizational form that may involve ongoing deception and affect the perceived
legitimation of the target population in negative ways, thus impacting its survival chances
(Kloock, 2000). Among the most common forms are fraudulent investment schemes that
appear attractive and "serious" and seem to be offered by legitimate companies. "Pyramid
schemes" such as the Ponzi (Pressman, 2009; Zuckoff, 2005) exemplify a deceptive
investment form that appears legitimate when encountered by an unsuspecting,
77
relatively inexperienced investor. D. J. Phillips and Kim (2008) revealed that jazz music
producers worried about offending their high-brow clientele by entering the market for
popular jazz during the early 20th Century. Thus, they resorted to the use of pseudonyms
in order to be able to compete in the lucrative and growing market without damaging
their identity as elite producers.
Many examples of strong aggressive mimicry appear to be rather specialized
organizational forms, which attests to the success with which their adherents are able to
avoid identification by large audiences. However, few people are aware of the ubiquity
and large-scale involvement of aggressive mimics in civil society. Particularly in the
health care sector, both "nonprofits-in-disguise" and "for-profits-in-disguise" (Steinberg,
2006, p. 118) are blurring the lines between charity and social agency and are claiming
membership in a different identity group when it satisfies their demand for acquiring
support or financial resources. However, such practices are usually not considered
illegitimate and they do not experience strong sanctioning.
In contrast, interest group scholars use the term "astroturfing" to refer to the
corporate sponsorship of organizations that appear to be genuine grassroots-level
movements (Lyon & Maxwell, 2004). The label of the phenomenon already hints at how
astroturfing works. Artificially created, strategically designed groups pursue astroturfing
when they exhibit all the characteristics of concerned citizens rallying for a cause close to
their hearts. Their model populations consist of grassroots NGOs such as environmental
groups or citizen rights groups. In their case, the dupes are policy makers and other
stakeholder groups, who mistake the impostors for legitimate social movements actors.
Astroturfing can be very effective as it is able to tap into the strong credibility and
78
legitimacy associated with a genuine grassroots movements. If a seemingly bona fide
local nonprofit group with a declared environmentalist mission supports the construction
of a new chemical plant, political decision-making can be influenced forcefully. Astroturf
campaigns appear convincing as they are seemingly based on broad and enthusiastic
support from various constituents, even though this mobilization is staged. Klotz (2007)
refers to the phenomenon aptly with the term "plagiarized participation" (p. 3).
The Coalition to Protect Patients' Rights (CPPR) serves as a recent example for
such an artificially constructed interest group. It was created by the DCI Group, a public
relations company specializing in providing front groups for clients from industry and
politics (Fang, 2009, p. 1161). CPPR appears to be a legitimate social movement actor
challenging the health care reform plans by the Obama administration. As it was not
formed by grassroots activists mobilizing for their cause, but by paid professionals
representing the interests of industrialists, it violates the authenticity assumptions with
which social movement actors are evaluated. In the past, other front groups created by the
DCI Group for clients from the tobacco industry appeared as interest groups concerned
with defending smokers' rights (Fang, 2009). People for the West! represent an older
example of astroturfing. The organization seemed to be an alleged wise-use-movement
activist group, but was actually funded by mining companies (Lyon & Maxwell, 2004).
Wise-use organizations are usually founded by conservative activists seeking to ward off
governmental control of private land ownership. Mining companies profited from
mimicking their population as they shared their interest in keeping governmental
regulations of land ownership at a minimum. Again, policy-makers and other social
movement actors functioned as dupes. During a period in the 1990s, People for the
79
West! was able to expand to more than 120 chapters nationwide before its allegiances
became public and its influence waned. Illustrating another case of astroturfing, Rao et al.
(2000) refer to "identity theft" when they describe how aggressive mimicry was used as a
social movement tactic designed to counter the environmental movement "by lumber
firms who founded organizations with names identical to environmentalist organizations
in the American Northwest" (p. 266).
Carpenter's (1933) three rules of mimicry suggest how mimicry in biology
compares with DiMaggio and Powell’s (1983) concept of mimetic isomorphism. Usually,
mimetic practices describe convergent intrapopulation evolution in which members of the
same population imitate each other due to various reasons (Miner & Raghavan, 1999). In
contrast, three-population mimicry systems entail that members of one population
approximate the form of a competing population through coevolutionary processes. The
mimicking population seeks to emulate the target population’s appearance and behaviors
as an opportunity to engage in similar relationships with a third population.
Aggressive mimicry strategies can ensue if individual mimics copy the
appearance of individual models by direct observation. However, aggressive mimicry in
organizational communities can also derive from mimetic isomorphism (DiMaggio &
Powell, 1983; Haveman, 1993). Mimetic isomorphic pressures originate within a
population when organizations imitate intrapopulation peers they regard as industry
leaders (Milstein, Hart, & York, 2002). If these role models engage in mimicry practice
and their observant peers follow due to isomorphic pressures, the effective outcome is
still some blending of the populations. Thus, in symbolic isomorphism dynamics (Glynn
& Abzug, 2002) individual mimics appropriate the identity of their model and serve as
80
role models that entice other members of their population to follow. Mimetic
isomorphism pressures may then affect the ways in which the mimic population intrudes
into their model's identity niche. In this sense, aggressive mimicry can also follow a two-
step process in which imitation is spearheaded by a group of influential mimics.
If role models champion mimicry behaviors within their population, aggressive
mimicry constitutes a special case of mimetic isomorphism. It is unique as the mimic
population as a whole is observed to migrate laterally in the models' resource space over
time. The model population remains their primary coevolutionary target of interest for
strategy adjustment. Essentially, direct aggressive mimicry by many individual mimics
and indirect aggressive mimicry through mimetic isomorphism both result in a collective
intrusion of the mimics into their models' identity niche. Both variants seem likely to
occur in organizational populations and likely operate together at the same time.
Mimics can achieve surface similarity with members of a model population
through a variety of means, which largely correspond with form definitions discussed
previously. They may approximate the structural embeddedness pattern of the models by
choosing similar linking partners, they can acquire organizing elements that resemble
those of the models, and they can abandon membership claims to a unique organizational
form in favor of pretending to share the form identity of the models. Label switching
constitutes by far the most cost-effective mimicry transformation as all it requires is
communicative effort by the mimics to establish a new form allegiance towards its
audiences. Paradoxically, labels are powerful identity markers that do not have to
coincide with the true internal identities of organizations (Allaire & Wolf, 2004), even
though name changes are usually taken to signify identity changes (Kogut & Zander,
81
1996). Thus, label changes constitute an organizational marketing strategy that is
comparatively easy to achieve. In contrast, internal organizational identities are prone to
inertia as they form the core of what an organization is about, so modifying them is
highly difficult and dangerous to organizations (Hannan, et al., 2006). As long as the
internal identity of a mimic does not preclude using incorrect labeling, congruence
between internal and external identities is not vital for organizational survival. Not
surprisingly, most cases of aggressive mimicry occurring in the world of organizations
include relabeling strategies.
Conditions for Aggressive Mimicry
Acquiring a new organization name may be comparatively inexpensive for
organizations, but aggressive mimicry is not a strategy chosen routinely as the potential
repercussions for mimics that are detected can be severe. Gauging the severity of looming
social sanctions for violating identity codes, individual organizations may therefore select
aggressive mimicry strategies based on weighing the expected benefits against the
expected risks. Among social movement organizations, aggressive mimicry can be
considered a genuine strategic option if it is kept in mind that the ideational stakes may
very well justify a range of partially quite drastic organizational tactics. Impostors may
lose legitimacy if they are found out, but false identity claims constitute a much more
benign transgression than violent protest or other disruptive strategies common among
activists.
Considering how aggressive mimicry works, it seems that it may fail to provide a
long-term evolutionary strategy if the signal-receivers and models are quick learners able
to adapt. As noted by Griebel and Oller (2008), "it appears that both deceiver and
82
victim are capable of learning, and an arms race of deception and detection ensues" (p.
27). As examples from biology highlight, the success of mimicry can be of limited
duration. If insects are able to learn that they may encounter flesh-eating plants that look
like harmless flowers, they will avoid these predators in the long run. This is particularly
true for organizational mimicry considering that form communication tends to diffuse
through an organizational community over time. With ongoing community evolution,
organizational forms tend to stabilize communicatively as their social environment
becomes increasingly familiar with them. Thus, it is possible that communication within
an organizational community will also spur the emergence of a cognitive category that
serves to correctly identity a population of mimics.
With audiences becoming increasingly sophisticated in terms of recognizing
organizations that are mimicking a different organizational form, the survival of the
mimics becomes less likely and the overall population numbers can be expected to drop.
The fluctuation in the prevalence of business models such as pyramid schemes attests to
the negative effects of increasing recognizability of a mimicking organization form.
While enterprises like the Ponzi scheme have emerged periodically for over a century,
their success has always faltered once public awareness of their existence increased.
There are three broad types of conditions under which aggressive mimicry may
emerge as a relatively enduring strategy for organizational actors. First, a variety of
factors can interfere with the recognizability of mimics, and if cognitive legitimacy stays
low, the risk of detection remains small as well. Second, in organizational communities
that exhibit a lack of regulatory oversight, there are fewer mechanisms for sanctioning
violations of social codes, which lowers the costs associated with adopting an
83
aggressive mimicry strategy. Third, the evolutionary maneuverability of the model
population may be limited, which prohibits its evolutionary flight into a more distinct
niche. In the following, all three conditions are explicated in greater detail.
In contrast to ecological communities commonly examined by biologists, human
practices of record-keeping allow mimics to be identified over time. However, if
available information about a form is sparse and conflicting, cognitive legitimacy levels
may be slow to rise. Ashforth and Humphrey (1997) note that as "the meaning of social
objects is inherently ambiguous, meaning is derived from the triangulation of
perceptions" (p. 49). In some organizational communities, stakeholder audiences are
heterogeneous and disconnected groups that exhibit little discursive overlap, which
reduces their overall influence (King, Lenox, & Barnett, 2002, p. 397). This makes it
easier for aggressive mimics to deceive one or several groups of audiences (Glynn &
Abzug, 2002; D. J. Phillips & Kim, 2008). Additionally, aggressive mimicry strategies
work well in the early stages of the evolution of an industry when novel forms of
organizations have not yet become institutionalized (Dobrev, et al., 2006) and few form
conventions have been established. This is illustrated by Saini's (2008) discussion of
organizations in the UK that appear as legitimate higher education institutions. By
recruiting mostly geographically dispersed foreigners for enrollment and acquiring their
money in advance, such front organizations run profitable business models that only
require websites. The business idea of educational tourism, which is based on attracting
foreign students to earn valuable degrees abroad, is in and of itself relatively new. Thus,
many students are inexperienced when it comes to judging if they are dealing with a real
university and do not know how to verify its authenticity conclusively.
84
As mentioned previously, if the ratio of mimics to models remains relatively
small, the risk of detection is low. Deceptive actions have to be comparatively rare to
ensure that mimicry remains a durable strategy (Sterelny, 2008) as their success is
frequency-dependent (Mallet & Joron, 2007; Ruxton, et al., 2004). In their discussion of
mimics posing as craft brewers, Carroll and Swaminathan (2000) argue that the spread of
a mimic form "would exacerbate identity problems because greater numbers of
consumers and others will come into closer contact with organizations using it and will
therefore learn about the inconsistencies" (p. 734). According to McKendrick et al.
(2003), form recognizability increases through "high-frequency interaction" (p. 69) with
those that display the form. Overall low numbers of mimics ensure that the group of
signal-receivers gets few opportunities to learn about the true identity of the mimics.
There are limited options for experiential learning through immediate contact with
deceivers and the diffusion of such rare experiences through the population is hampered
(Mallet & Joron, 2007). If there are few such learning opportunities for dupes, mimics
can also be less careful about how faithfully they imitate the models as a very superficial
resemblance may suffice.
Low cognitive legitimacy reduces the risk of detection in general, but in
organizational communities with strong policy regimes, those mimicry behaviors that did
get noticed are likely to be sanctioned less severely. Thus, one approach to reduce
deception in a community is "by making it too costly for the agent to engage in" it
(Gerschlager, 2005, p. 11). As sociopolitical environments are able to affect the cost
structure of the organizational strategies chosen (D. S. Meyer & Staggenborg, 1996),
communities that are largely unregulated offer the best conditions for aggressive
85
mimics as fewer norms and regulations are in place that determine how deceptive acts
will be sanctioned. Again, new industries tend to have policy environments that are less
well established and their "rules of the game" are not as rigid as in highly regulated
communities. Societal sectors in general have very different rules when it comes to
punishing organizations that engage in identity deception. While in commercial
industries, many deceptive practices are common and considered part of the sectoral
culture, civil society actors face severe consequences when they engage in false
signaling. In fields where authenticity is extremely important, which is the case in all
craft industries, impostors find it harder to engage in acts of identity "appropriation" (Rao
& Giorgi, 2006, p. 280) due to the steep costs attached to sanctions.
As Levins (1968) argues, a quick adaptive response would be the only way a
model population could adjust adequately to a population of mimics crowding into its
resource space. The third condition for aggressive mimicry depends upon the ability of
model populations to react swiftly when mimics are bound to invade their identity niches.
Aggressive mimicry as a strategy can be sustained better if the specific policy regimes
governing the population of models are so restrictive that models have limited
opportunities for adjusting their niche positions. If model populations lack the power to
ensure that their mimics are sanctioned for their behaviors and are further unable to
mobilize the dupes into developing collective strategies to unveil the impostors, their
only chance to get away from their competitors is to adjust their niche position. However,
this may not always be possible given the rigidity of their policy environment. Health
care organizations such as dentist's offices, for example, are subject to such heavy
regulations that it becomes very difficult for them to reposition themselves when
86
facing the competitive pressure of mimics. On the other hand, mimics such as cosmetic
dentistries are free to experiment with their form identities without incurring the same
regulatory regime as their models.
Evolutionary Implications of Aggressive Mimicry
The introduction of aggressive mimicry into organizational ecology is intriguing
as the concept points to a range of phenomena that may not be captured well if existing
theoretical expectations are maintained. At least three common assumptions built into
current models are questioned if aggressive mimicry strategies are taken into account.
First, the expected association between low cognitive legitimacy and organizational
mortality should not hold for mimics. Second, populations of aggressive mimics should
be found to converge upon their model populations through both selective and adaptive
evolutionary means, even if density levels are rising and competitive pressures are
mounting. Third, successful aggressive mimics should be able to switch form identities
without being penalized with heightened mortality risks.
The heavy reliance of aggressive mimics on low cognitive legitimacy as a
resource requires that the relationship between legitimation and competition dynamics
and their effects on organizational vital rates have to be examined yet again. Community
ecologists typically emphasize the importance of establishing form legitimation for the
survival of populations adhering to novel organizational forms (cf. Aldrich, 1999; Hsu,
Hannan, & Kocak, 2009). Among others (Hsu, 2006b; McKendrick, et al., 2003), Hsu
and Hannan (2005) argue that the degree to which relevant constituencies understand an
organizational form has implications for the long-term success of populations adhering to
that form. In that sense, cognitive legitimacy, associated with the ability of an
87
organization’s external environment to recognize what that form entails, is linked to long-
term population survival (Abrahamson & Fairchild, 1999). Consequently, organizations
claiming form identities that lack cognitive legitimacy are considered at risk for
heightened mortality.
However, populations of aggressive mimics feed on the ambiguities surrounding
their organizational form. Thus, they employ a survival strategy within organizational
communities that does not depend upon achieving cognitive form legitimacy. Their very
source of success lies in their ability to closely resemble a model population in important
aspects, and if their community lacks clarity about their true identity, their imitation
projects are less challenging. Thus, for aggressive mimics, the expected relationship
between the cognitive legitimacy levels associated with their form and organizational
mortality should be reversed. Hudson and Oykhusen (2009) studied the proliferation of
men's bathhouses, organizations which engage in a mixture of protective mimicry and
camouflage, and argued along similar lines. They noted that the "findings also challenge
conceptions of broad-based legitimacy and conformity as necessary conditions for
organizational survival" (Hudson & Okhuysen, 2009, pp. 149-150).
In the case of men's bathhouses and other "core-stigmatized organizations"
(Hudson, 2008; Hudson & Okhuysen, 2009, p. 252), which proliferate despite a serious
lack of sociopolitical legitimacy, their persistence makes sense if it is considered that the
strategies they pursue are mimicry-related, and as such, associated with low levels of
cognitive form legitimacy. Even though current ecological models do not account for
mimicry or camouflage, they are based on theoretical expectations that can be revised in
88
order to capture these phenomena. This way, it becomes possible to test the conditions
under which mimicry occurs and to examine its effects.
Drawing on ecological perspectives on interpopulation competition, D’Aunno et
al. (2000) hypothesized that within a regulatory environment that fosters competition,
organizations would be motivated to adopt form changes that differentiate them more
strongly from their competitors. Their analysis of transformation events within a
population of rural hospitals confirmed this proposition. In general, populations will seek
out evolutionary strategies of distinction if the densities of close competitors heighten the
pressures of competition. However, applied to mimicry strategies by populations, this
theoretical expectation would have to be reversed, as it is exactly the possibility to be in
close competition with their imitation target that the imitators will seek to retain.
Nevertheless, the tendency among mimics to seek out selection and adaptation events that
will increase resemblance with the model population should only occur under certain
legitimacy conditions.
As described before, aggressive mimicry is only a viable evolutionary strategy if
the cognitive legitimacy levels associated with the mimic form are sufficiently low. Table
1 describes the viability quadrants that are opening up when organizations consider
mimicry strategies within specific legitimacy environments. By considering these
options, it is possible to determine when organizations would be likely to adopt mimicry
strategies and thus, increase their level of competition with the model population. As
Table 1 shows, the two lower quadrants open up the most viable opportunities for
successful mimicry. Under these conditions, rising model population densities should be
associated with increased entry into the mimic population. Such entries would be most
89
likely to occur when both sociopolitical and cognitive legitimacy levels are low. The
second best viability region for aggressive mimicry is given by conditions with low
cognitive form legitimacy and high sociopolitical legitimacy. Under these circumstances,
aggressive mimicry is likely to be successful and even if it is not, the sociopolitical
penalties associated with it are small.
Table 1
Viability Regions for Aggressive Mimicry Strategies
High sociopolitical
legitimacy
(= favorable policy
environment)
Low sociopolitical
legitimacy
(= unfavorable policy
environment)
High cognitive legitimacy
(= high recognizability)
No benefits from
aggressive mimicry
Small potential benefits from
aggressive mimicry, high
risk of sanctions
Low cognitive legitimacy
(= low recognizability)
Small potential benefits
from aggressive mimicry,
low risk of sanctions
Large survival benefits from
aggressive mimicry, high
risk of sanctions
The third type of revision recommended for ecological models refers to the
expected relationship between adopting a mimicry strategy and the mortality risks
associated with such an identity-related transformation. As mentioned previously, these
transformations have not been found to be advantageous for organizations as they are
assumed to be too disruptive. Results from empirical studies indicate that the negative
effects prevail even though the change initiatives may offer the potential to bring the
organizations into better alignment with their environments (Hannan, et al., 2006). In
90
the context of aggressive mimicry, these expectations should not hold true. Rather than
experiencing survival penalties due to their form switching, aggressive mimics are
expected to improve their chances of survival.
91
CHAPTER 5: AGGRESSIVE MIMICRY WITHIN THE POPULATION OF CRISIS
PREGNANCY CENTERS
This chapter introduces a population of charitable organizations in the U.S. that
appears to successfully employ aggressive mimicry tactics. Its analysis offers an
opportunity for testing a variety of hypotheses about the conditions and effects of
aggressive mimicry in organizational communities. The following paragraphs provide a
brief overview of the chapter. Crisis pregnancy centers (CPCs) in the United States
represent a population of aggressive mimics with which some of the suggested revisions
to existing ecological models can be tested. CPCs are nongovernmental organizations
(NGOs) that originally emerged from a pro-life grassroots mobilization in North America
against the legalization of abortion in 1973. While some of them still carry the phrase
"crisis pregnancy" in their organization names, members of the population of CPCs are
not easy to identify as they refer to themselves in many different ways. However, the
label "CPC" will be consistently used to refer to all members of this population regardless
of their subpopulation membership throughout the remainder of the document.
The earliest CPCs were founded in the late 1960s (McCaffrey & Keys, 2000).
Popular estimates of the total number of CPCs in the U.S. range from 3,400 (Doan, 2007)
to 5,000, but the number of physical CPC locations might include as many as 6,500
(author's own estimate, based on the number of CPC locations extracted from online
directories as described in chapter 6). CPCs exemplify the tremendous diversity of the
American nonprofit sector (Anheier, 2005) as they appear to be charitable health
organizations comparable to nursing homes, teaching hospitals, and other clinics,
according to Anheier's (2005) classification of NGOs. In contrast, Webster's
92
Quotations, Facts and Phrases defines a crisis pregnancy center as "a non-profit
organization, generally established by Christian pro-life supporters, as a means of
encouraging pregnant women not to have abortions" (Abortions: Webster's quotations,
2008, p. 122). These widely differing notions of how to categorize CPCs attest to the
multi-faceted organizational identities portrayed by these organizations. CPCs have been
frequently accused of seeking to mimic legitimate reproductive health care providers
(RHPs) offering full gynecological services to women (Zwerling, 2008). As a population,
CPCs can be found to pursue different kinds of strategies in order to fulfill their
organizational goals. Doan (2007) summarizes the controversy around CPCs:
All crisis pregnancy centers offer free pregnancy testing, which is often the hook
for bringing women into the centers. Once there, women are counseled, especially
those who have positive pregnancy tests. The quality of this counseling varies
tremendously from untrained volunteers to professionally trained volunteers to
paid social workers (Munson 2002). Pro-choice advocates have accused many
crisis pregnancy centers of being "bogus clinics" and relying on deceptive
advertising practices to entice women in, then using graphic slide shows or films
to "counsel" pregnant women (Staggenborg 1991). (p. 91)
CPCs representing the first cohort openly defined their organizational mission in
ideological opposition to abortion. Their main support, both financially and in the form of
volunteers, came from faith-based organizations, so they often sought to incorporate
evangelical missions into their center activities. Early CPCs were very aware of the
importance of fostering local community relationships (Summerhill, 1984) as they
commonly operated with few resources of their own. Instead, they assumed the function
of community brokers by connecting their clients with resources such as maternity home
placements, adoption referrals, and prenatal health care. Hartshorn (2003) notes that
CPCs considered themselves as emergency intervention organizations who saved unborn
93
children from pregnancies that were conceived of as crises, problems, or risks, which she
describes as the "baby-centered approach" (p. 109). They identified "abortion-vulnerable"
and "abortion-minded" women, those clients who were open to choosing abortion as an
option of terminating their unwanted pregnancies, as the primary targets for their efforts.
Since the crisis frame became common in the late 1970s, many of the CPC practices
associated with a focus on fetal development have been described by pro-choice critics as
scare tactics employed to threaten clients (Mertus, 1990; Staggenborg, 1991; Zwerling,
2008).
Since their emergence, CPCs have grown tremendously in numbers and they also
have undergone significant form changes. Using aggressive mimicry as a tactic, many
CPCs have transformed from a crisis-centered form to a clinic form and accomplished a
great degree of organizational change. The mimics among the CPC population have
renamed their organization using titles devoid of "crisis terminology," they are
increasingly forging affiliation links with national umbrella organizations, and they have
been collectively engaging in the process of acquiring the skills and certifications
necessary for offering a broader range of medical services. The current CPC population is
comprised of pregnancy centers that focus on counseling, some that provide a range of
medical services, and others that operate as Christian counseling centers, maternity
homes, adoption agencies, and programs specializing on post-abortion counseling and
abstinence-only education (Hartshorn, 2003, as qutd. in Doan, 2007).
Given the potential impact that the organizational activity of CPCs may have on
the lives of women that seek reproductive health care, it seems worthwhile to examine
how CPCs as organization forms are reproducing and transforming over time.
94
Opposing movements coevolve by attempting to alter the environmental opportunity
structures of their enemies, but they may also affect each other in more direct ways
through the interdependent activities of populations (D. S. Meyer & Staggenborg, 1996).
The targeted membership claims made by CPCs to the form identity of RHPs have to be
viewed as occurring within the larger social movement battle around the issue of
abortion, and their coevolutionary relationship with the population of RHPs deserves
careful study.
Crisis Pregnancy Centers and the Pro-Life Movement
CPCs make up only one portion of a large and diversified community of
populations involved in the pro-life movement. D. S. Meyer and Staggenborg (1996)
noted in the mid-1990s that "the antiabortion or 'pro-life' movement has grown over the
past two decades in response to the successes of the abortion rights or 'pro-choice'
movement, notably legalization of abortion in 1973" (p. 1629). While the first CPCs were
founded even before the first American states legalized abortion in 1970 (Doan, 2007),
pro-life activism was not as highly organized then and only slowly began to gain
momentum after the Roe v Wade decision in 1973 (Doan, 2007). After a period of slow
growth, CPCs began to flourish in 1983, when the failure of the Human Life Amendment
had galvanized the pro-life movement into action (Maxwell, 2002; Staggenborg, 1991).
The purpose of this amendment was to render the Roe v Wade decision ineffectual by
ensuring that this federal policy could no longer overrule abortion bans introduced in
individual states. The constitutional amendment from 1983, which was sponsored by
Senator Orrin Hatch, represented one of the last of a series of unsuccessful attempts to
amend Roe v Wade (Munson, 2008). After these repeated failures to fight abortion
95
rights through high-level advocacy, pro-life activists started to focus on grassroots
intervention efforts. CPCs allowed for the direct influence of women pondering having
an abortion and their operation became an effective means to prevent pregnancy
terminations one at a time. In 1984, alone the number of U.S. affiliates of the Canadian
umbrella organization Birthright had already climbed to 600 CPCs (Summerhill, 1984).
The estimated number of CPCs by 1986 further rose to approximately 2,100 individual
organizations (Diamond, 1989). This trend has continued and the CPC population has
maintained its growth rate since then.
An essential aspect of the pro-life movement in general and the population of
CPCs in particular is its strong connection to conservative Christians. Until the mid-
1970s, the doctrines of the Catholic Church served as the most important driver behind
pro-life mobilization. Since then, movement participation has broadened beyond
Catholics as "the development of an ideology about the relationship of abortion to family
life and the role of women in society" appealed to a much larger group of Americans (D.
S. Meyer & Staggenborg, 1996, p. 1640). Evangelical Christians started to become
involved in the 1980s (Doan, 2007). They have been influential in shaping the
development of the movement ever since (Steiner, 2006) and multi-denominational
activities have become increasingly common within the pro-life movement (Blanchard,
1994).
The modern pro-life movement consists of three major branches with very
different functions that do not overlap much (Munson, 2002). First, "the political branch
of the movement consists of organizations and individuals who are trying to shape
change through traditional political channels such as lobbying, political campaigns,
96
and litigation" (Doan, 2007, p. 89). Large national organizations such as the National
Right to Life Committee (NRLC) and the Christian Coalition of America are the most
notable actors in this segment of the pro-life movement. Second, the "direct action arm"
(Doan, 2007, p. 89) of the movement consists of grassroots activists that are mostly
focused on abortion clinic protests, which includes tactics such as sidewalk counseling
and clinic vigils. The activist arm has a more militant and confrontational history than
any other segment within the pro-life movement, which can be partially attributed to the
activities of Operation Rescue, an organization that was "prominent between 1988 and
1992 [and] chiefly responsible for the incorporation of confrontational social protest as an
accepted strategy of the general pro-life movement" (Steiner, 2006, p. 5).
The third and fastest-growing (Mathewes-Green, 1996) segment of the pro-life
movement has been referred to as the "outreach stream" (Doan, 2007, p. 90) and it
comprises the crisis pregnancy center movement. Not surprisingly, there is a noticeable
disassociation of CPCs with more activist parts of the pro-life movement as "more
moderate groups within a movement may suffer from a 'negative flank effect'" (D. S.
Meyer & Staggenborg, 1996) when their activities are linked in stakeholders' minds with
radical action (Blanchard, 1994). Historically, CPCs have also been careful to not engage
in ostensive linking with pro-life lobbyists in order to avoid alienating their clients
(Summerhill, 1984). However, CPCs do not exist in isolation from other pro-life
organizations and they are embedded in a strong organizational support network.
Over the years, the movement activities of CPCs have spawned a sizable industry,
which is a development comparable to other social movement sectors such as the slow
food movement (K. Weber, et al., 2008). Apart from several influential national
97
associations that offer affiliation opportunities, CPCs also rely on organizations such as
Ramah International and Heartbeat International, which conduct managerial training for
CPCs and produce informational material such as brochures, DVDs, and fetal
development models shown to clients during counseling. Other CPC cooperation partners
include organizations such as Focus on the Family and the National Institute of Family
and Life Advocates (NIFLA), which have been promoting the equipment of CPCs with
prenatal ultrasound diagnostics, and Life Dynamics, an NGO specializing in health-
related post-abortion litigation.
Even though the number of direct linkages between the three segments of the pro-
life movement is limited, CPCs benefit greatly from higher-level regulatory activism at
state and federal courts. Lobbyist organizations are both responsible for the benevolent
regulatory environment most CPCs experience as well as for the unfavorable
sociopolitical climate within which RHPs operate. Additionally, many symbolic
innovations are created and propagated in other branches of the pro-life movement and
subsequently adopted by CPCs. The "Post-Abortion Stress Syndrome" exemplifies a
cultural product that has been imported by CPCs and incorporated into their counseling
routines. Many CPCs provide post-abortion counseling for women (Wolf-Devine &
Devine, 2009) based on the premise that they are suffering from this psychological
condition, even though it is not a disorder officially recognized by the American
Psychological Association (Buttenweiser & Levine, 1990). As the promotion of this
concept was systematically advanced in other segments of the movement, CPCs were
able to make use of this innovation by simply integrating it into their regular
organizational functioning (Reeves, 2003).
98
The Model Population: Reproductive Health Care Providers
Reproductive health care providers offer comprehensive health care options for
pregnant clients, which includes provision of services related to all three avenues:
Parenting, adoption, and pregnancy termination (A. Baker, 1995). In contrast, CPC
counseling focuses on parenting and adoption as the two available and preferred options.
As CPCs neither refer for abortions nor do they perform them, even if they are licensed
medical centers, the comprehensive attention of RHPs to all legal pregnancy options
constitutes the decisive difference between the two organization types. Kaufmann (1997)
defines RHPs as professional clinics "equipped to provide a wide range of women's
health care services. Most perform pelvic and breast examinations and offer family
planning information, birth control, and prenatal care as well as pregnancy testing and
abortions" (p. 2). Planned Parenthood is the largest national association of RHPs, but
only a portion of its affiliates offers abortion procedures. More than 300 RHPs offering
abortions are associated with the National Abortion Federation (NAF), which is the
national professional organization for abortion providers. Approximately half of all
abortions in the United States are performed by NAF-affiliated RHPs (Kaufmann, 1997).
Throughout the remainder of the document, the term "RHP" will be used to refer to the
subset of reproductive health care providers that offers pregnancy termination.
The population of RHPs in the United States has been declining steadily since the
mid-1980s, as can be seen in Figure 5. The first RHPs appeared after the
decriminalization of abortion in the early 1970s. As RHPs are always potential targets for
clinic violence, data on the evolution of RHPs in terms of their annual founding and
disbanding rates and other organizational characteristics are sparse. Due to safety
99
concerns about their staff, RHPs usually resort to hiring security personnel and their
physical addresses are rarely listed in directories. The only institutions keeping track of
the longitudinal population development of RHPs are The Alan Guttmacher Institute
(AGI) and NAF. All statistics reported about RHPs here are based on data provided by
AGI.
Figure 5: Changes in the population numbers of RHPs in the U.S. since 1973.
While the number of hospitals among RHPs used to exceed the number of
freestanding clinics throughout the early years of the population's existence, it is
noticeable that most current RHPs are no longer housed by hospitals. The migration of
RHPs into independent clinic facilities has been spurred by incidents of violence against
abortion providers, which became more frequent starting in the early 1980s
0
500
1000
1500
2000
2500
3000
3500
1973 1978 1984 1992 2004
Number of Organizations
Freestanding RHPs
Hospitals
100
(McCaffrey & Keys, 2000). By outsourcing abortion services to designated providers,
hospitals were able to externalize altercations with pro-life protesters.
Judging from anecdotal evidence alone about the number of lawsuits in which
RHPs and CPCs have been involved, it can be concluded that the coevolutionary linkage
between CPCs and RHPs must be tight. Baird-Windle and Bader (2001) documented a
string of court battles between RHPs and CPCs, the earliest of which date back to the
early 1980s. Many of the lawsuits were initiated based on RHPs claiming that CPCs
acted fraudulently by posing as abortion providers (Baird-Windle & Bader, 2001).
Calhoon (2007) notes that "CPC offices have also intentionally mimicked the signage of
recognized family planning clinics" (p. 18). In a well-known case from 1982, a CPC
called "Problem Pregnancy" rented office space right next to a Planned Parenthood
affiliate and displayed a sign with the abbreviation "PP," which often led to confusion
among women who were on their way to the offices of Planned Parenthood (Baird-
Windle & Bader, 2001). The advertising techniques of CPCs in their local Yellow Pages
directories were a topic of particular concern for RHPs, which argued that CPCs were
engaging in deceptive practices by advertising in the category "abortion services." After
several congressional hearings were scheduled, the Yellow Pages Publisher's Association
published a list of "revised guidelines for anti-abortion organizations wishing to advertise
their services" (Baird-Windle & Bader, 2001, p. 166). Occurrences such as these have
made the relationship between CPC and RHP populations fraught with friction during
most of their coexistence.
RHPs in most U.S. states are facing an increasingly hostile state policy
environment (Gelb & Hart, 1999; Munson, 2008) as many restrictive laws impede the
101
provision of abortion services. As mentioned by McBride (2008), "ever since states were
prohibited from criminalizing abortion in 1973, a central strategy of the pro-life
movement has been to lobby states to enact laws that would limit the practice of abortion
and women's access to the procedure" (p. 42). Almost all states have policies in place that
restrict accessing RHP services, for example by requiring parental and spousal consent
and mandatory waiting periods. RHPs and their clients may face specialized building
code regulations, late-term abortion bans, conscience-based exemption clauses,
restrictions to abortion funding and emergency contraception, and many other kinds of
limitations. Additionally, some states have specific laws that promote CPC funding, for
example with specialty "Choose Life" license plates (J. T. Berry, 2002). In contrast, a few
state legislatures also have passed facilitative laws, which promote access to abortion
services. Among these are regulations curbing deceptive advertising of CPCs and laws
protecting clinics from violent protest (Mertus, 1990).
Policies restricting access to RHPs are commonly coupled with rules that create a
benevolent climate for CPCs, and in states that facilitate access to RHPs, CPCs are often
more heavily regulated. In that sense, a hostile environment for RHPs also constitutes a
favorable environment for CPCs, and vice versa. As mentioned previously, as a reaction
to lobbying efforts from pro-life groups that resulted in an increase of restrictive
regulations governing RHP activities, abortion rights groups have tried to introduce
legislation that restricts CPC activities. Among these rules are laws that would help to
"monitor crisis pregnancy centers, regulate the information they provide, and require
tighter oversight and disclosure from centers that offer free ultrasounds" (Silverman,
2007, para. 10). Apart from a few exceptions, these attempts have been largely
102
unsuccessful, and most introductions of laws to regulate CPC advertising did not survive
the process in most state legislatures. Thus, CPCs are still able to operate mostly without
regulations that restrict their operations.
As state abortion policies are very influential indicators for sociopolitical climate
(Halva-Neubauer, 1993), it is worth examining if they are stable over time. The maps
provided in Figure 6 allow for a quick visual comparison of the state abortion policy
climates in the years 1989 and 2009. The shading is based on standardized and
categorized state scores that measure the number and types of abortion regulations active
in a state. They are derived from an annual state abortion policy grade compiled and
assigned by the National Abortion Rights Action League (NARAL). As can be seen fairly
quickly, there is in fact noticeable fluctuation within states when it comes to their
abortion policies. As these changes are expected to both affect CPCs and RHPs, they
have to be taken into account when it comes to the determination of the sociopolitical
environment of these two populations.
103
Figure 6: State abortion policy climate in the years 1989 and 2009.
104
Among CPCs, professionalization is limited and environmental regulations
governing their activities scarcely exist. Thus, the population of CPCs is relatively nimble
and able to switch identity categories and strategies with ease. In comparison to this form
diversity, RHPs are medical health care providers, which requires them to follow a host
of rules and regulations. In addition, they are facing the specific policy regulations
associated with the provision of abortion services. Furthermore, in order to avoid
becoming targets of clinic violence, RHPs have to engage in camouflage strategies
comparable with those employed by core-stigmatized populations as described by
Hudson (2009; see also Huheey, 1988). As noted by Carroll et al., (1990) higher levels of
regulatory restrictions on populations "create constraints to organizing and acquiring
resources" (p. 85). Expressed in evolutionary terms, RHPs are less flexible than CPCs in
terms of their ability to respond to coevolutionary pressures they might experience from
mimics crowding into their niche. Their range of options for "evolving away" is therefore
limited.
The Mimic Population: Crisis Pregnancy Centers
Not all CPCs engage in a mimicry strategy towards RHPs; in fact, it is only a
portion of CPCs that can be considered as doing so. A quick glance at the population of
CPCs reveals a dazzling variety of organizations. Some present themselves as distinct, in
the form of anti-abortion ministries that are openly engaging in proselytization and
Christian spiritual counseling. Others appear almost indistinguishable from RHPs,
featuring a large, professionally trained group of staff providing an impressive range of
medical gynecological services, which may even include STD testing and prenatal care.
However, despite of these variegated appearances and identity claims, it still makes
105
sense to differentiate CPCs from RHPs and consider CPCs as one population defined by
an overall shared mission – to dissuade women from obtaining abortions. As mentioned
by Richards (2009), entities may look alike in terms of their morphology, but if they still
serve entirely different functions, it is advantageous to examine their genealogy in order
to prevent miscategorization, a technique that has also been utilized by students of
mimicry in the life sciences (Wickler, 1968). Before the various subpopulations of CPCs
are presented in greater detail, a quick review of their commonalities may serve to help
characterize CPCs in greater detail.
As mentioned previously, CPCs depend strongly on their local communities when
it comes to recruiting volunteers and finding resources for client referrals. Comparable to
other voluntary organizations, CPCs tend to build relatively stable patterns of recruitment
and their volunteers are similar to each other in many respects (McPherson, 1990). Most
CPCs find their volunteers in local congregations, for example by advertising on parish
bulletins (Summerhill, 1984). CPC volunteers and directors are usually motivated by
strong ideological convictions about opposing abortion (Conway, 2000). Often, both
Catholic and Protestant congregations will provide meeting space for CPCs and clergy
will aid in training volunteers (Summerhill, 1984). Among other tasks, volunteers answer
the phones, conduct counseling sessions with clients, collect material aid and donations,
and organize fundraisers.
Apart from depending on private donations, CPCs have been increasingly able to
thrive on public funding sources. Beginning in 1981, a new grant source opened up for
CPCs when they became eligible for support under the Adolescent Family Life Act,
which was a federal grant program to promote sexual abstinence (Ginsburg, 1989).
106
The Charitable Choice Act from 1996 and additional federal abstinence-only funding
during the George W. Bush administration also served to infuse the crisis pregnancy
center movement with financial resources (Lin & Dailard, 2002; Silverman, 2007). Until
federal abstinence-only grants were discontinued under the Obama administration, CPCs
have received at least $60 million in governmental funding (Calhoon, 2007) through
these sources. The beginning of the Charitable Choice era was consequential for CPCs as
it provided incentives for faith-based organizations to officially register with the Internal
Revenue Service (IRS) as 501(c)(3) tax exempt entities. Charitable Choice laws provide
an opportunity for faith-based organizations such as CPCs to become contractors for
governmental services (Esau, 2005). Charitable Choice made it possible for CPCs to
fulfill some of the functions traditionally provided by social service agencies.
Traditionally, faith-based organizations have been reluctant to pursue government
funding due to the restrictions this financial model places on the display of religiosity
(Ebaugh, Chafetz, & Pipes, 2005), but CPCs have systematically increased their pursuit
of government grants over time due to programs such as Charitable Choice.
CPC Subforms Based on Varying Degrees of Mimicry
Despite their many similarities, CPCs can be distinguished based on differences in
their membership claims to different organizational form categories. Throughout the first
chapters, the emergent and subjective nature of organizational forms has been
emphasized and the value of organizational taxonomies has been questioned.
Nevertheless, the study of differences in form communication strategies is dependent
upon employing an a priori typology of forms, however imperfect, in order to capture
transformations over time. This procedure is "useful for testing a cultural argument in
107
which the researcher is required to show . . . how cultural effects vary within the
population or across the time span studied" (Thornton, 2004, p. 25). As argued by
Freeman (2006), "single populations are often best seen as organizational communities
within which classes of organizations occupy distinct locations in geographical and social
space" (p. 148). Given the seeming difference in evolutionary strategy, it may be
worthwhile to stratify CPCs according to the identity spaces they claim towards their
audiences. Based on an extensive content analysis of CPC labels, which is explained in
more detail in chapter 6, four main form categories can be distinguished. Before they are
described in more detail, a quick introductory overview highlights the main differences.
CPCs can be thought of as stratified along an identity form continuum that ranges
from strong distinction to strong mimicry. For the purpose of the study, four ranked
categories of identity-based CPC subforms were differentiated. "Strong distincts"
constitute the one end of the spectrum and include CPCs with names that place them
firmly into the pro-life movement. "Strong mimics" are located on the other end of the
spectrum and comprise CPCs that engage in labeling strategies strongly resembling the
naming conventions common among RHPs. The subforms referring to "weak distincts"
and "weak mimics" both refer to groups of CPCs that pursue somewhat distinctive or
ambiguous strategies, respectively.
By clearly labeling their distinctiveness from RHPs, strong mimic CPCs pursue
identity strategies much like those that are common among many other form
entrepreneurs. As discussed earlier, the benefits associated with exhibiting a unique form
that corresponds to an easily recognizable discursive category are great, and there are
CPCs that are choosing a distinct identity form to follow exactly this route. Not quite
108
as straightforward as the strong distincts, there are CPCs that are referred to as the weak
distincts as they display some unique identity markers that differentiate them from RHPs.
All CPCs fall under this second category if they use organization names with elements
that refer to crisis, emergency, or problem. Historically, the crisis form of CPCs has been
the most frequent form template chosen by new form adoptees, but over the years, this
type of label has waned in popularity among CPCs.
CPCs with neutral-sounding labels comprise a category associated with weak
mimicry. They are devoid of references to the pro-life movement and appear to be
woman-centered organizations. As published in a marketing study conducted for CPCs,
clients evaluate such labels positively as they feel that the staff in such organizations will
be professional and nonjudgmental (Entsminger, 2005). As there is no terminology
characterizing neutrally labeled CPCs as representatives of the pro-life movement, there
is some chance for such CPCs to get confused with an RHP by clients. However, many of
them still retain the word "pregnancy" in their organization titles, which is not typical for
RHPs, so their attempts at mimicry are not as full-blown as the last category of
organizations described.
On the very other end of the mimicry spectrum, CPCs are found which have
chosen organization names that are basically indistinguishable from labels common for
RHPs. Most of them pursue strong mimicry by adopting the label "clinic" or "health
center" and combining it with references to women. These label elements are strong
markers with which CPCs can position themselves as health professionals offering
reproductive health care. The labels usually also signify that the organization has in fact
acquired medical licensing in order to be able to operate a prenatal diagnostic
109
ultrasound (Chandler, 2006). Organization label and feature combine and make it very
likely that potential clients believe that the CPC is in fact an RHP.
Among the earliest entrants into the CPC population, many representatives of the
crisis form can be found, which can be considered weak distincts. Their association with
negatively tinged vocabulary stems from the way activists viewed the role of their own
activities in the late 1970s and early 1980s (Ginsburg, 1989). CPCs considered
themselves as crisis intervention organizations saving a pregnancy that was conceived of
as a problem (Hartshorn, 2003). Averting the pending abortion was regarded as an
emergency situation in which the life of the unborn child was threatened and required
urgent rescue (Steiner, 2006). This emphasis on the fetus or the baby correlated with
references to "abortion-vulnerable" and "abortion-minded" women. These clients were
considered primary targets for intervention efforts as they were open to choosing abortion
in order to terminate their unwanted pregnancies. Prime importance was placed on direct
contact with pregnant women, which was successfully accomplished through a whole
generation of CPCs that formed a thriving "problem pregnancy industry" (McCaffrey &
Keys, 2000, p. 48). As crisis form adherents used to be very common among CPCs, the
label "CPCs" is still used to refer to the population in general even though the labels of
current CPCs often differ from their early names.
In later years, references to problem, crisis, or emergency pregnancies in CPC
naming conventions vanished. This small rhetorical innovation enabled CPCs to engage
in a weak form of mimicry as their new titles appeared more neutral and evaded easy
classification. The new labels signaled neither affiliation with faith-based sponsor
organizations nor did they allude to the problematic crisis frame that had led to a
110
reduction of the influx of clients (Hartshorn, 2003). As Kaufmann (1997) describes, a
neutral CPC label constitutes "a vague-sounding name, often with the word pregnancy in
it" and among the services offered will be "options counseling" and "postabortion
counseling" (p. 16), which is terminology reminiscent of RHP services.
The CPC clinic form became particularly prevalent after 1990, when the national
organization NIFLA launched its promotion of ultrasound technology as an integral
service for abortion dissuasion (Glessner, 2001; Hartshorn, 2003). Since then, an
increasing number of CPCs have adopted this new form identity, and with it, are
engaging in a strong form of identity mimicry. The medical frame accomplishes several
goals: (1) It includes references to women as the main beneficiaries of CPC services, (2)
it alludes to the provision of medical testing services as an added organizational feature to
attract clients, and (3) it allows CPCs to strongly compete with RHPs in an area of
competence that was previously not easily accessible to CPCs.
By adopting the labels "health care" and "clinic" and appearing as professional
health care providers, CPCs are able to mimic commercial nonprofit organizations such
as clinics, hospitals, and nursing homes (Steinberg, 2006). Together with the new label,
CPCs usually acquire the medical licensure necessary to operate ultrasounds. As
Kaufmann (1997) describes, "this means the clinic can legally perform certain medical
tests and procedures and may have a doctor or licensed counselor on staff" (p. 14). As
such, the conversion to the clinic form constitutes a valuable discursive signal that makes
CPCs more competitive in the market for potential clients (Podolny, 1993; Spence,
1974). The only drawback to adopting the clinic form is that the acquisition of ultrasound
equipment is relatively costly, which means that even though many CPCs might like
111
to convert to strong mimicry, they have to solicit funds that will enable them to do so
first.
A fourth type of CPC is represented by organizations that do not pursue any kind
of mimicry strategy. On the contrary, they are using names that emphasize their
allegiance with the "alternatives to abortion" movement (Ginsburg, 1989, p. 122) or
openly signal their affiliations with Christianity. These organizations often specialize in
offering ideological services such as Christian counseling and abstinence-only education
(Blanchard, 1994, p. 48). Among such strong distincts are both very old CPCs and
relatively recent population entrants. Distinct form adherents will often foreground their
religious connections, but they may also be strictly non-denominational. Louise
Summerhill, the founder of the CPC chain Birthright, expressed a very typical desire for
establishing a distinct organizational identity when she noted that "the name 'Birthright'
in every city, from coast to coast, all over North America, to every pregnant girl who
needs us, must have the same essence and follow, as closely as possible, the vision of its
Founder, The Creator of human life" (Summerhill, 1984, p. 24). While Birthright is the
oldest brand of CPC, a rising number of newly founded CPCs have been choosing a
distinct form identity in recent years.
It appears that the continuing decline of the number of RHPs as well as the highly
favorable policy environment under which certain CPCs operate have made mimicry
tactics obsolete for some of them. A CPC director, who recalls her organization's name
change in 1997 to "Women's Choice Pregnancy Clinic," describes that at that time, it
appeared to be a perfect name for attracting clients (Monahan, 2007). With its allusion to
women-centered health-related offerings, this label represents a strong mimicry
112
strategy. However, due to the conservative population served by this CPC, the usage of
the word "choice" in the name did not prove beneficial. As an explanation, the director
notes that "[w]e are two and a half hours from the nearest abortion clinic. With no
competition and a half page ad in the phone book, our name is not the issue. . . . With our
demographics, we do see a great number of abortion-minded clients anyway" (Monahan,
2007, para. 3).
Figure 7: Newspaper coverage of CPC subform labels from 1980 to 2009.
Figure 7 gives an overview of the prevalence of the four types of labels in U.S.
national, regional, and local newspaper stories from 1980 to 2009. Table 2 provides
examples for typical labels representing these categories. Until the mid-1980s, very low
numbers of news items were devoted to CPCs. From that point on, coverage
0
50
100
150
200
250
300
350
400
1980 1985 1990 1995 2000 2005 2010
Number of articles per year
Distinct Form
Crisis Form
Neutral Form
Clinic Form
113
increased and particularly the proportion of articles featuring neutral labels expanded
swiftly. The coverage frequencies suggest that while all labels are commonly used by
CPCs, there has been an intermittent trend towards the increased usage of the crisis and
neutral labels. In later periods of time, the crisis label appears to have become less
fashionable among CPCs whereas the neutral and clinic labels are continuing to rise in
prominence.
Table 2
Examples for Common CPC Labels Covered in Newspapers From 1980 to 2009
Distinct form
(= strong distincts)
Crisis form
(= weak distincts)
Neutral form
(= weak mimics)
Mimic form
(= strong mimics)
Alternative to Abortion
Center
Alpha Center CareNet Care Center
Birthright Alpha House Life Network Life Center
Christian Crisis
Pregnancy Center
Birthchoice Community Life Service
Pregnancy Care
Clinic
Christian Pregnancy
Clinic
Birthline
Community Pregnancy
Center
Pregnancy Clinic
Christian Pregnancy
Help Center
Crisis Pregnancy Center Life Care Center
Pregnancy Help
Clinic
Christian Pregnancy
Center
Emergency Pregnancy
Center
Pregnancy Assistance
Center
Women’s Care
Center
Right-to-Life Center Life Choice Ministry
Pregnancy Counseling
Center
Women’s Health
Services
Right-to-Life Pregnancy
Center
Pregnancy Crisis Center Pregnancy Help Center
Women’s Help
Center
Right-to-Life Pregnancy
Help
Pregnancy Problem
Center
Pregnancy Options
Center
Women’s Clinic
Save-A-Life
Emergency Pregnancy
Services
Pregnancy Resource
Center
Women’s Medical
Center
Pregnancy Ministry
Pregnancy Support
Services
Women’s Resource
Center
Number of articles per form
374 2969 4282 1016
Note. Total number of articles: N = 8,641.
114
Hypotheses about CPCs
The evolution of CPCs within a larger organizational environment has been
described in some detail in the previous sections. In the following pages, four sets of
hypotheses are presented that allow for the systematic scrutiny of the relationship
between aggressive mimicry among CPCs and their population vital rates. All four sets
together explore the conditions under which mimicry occurs as well as the effects it may
have. The first group of hypotheses examines the founding rates of all CPCs. Its
predictors include CPC density indicators, legitimacy variables, and resource indicators,
which all constitute common types of predictors used in ecological modeling. The second
set of hypotheses focuses on the founding rates of the subpopulation of mimics among
CPCs. In this context, the effects of both CPC as well as RHP densities are evaluated as it
is assumed that mimic foundings are related to the vital rates of the models. In addition,
the role of legitimacy is investigated. The third hypothesis group explores under which
circumstances CPCs engage in mimicry transformations via adaptation. It focuses on
organizational name changes mimicking RHP labels as the outcome of interest.
Predictors for this model include cohort effects, legitimacy effects, mimic density effects,
and RHP density effects. The fourth set of hypotheses revolves around the rate of
organizational dissolution among CPCs. Most importantly, it tests the effects of selecting
a mimicry strategy on CPC survival. Additionally, it includes legitimacy indicators,
resource indicators, and CPC density variables.
Founding Rates of CPCs
In order to study the population of CPCs as a whole, it is useful to test a variety of
hypotheses that are derived from general ecological dynamics surrounding founding
115
events. Among those aspects found to play a role, three types of factors deserve closer
examination. Existing studies have highlighted the importance of resource availability for
foundings. They also suggest that the sociopolitical legitimacy associated with a form
affects the rate of entry into the corresponding population. Additionally, density-
dependent processes are likely to play a role for in the context of CPC foundings just like
they have been identified as influential for start-ups in many other populations (Carroll &
Hannan, 1989b; Delacroix, et al., 1989). If density dependence affects CPC foundings,
their entry rates would be related to CPC density in an inverted curvilinear pattern. In
other words, with increasing population density, the founding rates would increase until
the population has reached its carrying capacity and decline thereafter (Carroll, 1997).
Hypothesis 1a. There will be a quadratic relationship between the overall
density of CPCs and the founding rates of CPCs.
Institutional theorizing emphasizes that legitimacy effects are factors that
influence foundings rates. High sociopolitical legitimacy exists if the policy
environments of organizations (Dobbin & Dowd, 1997) are favorable, which is known to
provide an incentive for foundings (Aldrich & Fiol, 1994). Such environmental effects
associated with legitimacy are particularly important for voluntary organizations such as
CPCs and have been articulated as part of the political opportunity structures available to
NGO entrepreneurs (D. S. Meyer & Staggenborg, 1996; Stuart & Sorenson, 2003).
Accordingly, J. D. McCarthy and his colleagues (1988) refer to "environmental
conduciveness" (p. 74) as an important sociopolitical determinant of organizational
founding. Thus, a beneficial state policy environment is expected to encourage foundings
of new CPCs.
116
Hypothesis 1b. There will be a positive relationship between sociopolitical
legitimacy and the founding rates of CPCs.
Several authors have suggested that presidential attitudes may serve as indicators
for sociopolitical legitimacy on the national level. As presidents are highly visible public
figures, their positions and political allegiances have been found to spark as well as
dampen social movement activities (J. Walker, 1991). Therefore, Minkoff (1999)
examined the relationship between democratic presidencies and organizing levels within
the women's movement. Similarly, Soule and King (2008) operationalized political
opportunity structures by checking if activism in the women's, peace, and environmental
movements would increase during the terms of democratic presidents. While their
findings did not yield conclusive results about the role of presidential attitudes on social
movement organizing in the sectors mentioned before, it is worthwhile to examine if it
might play a role in the context of CPC foundings. The presidential stance towards
abortions captures a dimension of sociopolitical legitimacy on the level of the nation,
which makes it an interesting factor to study. During the terms of pro-life presidents,
which can be considered favorable sociopolitical environmental conditions, the founding
opportunities for CPCs are expected to improve.
Hypothesis 1c. CPCs will be preferentially founded during the terms of pro-life
presidents.
As mentioned in chapter 2, environmental resources are paramount when it comes
to the vital rates of populations. The availability of a variety of resources provided on
several levels of analysis is likely to accelerate the founding rates of CPCs (J. D.
McCarthy, et al., 1988). Organizations require resources to function and
117
entrepreneurs prefer industries in which ample resources are available (Scott, 1981). W.
P. Barnett et al. (2003) refer to Stinchcombe (1965) when they observe that foundings
depend on "resources . . . such as financial capital, an available and trained workforce,
supportive legal institutions, and existing organizations that facilitate trade and provide
models for organizing" (p. 675). In the context of social movement organizations such as
CPCs, resource mobilization theory (J. D. McCarthy & Zald, 1977) articulates the
relationship between resources and level of organizational activities. Resource
mobilization theory suggests that "the greater the absolute amount of resources available
to the [social movement sector], the greater the likelihood that new [social movement
industries] and [social movement organizations] will develop to compete for these
resources" (p. 1225). Accordingly, the overall availability of financial resources to NGOs
should be influential for the rate at which new CPCs are emerging.
Hypothesis 1d. There will be a positive relationship between the influx of
financial resources into the NGO sector and the founding rates of CPCs.
For nongovernmental organizations, human resources in the form of local access
to volunteers are an important environmental factor. As noted by Sorenson and Audia
(2000), entrepreneurs are encouraged to engage in foundings due to their already
established networks with existing organizations in a region. As CPCs tend to recruit
heavily from local religious networks (Summerhill, 1984) and emerge from pro-life
communities that have already mobilized Christian volunteers to become active for their
cause, CPCs should be founded at a higher rate in areas with a large proportion of the
population actively involved in churches.
118
Hypothesis 1e. There will be a positive relationship between human resource
availability and the founding rates of CPCs.
Demand-side effects have been found to influence the vital rates of RHPs
(Kahane, 2000) and are also likely to play a role for CPC foundings. In this sense, the
degree to which clients demand the services of CPC becomes an important dimension of
their resource environment. Areas with a large number of potential CPC clients feature a
higher level of demand for their services as an organizational resource. Thus, the regional
availability of clients in terms of the proportion of the female population potentially
facing unwanted pregnancies should constitute another environmental variable important
for CPC founding activities.
Hypothesis 1f. There will be a positive relationship between client availability and
the founding rates of CPCs.
Founding Rates of CPC Mimics
CPC mimics constitute a special subpopulation of the overall population of CPCs.
Given the particular conditions for mimicry and its characteristics, their founding rates
are expected to be influenced by a unique set of factors. As outlined previously, both
cognitive and sociopolitical legitimacy levels likely matter for de novo entry into the
mimic subpopulation. Additionally, the densities of mimic role models within the CPC
population as well as the density and proximity of the model population should influence
the founding rates of CPC mimics. All of these factors are examined in hypotheses
examining the founding rates of CPC mimics.
Based on the notion of aggressive mimicry as a two-stage process via mimetic
isomorphism, the foundings of aggressive CPC mimics may be associated with the
119
availability of intrapopulation role models already engaging in the same strategy
(Freeman & Audia, 2006; Rao, et al., 2003). As mentioned previously, the mimics' lateral
niche migration observed as an outcome of aggressive mimicry can be due to direct
imitation of RHPs as well as due to mimetic isomorphism. In the latter case, influential
mimics within the CPC population provide an incentive for new entrants to follow in
their footsteps. Entrepreneurs deciding to enter the CPC population are likely to select
mimicry labels based on the principle of "frequency-based imitation" (Miner &
Raghavan, 1999, p. 39), which means that they will preferentially adopt the mimic
subform upon entry if they perceive that a sizable portion of existing CPCs has chosen
this form as well. Accordingly, the rate of mimic foundings can be expected to rise with
the growing density of existing role models.
Hypothesis 2a. With increasing density of CPC mimics within the CPC
population, CPC mimics will be founded at a higher rate.
Traditionally, organizational ecology models emphasize that rising levels of
cognitive legitimacy are associated with increased founding activity (Aldrich & Ruef,
2006; Strang & Macy, 2001). Aggressive mimicry as a concept may challenge this
relationship. As a basic tenet of the theory presented about aggressive mimicry, it was
argued that a low degree of form recognizability makes it more feasible as an
evolutionary strategy. If audiences know little about the existence and practices of a
population of aggressive mimics, it is difficult for them to identify its members as
impostors. Mimic subforms should therefore appeal to CPC entrepreneurs particularly
under conditions in which pursuing mimicry is likely to be successful. Based on this
120
rationale, it follows that low rather than high cognitive legitimacy levels make the entry
into the mimic subpopulation more attractive for CPC entrepreneurs.
Hypothesis 2b. There will be a negative relationship between the cognitive
legitimacy of the CPC form and the founding rate of CPC mimics.
In addition to cognitive legitimacy, social movement theorizing points to the
importance of the institutional environment within which aggressive mimics operate as it
is also the major source of sociopolitical legitimacy for a form. Swaminathan (1995)
studied founding rates of specialist wine producers in the U.S. and identified the
favorability of policy environments as an important exogenous predictor for entry rates.
If the regulatory environment sanctions mimicry, e.g. if a CPC is prohibited from
advertising its services under the heading "abortions" in the Yellow Pages due to
deceptive advertising bans, it becomes less tempting to engage in mimicry as an
organizational strategy. On the other hand, if the sociopolitical environment has few
regulations that restrict mimicry strategies, it becomes less costly to experiment with
them and more organizations may engage in them. Therefore, a sociopolitical
environment favorable towards CPCs should not only be beneficial for CPC foundings in
general as suggested in Hypothesis 1b, but also specifically spur the founding rates of the
CPC mimic subform.
Hypothesis 2c. There will be a positive relationship between the sociopolitical
legitimacy of the CPC form and the founding rate of CPC mimics.
Aggressive mimicry involves a tight coevolutionary link between the densities of
mimic and model populations. However, the expected direction of their relationship does
not concur with traditional assumptions derived from organizational ecology. As
121
described previously, mimicry dynamics result in niche positioning patterns that do not
follow typical competition models. Rather than fleeing from competition with their model
population, mimics tend to migrate into regions of the resource space that are occupied
by their models (Huheey, 1988; Kloock, 2000). Therefore, increasing levels of the
density of models should be associated with accelerated entry rates into the mimic
subpopulation.
Hypothesis 2d. With increasing density of RHPs, CPC mimics will be founded at
a higher rate.
As mentioned previously, Carpenter (1933) argues that mimicry strategies require
close geographic proximity between mimics and models. Aggressive mimicry not only
entails a convincing imitation of a rival organizational form, but it also requires that
populations of organizational mimics directly draw on the resource base of target form
adherents. Both of these aspects suggest that aggressive mimicry strategists are likely to
seek out locations in close proximity to their organizational models. Referring to the
location choices of CPCs, McLaughlin and Haas (2003, para. 5) summarize the viewpoint
of mimicry entrepreneurs: "If this is truly God's work, why should we settle for seconds
or thirds? Why not select prime real estate across the street from Planned Parenthood or
the college campus? Go where the action is." It is noteworthy that already in the mid-
1990s, 84% of counties did not have an RHP performing abortions any longer
(Kaufmann, 1997). In 2007, almost 90% of counties in the U.S. were without an abortion
provider (Doan, 2007). As aggressive mimics are likely to be founded with the intention
to increase competition for an RHP, it is expected that foundings will occur preferentially
in proximity to their model rather than in a county without an RHP.
122
Hypothesis 2e. CPC mimics will be founded at a higher rate in close proximity to
a RHP.
Aggressive Mimicry Label Transformations
In addition to entering the mimic subpopulation via founding events, CPCs can
also adopt the CPC mimic form by changing their organization labels in an adaptation
event that occurs at some point during their existence. Many of the ecological predictions
about entry into the CPC mimic population are expected to hold when it comes to
engaging in adaptive aggressive mimicry by changing the label of the organization. But
even though both mechanisms of subpopulation entry are likely to be influenced by a
similar set of predictors, there are also important differences between the two processes.
As will be hypothesized later, similarities should be found in terms of the effects of
legitimacy and the density of intrapopulation role models as well as the density and
proximity of mimics. However, transformation events occur among already existing
organizations, which means that differences in their propensity for organizational change
have to be taken into account. In this sense, adaptive entry into the mimic subpopulation
differs greatly from a de novo entry via foundings. Research examining organizational
inertia (Dobrev, Kim, & Carroll, 2003; Hannan, 1997; Hannan & Freeman, 1984) has
repeatedly suggested that environmental imprinting (Stinchcombe, 1965) at the time of
founding increases inertial pressures operating on organizations. Thus, the set of
hypotheses about label transformations also includes an examination of birth cohort
effects.
Among older organizations, cognitive inertia (Porac & Thomas, 1990) may inhibit
their ability to adopt a mimicry form. As "organizational identities generally become
123
stronger and more resistant to change as organizations age" (Hannan, et al., 2006, p. 760),
the effects of interpretive imprinting (G. P. Hodgkinson, 1997; Weyer, 1989) are
expected to increase, which means that these organizations will exhibit reluctance to
switch to a mimicry strategy that involves identity-based form modifications. Particularly
the oldest cohort of CPCs, which was founded in the era of the "emergency pregnancy,"
may experience interpretive imprinting (Weyer, 1989). Consequently, the tendency of
members of the very first cohort to pursue aggressive mimicry name transformations
should be reduced.
Hypothesis 3a. The first founding cohort of CPCs will engage in aggressive
mimicry at a lower rate than subsequent cohorts.
In the context of CPC mimic foundings, Hypothesis 2b tests a reversal of the
commonly assumed positive relationship between the cognitive legitimacy of a form and
the rate of entry into the corresponding population. This adjustment of expectations also
applies when CPCs enter the mimic subpopulation via organizational label
transformations. As was argued previously, low cognitive form legitimacy should
increase the likelihood of success associated with aggressive mimicry as form ambiguity
aids mimics in escaping detection. CPC entrepreneurs attempting to implement promising
adaptations (Strang & Macy, 2001) may therefore select mimicry strategies preferentially
in situations when cognitive form legitimacy is low and the prospect of adopting them
successfully is high.
Hypothesis 3b. There will be a negative relationship between the cognitive
legitimacy of the CPC form and the rate of aggressive mimicry adoption among
CPCs.
124
The influence of sociopolitical legitimacy on the transformation rate of CPCs is
expected to resemble the relationship between sociopolitical legitimacy and CPC mimic
foundings posited in Hypothesis 2c. In both cases, sociopolitical environments influence
the costs associated with choosing certain organizational strategies. This is possible as
policy environments signal which "regulatory mechanisms of control" (Scott, 2000, p.
127) are in place which determine the kind and severity of sanctions against illegitimate
behaviors (D. S. Meyer & Staggenborg, 1996). Again, high sociopolitical legitimacy
should be conducive to experiments with form transformations as favorable policy
environments reduce the risks associated with adopting an alternative organizational form
(Rao, et al., 2003).
Hypothesis 3c. There will be a positive relationship between the sociopolitical
legitimacy of the CPC form and the rate of aggressive mimicry adoption among
CPCs.
The notion of mimetic isomorphism is not only relevant for CPCs entering the
mimic subpopulation via founding events as tested in Hypothesis 2a, but it also applies to
CPCs which switch into the mimic form by transforming their organization labels during
their existence. As Lee and Pennings (2002) found, organizations watch each other
closely and will engage in outcome-based imitation if they feel that their peers have made
beneficial adaptations. Thus, they "tend to model themselves after similar organizations
in their field that they perceive to be more legitimate or successful" (DiMaggio & Powell,
1983, p. 152). As entrepreneurs cannot always verify that a specific organizational
adaptation is in fact beneficial, they are prone to emulate those strategies that they
consider most popular (Strang & Macy, 2001). With an increasing number of CPCs
125
switching to a mimic form, "cascades of adoption" (Strang & Macy, 2001, p. 147) can be
expected and even more transformations are expected to ensue.
Hypothesis 3d. With increasing density of mimics within the CPC population,
CPC mimics will engage in aggressive mimicry at a higher rate.
Scholars have pointed out that competing populations of social movement
organizations coevolve as they engage in repeated strategic interactions in order to
weaken their opponents (D. S. Meyer & Staggenborg, 1996). This and many other similar
observations suggest that CPC transformations should be tightly connected with RHP
activities. Audia et al. (2006) found that commensalistic relationships such as those
between CPCs and RHPs result in close "competitive monitoring" (p. 390). This
monitoring is not dependent upon direct competition between organizations, but refers to
a general tendency of competitors to react to diffuse competitive pressures exerted by
their rivals. In the case of aggressive mimics competing with their models, the expected
relationship again contradicts traditional assumptions about competition. As has been
posited in Hypothesis 2d in the context of mimic population entry via foundings, CPCs
should adopt a mimicry strategy more readily if the density of their model population is
increasing. Such mimicry moves intensify competition rather than alleviating it.
Hypothesis 3e. With increasing density of RHPs, CPC mimics will engage in
aggressive mimicry at a higher rate.
Studies abound that attest to the importance of examining the impact of
organizational proximity on the evolution of populations (Bigelow, et al., 1997; Simons
& Roberts, 2008; Stuart & Sorenson, 2003; Vasi, 2007). Baum and Haveman (1997)
studied location choices by hoteliers in the Manhattan area and found that
126
entrepreneurs were not afraid of direct competition and even tended to follow their
competitors in relatively close distance, which illustrates the salience of geographical
considerations for the study of entrepreneurial activities (Stuart & Sorenson, 2003). CPC
mimics are expected to react particularly responsively to the presence of RHPs as these
constitute their models and at the same time their direct competitors. Accordingly, as
mentioned by Mertus (1990), "fake abortion clinics attempt to mislead potential clients
by renting office space near well-known pro-choice clinics and adopting names and
initials . . . which resemble those of clinics that actually perform abortions" (p. 98).
Mimicry transformation events are therefore expected to occur more frequently in the
immediate presence of an RHP.
Hypothesis 3f. CPC mimics will engage in aggressive mimicry at a higher rate in
close proximity to a RHP.
Organizational Failures
Ecological models usually include expectations about organizational disbandings
and several typical mortality indicators have already been previously introduced. The set
of hypotheses about organizational failures also features legitimacy indicators as the
success of mimicry strategies in a community highly depends on the legitimation
dynamics that occur within it. Environmental resource availability indicators, which were
also already introduced in the context of organizational foundings, are further examined
in their effect on CPC survival. Additionally, density effects of various subforms of CPCs
are investigated as they are suspected to influence CPC mortality rates. In order to
specifically test the relationship between mimicry and mortality, the survival chances of
CPCs pursuing a mimicry strategy with those of non-mimics are compared. Another
127
comparison examines differences between the mortality rates of CPCs entering the mimic
subpopulation via adaptation versus those that enter it via selection.
Carroll and Swaminathan (2000) describe in their study of the U.S. brewery
industry that organizations potentially face sanctions if their deceptive behavior is
detected by the dupes. Based on the potential costs associated with aggressive mimicry, it
is expected that CPCs will only engage in it if it yields high returns for them. Bates
(1862) was the first biologist who concluded that mimicry would not persist as an
evolutionary strategy if it did not offer superior survival benefits, which should also apply
in the context of organizational communities. Thus, a basic expectation associated with
the ongoing occurrence of aggressive mimicry among CPCs is that organizations
pursuing a mimicry strategy are expected to fail less frequently than non-mimics.
Hypothesis 4a. CPC mimics will have a lower failure rate than CPCs pursuing a
distinctive identity strategy.
Most theorists currently argue that adherents of emerging forms can only thrive
and carve out lasting industrial niches if they achieve cognitive legitimacy (Archibald,
2004; Kuilman & Li, 2009; McKendrick, et al., 2003; Rao, 2001; K. Weber, et al., 2008).
However, aggressive mimicry fundamentally challenges this assumed relationship as
mimics depend upon form ambiguity for survival. If populations employ mimicry as a
dominant strategy, their survival chances should not be negatively affected by low levels
of cognitive legitimacy. On the contrary, they are expected to suffer from high levels of
cognitive legitimacy as the likelihood of detection (Ruxton, et al., 2004) increases if more
information about the organizational form diffuses among dupes.
128
Hypothesis 4b. There will be a positive relationship between the cognitive
legitimacy of the CPC form and the rate of failure among CPCs.
As mentioned previously, high levels of sociopolitical legitimacy have been found
to bolster organizational foundings as entrepreneurs preferentially select legitimate
business types. Given that a beneficial sociopolitical policy environment holds so many
advantages for organizations, it is not surprising that research has also demonstrated a
positive relationship between the sociopolitical legitimacy of a form and the survival
rates of organizations adhering to it (Hamilton, 2006; Singh, Tucker, & House, 1986). E.
T. Walker and McCarthy (2010) argue that "organizations that are taken as legitimate
tend to be rewarded with resources (Aldrich and Auster 1986; Pfeffer and Salancik 1978),
which, in turn, also promote survival" (p. 318). Conversely, higher levels of sociopolitical
legitimacy should be associated with a lower dissolution rate among CPCs.
Hypothesis 4c. There will be a negative relationship between the sociopolitical
legitimacy of the CPC form and the rate of failure among CPCs.
The availability of environmental resources does not only play an important role
for founding processes. It also affects the proliferation of populations as organizing
activities can only be sustained over time if they are fueled by the influx of resources
(Thornhill & Amit, 2003). As mentioned before, volunteers from religious congregations
constitute a vital source for human resources for CPCs. Cadge and Wuthnow (2006) note
that congregations greatly encourage their members to volunteer in faith-based
community organizations such as CPCs. This means that in regions with high
involvement in religious activities, the pool of volunteers for CPCs is large. Given that
CPCs overall tend to be small organizations that heavily depend upon the services of
129
volunteers (Summerhill, 1984), increased access to human resources is expected to
reduce the mortality risks of CPCs.
Hypothesis 4d. There will be a negative relationship between human resource
availability and the rate of failure among CPCs.
As suggested before in the context of foundings, voluntary organizations such as
CPCs do not only thrive on financial resources, but they also require that client
populations demand their offerings. In their longitudinal study on the mortality rate of
daycare centers in Toronto, Baum and Oliver (1991) conceptualized the age range of
children serviced by an organization as a salient dimension of the resources available to
it. This operationalization highlights that clients constitute a sustaining resource for
organizations. It is particularly important for nonprofit organizations, which routinely
solicit funds based on the number of clients served (Brown & Slivinski, 2006), to
demonstrate to donors that they serve a large number of clients. The larger the pool of
potential clients that can be accessed by a CPC in a given geographic region, the better
are the survival chances of organizations catering to these clients.
Hypothesis 4e. There will be a negative relationship between client availability
and the rate of failure among CPCs.
It is possible that organizations decide to differentiate themselves from population
members that are perceived as illegitimate (Milstein, et al., 2002), so distinctive identity
positioning may appeal to many CPCs. As their moves towards distinctiveness accelerate
the cognitive legitimation of CPCs, the population as a whole becomes more visible, and
with it, the mimicry behaviors of the RHP impostors among them. As noted by Brower et
al. (1960), populations engaging in mimicry run the risk of being detected if they do
130
not keep a low profile. The penalties for increasing visibility are likely to diffuse
throughout the CPC population and manifest themselves in increasing failure rates. Thus,
an increase in the density of subpopulation members pursuing distinct identity labeling
strategies should have adverse effects on the survival of the CPC population as a whole.
Hypothesis 4f. There will be a positive relationship between the density of CPC
non-mimics and the rate of failure among CPCs.
Mezias and Lant (1994) note that the evolution of populations may be altered by
Darwinian selection processes as well as Lamarckian transformations and refer to their
combination as the "joint effect of both institutional and ecological influences" (p. 180).
However, based on the literature about the failure risks associated with organizational
change, it might be possible to determine if their influence differs in degree. As
disruptive organizational change always bears a certain risk (Hannan, et al., 2006), it
might pose additional dangers for organizations to engage in adaptive mimicry. Thus,
"born" mimics, CPCs that entered the subpopulation of mimics via selection, are
expected to exhibit lower mortality rates than those CPCs that transformed to the mimic
form by means of adaptation.
Hypothesis 4g. "Born" CPC mimics will have a lower failure rate than CPCs that
entered the mimic subpopulation through adaptation.
Table 3 on the next page provides an overview of all hypotheses about the
foundings, transformations, and failures of CPCs and CPC mimics.
131
Table 3
Summary of Hypotheses
Founding rates of CPCs
Hypothesis 1a. There will be a quadratic relationship between the overall density of CPCs and the founding rates of
CPCs.
Hypothesis 1b. There will be a positive relationship between sociopolitical legitimacy and the founding rates of CPCs.
Hypothesis 1c. CPCs will be preferentially founded during the terms of pro-life presidents.
Hypothesis 1d. There will be a positive relationship between the influx of financial resources into the NGO sector and
the founding rates of CPCs.
Hypothesis 1e. There will be a positive relationship between human resource availability and the founding rates of
CPCs.
Hypothesis 1f. There will be a positive relationship between client availability and the founding rates of CPCs.
Founding rates of mimics within the CPC population
Hypothesis 2a. With increasing density of CPC mimics within the CPC population, CPC mimics will be founded at a
higher rate.
Hypothesis 2b. There will be a negative relationship between the cognitive legitimacy of the CPC form and the
founding rate of CPC mimics.
Hypothesis 2c. There will be a positive relationship between the sociopolitical legitimacy of the CPC form and the
founding rate of CPC mimics.
Hypothesis 2d. With increasing density of RHPs, CPC mimics will be founded at a higher rate.
Hypothesis 2e. CPC mimics will be founded at a higher rate in close proximity to a RHP.
Aggressive mimicry label transformations
Hypothesis 3a. The first founding cohort of CPCs will engage in aggressive mimicry at a lower rate than subsequent
cohorts.
Hypothesis 3b. There will be a negative relationship between the cognitive legitimacy of the CPC form and the rate of
aggressive mimicry adoption among CPCs.
Hypothesis 3c. There will be a positive relationship between the sociopolitical legitimacy of the CPC form and the rate
of aggressive mimicry adoption among CPCs.
Hypothesis 3d. With increasing density of mimics within the CPC population, CPC mimics will engage in aggressive
mimicry at a higher rate.
Hypothesis 3e. With increasing density of RHPs, CPC mimics will engage in aggressive mimicry at a higher rate.
Hypothesis 3f. CPC mimics will engage in aggressive mimicry at a higher rate in close proximity to a RHP.
Organizational failures among CPCs
Hypothesis 4a. CPC mimics will have a lower failure rate than CPCs pursuing a distinctive identity strategy.
Hypothesis 4b. There will be a positive relationship between the cognitive legitimacy of the CPC form and the rate of
failure among CPCs.
Hypothesis 4c. There will be a negative relationship between the sociopolitical legitimacy of the CPC form and the rate
of failure among CPCs.
Hypothesis 4d. There will be a negative relationship between human resource availability and the rate of failure among
CPCs.
Hypothesis 4e. There will be a negative relationship between client availability and the rate of failure among CPCs.
Hypothesis 4f. There will be a positive relationship between the density of CPC non-mimics and the rate of failure
among CPCs.
Hypothesis 4g. "Born" CPC mimics will have a lower failure rate than CPCs that entered the mimic subpopulation
through adaptation.
132
CHAPTER 6: METHOD
Data Sources
In order to construct a longitudinal dataset of the population of CPCs in the
United States, a variety of sources were combined. The data used for analysis were
collected from several archival databases. As Aldrich and Pfeffer (1976) and Ventresca
and Mohr (2002) recommend, VSR processes operating on organizational populations
can be studied well if longitudinal archival data are available. Like many nonprofit
organizations, CPCs are not listed in exhaustive, well-kept industry directories, but they
can be identified if multiple data sources are combined and used to buttress each other,
which is a procedure that has been outlined by W. P. Barnett et al. (2003). Therefore, the
data collection strategy consisted of supplementing tax-related incorporation data sources
supplied by the National Center of Charitable Statistics (NCCS) and GuideStar USA with
online directories listing CPCs and RHPs. Longitudinal data about the vital rates of RHPs
were obtained from The Alan Guttmacher Institute (AGI). The content analysis of
newspaper coverage about CPCs was based on annual searches in the source "U.S.
Newspapers and Wires," which is indexed in the article database LexisNexis Academic.
Data from the U.S. Census 2000, the National Abortion Rights Action League (NARAL),
the Association of Religion Data Archives (ARDA), and the Giving USA Foundation
were obtained for the creation of supplementary variables.
Tax-Exempt Charities Databases
Boris and Steuerle (2006) state that it is advantageous to use the Business Master
Files (BMFs) provided by the NCCS "for analysis of the organizational makeup of the
nonprofit sector" as this dataset constitutes "the most comprehensive list of nonprofit
133
organizations available" (p. 87). NCCS's BMFs and Core files are based on documents
filed by charities with the Internal Revenue Service (IRS). These documents have to be
submitted in order to become and remain eligible for tax-exempt donations. Incorporation
with the IRS entails the assignment of a federal employer identification number (EIN),
which is a unique identifier that organizations carry throughout the entire duration of
their existence. IRS databases tracking tax-exempt entities refer to individual
organizations by their EINs. This is an important advantage of IRS-based data given that
among CPCs, name changes are frequent and the very occurrence of a name change is
conceptualized as an important dependent variable signaling an identity change (Kogut &
Zander, 1996).
The BMFs are cumulative files updated several times per year in which all tax-
exempt charitable organizations are listed that are still considered active. The information
provided in the BMFs has been collected from the IRS Forms 1023 and 1024 (Twombly,
2003). The BMFs capture all public charities, but it takes some time until defunct
organizations are purged from them as the updating of listings depends upon contacting
organizations by mail (Lampkin & Boris, 2002). The Core files are based on information
from the IRS Form 990 and contain more detailed information, but only a portion of
public charities are listed in them as very small organizations are not required to file
Form 990 (Lampkin & Boris, 2002). The nonprofit databases include the so-called ruling
date, which is the year in which an organization acquired its status as a public charity.
This ruling date has been typically used as the organizational founding year in a variety
of studies as organizations often seek to become charities as soon as they are founded
(Esparza, 2007; Twombly, 2003). As noted by Twombly (2003), the
134
operationalization of organizational birth dates as their ruling dates is not entirely
unproblematic as the registration with the IRS does not necessarily have to coincide with
the actual founding date of the organization. Also, there is no mandatory incorporation
for small nonprofit organizations and faith-based groups, so it is possible to operate a
CPC without having acquired tax-exempt status. However, given that charities profit
from being able to offer tax-deductible donations, many CPCs decide to incorporate
immediately at the time of founding even though it is not a requirement (Lampkin &
Boris, 2002).
The NCCS databases based on yearly IRS returns provide valuable information
on the year of a CPC’s name change, its overall size in terms of its assets, and its
location. The BMFs are available in yearly intervals from 1989 to 2010 with a gap from
1990 to 1994, and the Core files cover the years 1989 to 2007, so in combination, they
allow for an analysis of the period 1989 to 2009. NCCS data have proven valuable for
nonprofit research (cf. Gordon, Greenlee, & Nitterhouse, 1999; Marquis, 2003).
GuideStar USA, which is a partner of NCCS, provides a freely searchable online
database based on the same tax information as NCCS. As GuideStar also allows for a
keyword search based on organizational mission statements, if available, it was used to
extend the organizational sample used for study. While GuideStar also references
organizations based on their EIN for easy matching, it only features organizations
currently in existence, so its main usefulness consists in the opportunities for data
verification it offers. Among other features, GuideStar listings also include information
about the Internet addresses of many organizations, which made it possible to conduct
quick fact-checking on individual CPCs.
135
The Alan Guttmacher Institute RHP Database
A longitudinal dataset containing the number, type, size, and county location of
RHPs was made available by The Alan Guttmacher Institute (AGI). AGI is a pro-choice
nonprofit research institute which maintains a yearly database on reproductive health care
organizations. AGI provided annual data on the number of RHPs between the years 1973
and 2005 in individual counties. Information about individual organizations was not
included in the dataset, which meant that a full-scale analysis of vital rates among RHPs
was not possible. Detailed analyses of births and deaths require information about the
exact number of foundings and disbandings rather than just change statistics, and due to
confidentiality reasons, organization-level data could not be provided by AGI. However,
density measures were created for a given year detailing the number of RHPs on several
levels of analysis, which included the density of organizations within an individual
county, a U.S. state, and the entire United States. Additionally, the data was stratified
according to provider type (non-hospital or hospital). The organization size of a provider
was calculated based on the annual number of abortion procedures provided. This size
variable was used to create weighted density measures that take the overall population
mass into account (Carroll, 1988).
Online Databases Listing CPCs
The most extensive online directories of CPCs are typically offered by national
CPC umbrella organizations such as Heartbeat International, which lists more than 1,800
CPCs in the United States alone. The online directory maintained by Heartbeat
International also includes approximately 850 non-Heartbeat affiliates. Other directories
are provided by OptionLine, a joint venture by Heartbeat International and the
136
national umbrella Care Net, and by Several Sources Shelter, an organization that
maintains the LifeCall CPC online directory with more than 2,700 listings. The umbrella
organization Birthright, which was originally founded in Canada, Ramah International,
an organization specialized in producing post-abortion counseling material for CPCs, and
a variety of smaller regional organizational associations host additional Internet
directories. The directory of Ramah International, which lists over 2,500 organizations, is
quite interesting as it also contains many organizations that are not affiliated with
national umbrellas. Both OptionLine and the Heartbeat International directory provide
additional information on CPCs beyond organization names and address, which makes it
possible to estimate that approximately one third of all CPCs in the United States have
added basic medical services to their range of activities. Most directories allow for CPC
directors to enter information about their organizations on their own.
While many of these directories exhibit a large overlap of listings and a
considerable number of them may be obsolete, they are useful for getting an idea of the
overall size of the CPC population in the United States. More importantly, the
organizations listed identify themselves as crisis pregnancy centers, which means that the
directories offer the opportunity to verify if an organization is in fact a CPC. Given the
ambiguity associated with many labels chosen by CPCs, checking directory listings is
often the only avenue for determining if a certain organization may be classified as a
CPC instead of a RHP or a social services agency of some kind. Therefore, scraping
software was employed to extract listings from the six largest CPC databases. Most
listings contained organization names, addresses, and website information. Altogether,
9,284 entries were collected and examined. After grouping the results according to
137
similar phone numbers and ZIP codes, 4,485 unique organizations remained. Many of
them were found to employ multiple names and they had several street addresses listed.
The organization list created in this fashion could be used to verify if potential CPC
listings from the NCCS database were in fact to be included in the sample.
LexisNexis News Database
The full-text online news database LexisNexis was used for an extensive content
analysis of U.S. newspaper coverage about CPCs, which provided the basis for the
analysis of the shifting form discourse about CPCs. LexisNexis was searched as far back
as 1970, the year in which the first CPC included in the population analysis was founded,
but as there were no references to any of the known CPC organization labels prior to
1980, the content analysis used for measuring cognitive legitimacy includes coverage
between the years 1987 and 2009. The annual frequency of articles referring to one of the
four CPC subforms between the years 1980 and 2009 was determined for the purpose of
evaluating the coverage fluctuations specific to the various label types (see Figure 7 in
chapter 5). More details are provided in this chapter's section outlining the
operationalization of cognitive legitimacy. Its measurement was based on network
measures derived from the content analysis of the co-occurrence of keywords. These co-
occurrence networks were constructed and clustered according to meta concepts, which
will also be explained in the same section. Then, density measures were calculated to
assess the annual levels of cognitive legitimacy associated with CPCs.
NARAL Policy Index
In order to capture the sociopolitical legitimacy of CPCs, measures of the state
abortion policy climate were constructed and combined with those already available.
138
As suggested by Nicholson-Crotty (2007) and Tucker et al. (1988), it is important to
consider the policy environment surrounding organizations on the state level rather than
nationally due to considerable policy fluctuations between states. The policy environment
of CPCs was gauged by evaluating the state abortion policy climate as it serves as an
indicator for the institutional environment within which pro-life organizations are trying
to function. States with stricter and more numerous regulations towards abortion exhibit a
favorable policy environment towards pro-life organizations. If the policy climate
towards abortion is very relaxed, it can be taken as an indicator that the pro-life
movement in that state is not as active. Thus, based on known social movement
dynamics, sociopolitical environment effects can be inferred (Ingram & Rao, 2004).
Ingram and Rao (2004) pursued a similar strategy when they operationalized the
enactment of rules and regulations against chain stores as an indicator of a beneficial
climate towards the movement for chain stores. In her study on health advocacy,
Nicholson-Crotty (2007) used scores from Who Decides, the annual index of state-level
reproductive health care policies created by NARAL. Based on NARAL's comprehensive
monitoring of federal and state legislative activities related to the issue of abortion,
numerical grades were assigned to every state on a yearly basis. This numerical, weighted
index "provides the most comprehensive measure of the reproductive health policy
environment within the states" (Nicholson-Crotty, 2007, p. 13). More details about the
construction of the sociopolitical legitimacy measure are provided in a separate section of
this chapter.
139
Supplementary Data Sources
The availability of human resources for volunteering was captured with data on
the level of church participation in a county, which was extracted from the Association of
Religion Data Archives (ARDA) and downloaded from their dataset based on the year
1990. Throughout the history of the crisis pregnancy center movement, CPCs have been
known to depend on recruiting their unpaid staff from Christian congregations (Conway,
2000; Summerhill, 1984), so the degree to which individuals in their proximity are
involved in church activities provides a measure of human resources available to them.
The membership data include participation in all religious groups, not just those of
Christian faith with pro-life ideologies. For the purposes of a similar study, Doan (2007)
included the Catholic congregation membership only. However, as Catholics make up
only a portion of Christians active in the pro-life movement, it appeared preferable to not
restrict the measure to Catholics. As the American Religious Identification Surveys
(ARIS) conducted by the Graduate Center of the City University of New York have
shown, the overwhelming majority of religious adults of 18 years of age and older adhere
to Christian faiths. In 1990, ARIS found that 86.2% were associated with various
Christian denominations (Kosmin, Mayer, & Keysar, 2001). Thus, the ARDA overall
church membership data were considered a sufficient approximation of church
participation in a county.
For the construction of additional variables, further datasets were explored and
utilized. Data on population demographics on the county level were obtained from the
U.S. Census 2000 dataset. In order to capture the amount of overall financial resources
available to the U.S. nonprofit sector over the years, the 52
nd
Annual Report on
140
Philanthropy (Giving USA Foundation, 2007) was perused. This data source is created by
The Center on Philanthropy at Indiana University and published annually by the Giving
USA Foundation. It lists the total amount of donations provided to charitable
organizations in billions of current U.S. Dollars.
Data Preparation Procedures
Construction of CPC Database
In order to create an exhaustive list of tax-exempt CPCs based on searches from
the NCCS BMF and Core databases, a variety of data preparation steps had to be taken.
As an initial strategy, the searchable National Taxonomy of Exempt Entities (NTEE)
classification codes were inspected in order to determine under which categories CPCs
were listed. Similar to the North America Industry Classification System (NAICS), the
NTEE codes provide a means to classify charitable organizations in a systematic fashion
(Lampkin & Boris, 2002). In 1999, NCCS staff revised existing NTEE codes to ensure
that the activity information collected by the IRS about organizations can be translated
without error into the NTEE classification system (Lampkin & Boris, 2002). The
categorization of organizations according to NTEE codes is conducted by IRS staff
members and usually remains the same unless an organization switches its focus of
activity. Based on preliminary searches, it was determined that a distinct classification
code exists that is used exclusively for CPCs. They are typically classified as "P47"
organizations, which means that they are categorized as Human Services organizations.
Based on this NTEE code, a multiple-year search was conducted and 1,442
unique organizations were identified that were listed under "P47" in at least one year
between 1989 and 2009. Unfortunately, the NTEE categorization scheme was found
141
to be neither completely reliable (Lampkin & Boris, 2002) nor was it able to capture all
organizations in the database with names suggesting that they were members of the CPC
population. Therefore, it seemed necessary to not only conduct a database search based
on the customary NTEE code associated with CPCs, but an additional search had to be
designed that would also detect all potential CPCs that were not classified correctly. A
serious challenge in this process originated from the very ambiguity of CPC labeling
choices. Due to their often very indistinct names, CPCs could not be identified simply
based on common keywords contained in their organization labels.
In order to develop a search algorithm that could capture all potential CPCs
within the database, extensive keyword searches for typical CPC labels were conducted,
which yielded a number of common categories with which CPCs are classified in the
database. As can be seen in Appendix A, CPCs were found among a variety of code
classes associated with the "E" category, which denotes health-related activities. They
were also listed in a number of other "P" categories, and occasionally, they occurred in
the category "R62," which denotes organizations classified as "right-to-life." CPCs that
were included in the sample for analysis were listed in a total of 71 unique code classes,
which suggests that the NCCS classification routines also involved categorization errors.
Based on the visual inspection of popular CPC labels, a list of common keywords was
developed that would capture a large number of potential NCCS listings representing
potential candidates for inclusion in the dataset. The application of an extended search
algorithm, which combined NTEE categories and keywords, yielded 7,345 unique
organizations. As 354 of these search hits could be immediately identified as CPCs due to
their usage of labels that were distinct enough, they could be added to the inclusion
142
list immediately. Table 4 provides an overview over the searches performed in the NCCS
and lists the search terms that were used.
Table 4
Summary of Searches Conducted in NCCS and GuideStar Databases
Type of search Search terms
NTEE categories searched in NCCS
E19, E40, E42, E70, E99, P01, P20, P30,
P40, P42, P45, P46, P47, P50, P70, P80,
P82, P99
Keywords searched in NCCS
abortion, alternative, beginning, birth, care,
crisis, heart, hope, life, pregnancy, unborn,
way, woman
CPC abbreviations searched in NCCS CPC, PC, PCC, PRC, WRC
Keywords searched in GuideStar in six
separate full-text searches
right-to-life, right to life, R62, P46, P47,
pregnancy
Note: The NCCS keyword search included linguistic variants of the terms listed and was
combined with NTEE category and NCCS abbreviation searches in a single database search
term.
While organizations with labels suggesting that they were active in other
nonprofit sectors could be instantly excluded from the extended search, 3,042
organizations remained that exhibited ambiguous organization labels and required further
examination. Accordingly, more information was collected about them. This was
accomplished by a search in the GuideStar database. Some organizations were also cross-
referenced with a list of CPCs scraped from online directories. Based on the
organizational mission statements listed in GuideStar or their occurrence in a CPC online
directory, some of the potential CPC listings warranted inclusion in the dataset. In
143
addition, both a search in the Yellow Pages and a Google search were conducted to
gather more information about organizations not listed in online directories. It was
deemed a sufficient criterion for inclusion as a CPC if an organization was listed under
"abortion alternatives" in the Yellow Pages, for example. In some cases, organization
websites were used to determine inclusion, particularly to weed out organizations with
multiple or specialized purposes that did not provide options counseling to their clients.
It is a peculiarity of GuideStar and NCCS that few organizations that are
classified as "P47" in NCCS are members of the same category in GuideStar. On the
contrary, many CPCs appear in GuideStar listed as "R62," which might be due to the
possibility that organizations can modify their NTEE code listings in GuideStar
themselves. As GuideStar is a database specifically designed to connect organizations
with their donors, many CPCs have opted to abandon mimicry strategies in this context.
A probable cause might be that revealing their true identity in GuideStar ensures that pro-
life donors recognize CPCs as appropriate targets for donations. Similar conclusions
could be drawn when the mission statements of CPCs in GuideStar were inspected, which
usually appeared very explicit about the pro-life stance of CPCs. Based on such identity
declarations with which CPCs positioned themselves as pro-life organizations, additional
CPCs were discovered and added to the dataset. This strategy of verifying population
membership through means such as GuideStar mission statements yielded another 148
CPCs, which increased the total number of organizations to 1,944.
As a next step, six additional keyword searches were conducted in the GuideStar
database to find more CPCs based on the content of their mission statements, their NTEE
codes, and parts of their organization names. Table 4 also provides the search terms
144
that were used for these exploratory searches performed in the GuideStar database. The
results were combined, and the names and mission statements were inspected. After
eliminating irrelevant search hits and those already retrieved from previous searches, 170
additional organizations were identified, but only 159 of them also appeared in the NCCS
database, which produced a total of 2,103 organizations eligible for inclusion in the
dataset. Finally, the NCCS database search was repeated for both BMF and Core files,
this time restricted to the 2,103 EINs of organizations identified earlier, which resulted in
sets of annual tables from 1989 to 2009 containing all organizations previously selected
that were listed in that year.
After additional data cleaning strategies had been applied, 16 organizations were
removed from the dataset as their incorporation date was before 1970. These
organizations were mostly churches or other charitable faith-based organizations not
founded as pregnancy counseling services and thus, did not constitute typical CPCs.
However, pregnancy ministries founded after 1970 were left in the sample if they were
independent organizations created for the explicit purpose of counseling pregnant
women. Another 13 organizations were excluded as they had been identified incorrectly
with a "P47" code, probably due to their name similarities with CPC chains such as
Birthright (e.g. organizations such as "Birthright Israel Next"), or they were special-
purpose or multi-purpose CPCs that deviated markedly in size and structure from typical
CPC listings in the database. Among those excluded were organizations founded as
maternity homes, Catholic social service agencies, adoption agencies, shelters, parenting
help organizations, birthing centers, and teen pregnancy reduction and abstinence-only
education centers.
145
Finally, 29 organizations were excluded specifically from the event history
analyses as they only appeared in one annual file throughout the years, most often in the
year 2009, which made it impossible to observe a transition rate for them. The final
number of organizations entered into the dataset for event history analyses consisted of
2,045 CPCs. 1,377 of these were found to be present in both BMF and Core datasets,
which means that for approximately 67% of the organizations, analysis depended on both
data sources for the extraction of key organizational variables.
Organizational Vital Rates and Attributes
The NCCS BMFs and Core files were combined to serve as a primary source of
organization-level data on births, deaths, and transformations of CPCs over time. While
the BMF database constitutes a more exhaustive collection and captures all incorporated
charitable organizations, there are several disadvantages associated with this data source
that could be partially remedied by combining it with the Core files. First, it is a
cumulative file and extracting exact dates about an organization's date of disbanding from
BMFs is unreliable as a listing may take up to three years to get purged from the files
(Lampkin & Boris, 2002). Second, there is a coverage gap between the years 1989 and
1995, which poses a problem when it comes to pinpointing the exact year a name
transformation might have taken place, and it also precludes obtaining updates for time-
varying variables. On the other hand, the Core files feature fewer organizations given that
only those organizations are included in the Core database that file IRS Form 990, which
is not a requirement for smaller entities. While a defunct organization is purged in the
subsequent year from the Core files, they are not a completely reliable source for
determining organizational deaths as it is also possible that an organization missed the
146
filing deadline or that an organization not required to file Form 990 chose to no longer
return it (Esparza, 2007).
As a solution, both BMF and Core listings on individual organizations were
combined based on EIN identifiers, which resulted in a dataset that was more complete
and robust than could have been achieved from focusing on only one of the two data
sources. As mentioned previously, slightly more than two thirds of the organizations
appearing in the final dataset were listed in both BMFs and Core files. However, as it was
possible that information extracted from the two sources conflicted for a year, a set of
rules specifying how to deal with these data inconsistencies had to be created.
Additionally, several other peculiarities of IRS-based tax data had to be captured with the
application of consistent rules about data interpolation and streamlining. In the following,
several of the problems encountered in the process of deriving data on individual
organizations are described and the data cleaning efforts undertaken are listed.
The founding year of an organization was determined based on the entry for its
ruling date listed in the NCCS databases. For 8 organizations, no such entry was available
throughout all years within which they appeared, so their first year of appearance in the
database was used as a proxy for their founding date. For 31 organizations, two
conflicting dates were designated as their ruling date. After individually researching their
founding years, the more plausible entry was selected. 206 organizations were listed in
BMFs or Core files that preceded their ruling dates. In most of these cases, their
appearance predated their founding by one year. These entries were purged from the files
as it was decided that for purposes of consistency, the ruling date entry was to be treated
as the authoritative information about an organization's founding year. 285
147
organizations were interval-truncated in the sense that they were not tracked yearly due to
the gap in the BMFs between the years 1989 and 1995. For these organizations, no Core
file information was available to fill in the gaps, but if the variable entries between the
two years were the same, it was concluded that they had not changed and the gap was
interpolated with information from the year 1989 accordingly.
For 72 organizations, it was not possible to determine the exact year in which they
changed their organization label as information in BMFs and Core files conflicted or they
belonged to the group of entries listed only in BMFs and with a gap between 1989 and
1995. In many of the cases, the possible window for a name change only included two or
three years. In order to interpolate the transformation year, randomization procedures
were used to designate one of the possible years within which the name change could
have taken place. Further, 170 organizations in the dataset switched their county location
over the years, which complicated their location assignment to a single county. Given
that their location in a certain county was mostly important for founding analyses,
information about the county associated with their original incorporation was taken as a
time-invariant attribute and carried over through all remaining years. A total of 1,268
organizations did not enter into observation immediately at their time of founding, which
was corrected in the event history analyses by designating them as left truncated. 834 of
them were founded prior to 1989, so their yearly activities prior to 1989 were not
captured by data sources available from NCCS. Among the 434 organizations with
founding dates in or after 1989, but which appeared later than that, most were listed in the
year immediately following their ruling date.
148
As explained previously, both BMFs and Core files pose challenges to researchers
extracting information about organizational disbandings. A set of rules was designed
specifically to ward against the individual peculiarities of the data sources when they
were used in combination to determine organizational death dates. For organizations only
listed in BMFs, the year of disbanding was set to one year after their last appearance. For
organizations listed in both BMFs and Core files, the year after their last appearance in
the Core files was used for the designation of the dissolution date in cases where they
disappeared in or before the year 2005. As organizations may skip filing Form 990, this
was a conservative method of establishing disbanding dates. However, if the last
appearance of an organization in BMFs was in the year 2008 and they were listed in the
Core files in the year 2006, their Core file disappearance was used to determine their year
of disbandment as BMF records are known to register dissolution dates more slowly.
Organizational Label Transformations
A qualitative content analysis was conducted in order to determine the
subpopulation membership of CPCs in four different strata exhibiting various degrees of
mimicry. Once the organization name was coded, it was possible to determine label
transformation events designating that CPCs were switching identity categories. The
categorization of organization names was based on information from the NCCS data files
and allowed for a determination of the years in which organizations carried certain
names. Classifying organization names is somewhat arbitrary as different audiences may
assign various meanings to the labels used by organizations, so the main strategy
underlying this effort focused on the determination of the degree of mimicry noticeable in
the name choices of CPCs over the years. Due to the strong anecdotal evidence in
149
literature about the labeling tactics of CPC (Baird-Windle & Bader, 2001; Staggenborg,
1991; Zwerling, 2008), it was safe to assume that CPCs were taking RHPs as their
reference point for defining their competitors (Porac & Thomas, 1990) and picking
names. Thus, key dimensions for the classification of CPC labels captured the similarities
and differences between CPCs and RHPs.
The first step towards developing a classification scheme consisted of a content
analysis of common names used by RHPs. While there are no comprehensive lists of all
RHPs in the United States, it was possible to obtain 562 organization names of RHPs
from two online directories. The directory Abortion Clinics OnLine (ACOL) is available
at www.gynpages.com and listed 400 RHPs at the time of search. The directory
abortion.com listed 162 providers at the same time. Based on a preliminary analysis of
these names, the coding scheme for strong mimicry was developed. Table 5 lists
examples for labels corresponding to all four coding categories used for classification.
These are arranged on an ordinal scale ranging from strong distinction to strong mimicry.
Location names, proper nouns, and acronyms such as "AAA" were not considered in the
coding process. If names contained elements signaling membership in different
categories, a conservative coding strategy was pursued and the more distinct category
was chosen for classification to avoid the erroneous assignment of a mimic identity.
When organizations were listed with two names connected by the phrase "operating as"
or "Dba," an abbreviation for "doing business as," the second label was used for coding.
150
Table 5
Examples for Typical CPC Labels Classified Along a Mimicry Continuum
Strong distincts
Type 1
Weak distincts
Type 2
Strong mimics
Type 3
Weak mimics
Type 4
Birthright of Kokomo
Alternatives A Crisis
Pregnancy Center
Community Care
Pregnancy Center
New Dawn Womens
Clinic
Sav-A-Life of the Pearl
River Area
Crisis Pregnancy
Center of the Jersey
Shore
Pregnancy Support
Center of Warren
County
Alpha Womens Center
of Barry County
Mother and Unborn Baby
Care Pregnancy Problem
Centers
Light of Hope Crisis
Pregnancy Center
Agape Pregnancy Care
Center
Womans Choice
Services
Helpers of Gods Precious
Infants
Advise and Aid
Pregnancy Problem
Center
First Choice Pregnancy
Center
Crisis Pregnancy Help
Clinic of Caldwell
County
Hill Country Christian
Action Council
Mary & Elizabeth
Crisis Pregnancy
Center
Hamilton County
CareNet
Womens Resource
Centers of Jacksonville
CPC brand names with highly suggestive pro-life references such as Birthright
and SAV-A-LIFE were coded as strong distincts as the establishment of brand names
obviously helps in the process of coalescing into a recognizable industrial category
(McKendrick, et al., 2003). Further, any clear symbolic association with Christianity and
the pro-life movement was coded as a strongly distinct identity label. As mentioned by
Stevens (2002), the label "ministry" is often a signifier for CPCs associated with
conservative Protestantism. References to biblical words and various names for God are
also common ways to refer to Christian organizations (Stevens, 2002). Published
marketing research of CPCs shows that potential clients are aware that such connections
will mean that an organization labeled in that fashion will not offer comprehensive
reproductive health care (Entsminger, 2005). Thus, they were coded as strong distinctive
identity markers if they were found to be part of an organization's name.
151
Weak distinctive labels included references to crisis, emergency, and problem
pregnancies as well as to "life" and "birth." Neutral-sounding labels were coded as
evidencing weak mimicry as they were devoid of any hints that the organization may
have been anti-abortion, but they were not resembling RHP labels so closely as to
warrant deliberate ambiguity. The CPC brand name "Care Net" was also classified as a
weak mimicry label due to its symbolic association with the labeling common among
professional health care centers, which partially offset any branding effect that might
have occurred. Frequently, weak mimicry labels also included references that suggested
the organization might have been a social service agency of some kind as the words
"community" and "family" occurred frequently.
Among the first three label types, the word "pregnancy" was usually a part of the
organization name, but in many cases, strong mimic name variants did not contain this
word. Labels were coded as signifying strong mimicry if they suggested that the
organization was a health care provider and there were no hints that the health care
provided did not include a comprehensive range of services. A second type of strong
mimic appeared to be a feminist organization of some kind and employed terminology
such as "choice" and "option" accordingly, which is often considered pro-choice
terminology by CPC clients (Entsminger, 2005). As the word "clinic" was considered a
strong identity marker that would even override weakly distinct markers such as "crisis"
or "birth," organizations were categorized as strong mimics even if these words were
additionally present.
Due to the difficulty of creating a coding scheme that could be applied
consistently over a variegated range of labels in order to capture nuances in the
152
identity positioning of CPCs, a rigorous procedure was followed to assess the reliability
of the measure. First, Krippendorff's alpha was selected as an index for reliability
estimations as it is a conservative measure that corrects for chance agreements (Hayes,
2005). As a chance-corrected type of index, Krippendorff's alpha is not only suitable for
assessing reliability in situations where content is measured on a nominal scale, but it
also accounts for the special characteristics of ordered ratings (Krippendorff, 2004a,
2004b). Hayes (2005) wrote an implementation of the estimation logarithm for the
software package SPSS. Hayes's script was used to calculate Krippendorff's alpha
adjusted for an ordinal rating scale and two coders. A desired alpha value of at least .80
was determined before reliability test coding began.
As a first step, 40 labels were randomly selected for a pilot test. None of the units
included in the pilot were label exemplars used for training the coder. After coder
training, the units were coded by one judge and the researcher. Next, Krippendorff's
alpha was calculated. As the value for Krippendorff's alpha for this test exceeded .95, no
further training was deemed necessary and two judges (one of them was the researcher)
proceeded to code all organization names that occurred in the NCCS data over the years.
After combining the ratings for units from the pilot test with all remaining ratings, the
judgments of both coders were compared and 278 disagreements as well as 2 missing
ratings were determined, which resulted in an overall Krippendorff's alpha of .91. 262 of
these disagreements were discussed and subsequently resolved. As 16 disagreements
remained and could not be resolved, the corresponding labels were assigned 8 judgments
of each judge randomly for consistency purposes.
153
The coding system was applied to data listed in the NCCS database after spelling
corrections and typographical errors had been corrected. Some organizations changed
their names multiple times, but due to the extreme rarity of more than two name changes,
only the first three organization names and the first two label changes were documented.
A naming transition event was recorded for the year in which an organization in the
dataset changed its label.
As all labels were classified according to the four ranked categories, it was
possible to differentiate between three types of transition events: (1) A mimicry transition
was recorded when an organization switched from a label classified as type 1, 2, or 3 into
a higher category. (2) A within-category transition was documented when there was a
genuine change of names, but both labels were classified as representing the same label
category. (3) A distinct transition was noted when an organization changed from a label
categorized as type 2, 3, or 4 into a lower category. In some cases, organizations were
listed with a secondary name that warranted classification in a different category.
However, usually such secondary name listings were carried over in the NCCS data for
several years, so it was not possible to pinpoint the exact year in which the label change
had occurred. Thus, organizational labeling transitions suggested by secondary name
usage were not captured as regular label transformation events, but the information about
the organization's usage of multiple names was recorded in an additional variable.
Constructing a Measure of Cognitive Legitimacy
As mentioned previously, new scholarship argues for a conceptualization of the
achievement of cognitive legitimacy based on the discursive stabilization of form
communication among key audiences (Kennedy & Fiss, 2009). Thus, a longitudinal
154
measure of cognitive legitimacy should focus on textual data. As pointed out by Aldrich
and Baker (2001), the media play a substantive role in the legitimation discourse about
organizational forms as "stories are critically important sense making tools for most
people" (Rosa & Porac, 2002, p. 516). Press coverage is a good starting point for form
evolution analyses (Mazza & Alvarez, 2000) as it constitutes a vital source for acquiring
information about organizations (Deephouse, 1996), particularly if they are involved in
social movements (Gamson, 1995). An emphasis on media sources allows for surveying
all discourse types that are circulating about a form, which is advantageous as cognitive
legitimacy analyses should not focus on specialized discourse (Moore & Hala, 2002). A
corpus of newspaper stories may serve as a valid proxy for an extensive analysis of
different discursive arenas composed of the organization's audiences as described by Hsu
(2006b; 2005) and others (D. S. Meyer & Staggenborg, 1996) given that it constitutes a
collection of texts authored by many different writers. While these arenas cannot be
studied independently with such an approach, it is still possible to make sure that the
article types selected capture several of these audiences.
An important prerequisite for designing a suitable approach to the analysis of
cognitive legitimacy over time was that it should not simply rely on the overall number of
articles covering crisis pregnancy centers. While these "story-counting techniques"
(Papadakis, 1996, p. 143) are popular (cf. Lounsbury, et al., 2003; Rao, et al., 2003) due
to their ease of application, they are not fine-grained enough to track the discursive
evolution of cognitive templates. In addition, in the case of a search within the
LexisNexis article database, increasing numbers of newspapers are an unreliable means
of measuring that attention to a subject has increased substantially as the usually
155
observed yearly growth over time may simply be a by-product of the increasing number
of publications indexed in a specific source (J. Matheny, personal communication,
August 16, 2010).
A second requirement for the analysis was that it should be able to trace the
evolution of form discourse as it revolved around the features, structures, and identity
demarcations of forms, thus capturing all aspects that have been traditionally defined as
salient for form definitions. Hannan (2005) notes that "the codes that specify an
audience's default expectations can be considered as semantic objects, and the texts that
discuss these organizations can be studied directly as a step in identifying organizational
forms" (p. 66). Accordingly, the semantic content of newspaper coverage as
corresponding to the three definitional approaches to organizational forms can be
analyzed in order to take the salience of feature-based, structure-based, and identity-
related concepts into account. Further, the measurement of the cognitive legitimacy of a
population potentially engaging in mimicry should focus on an examination of those form
aspects that distinguish mimics from their models and compare them with form aspects
that are shared by both. It should be mentioned that in light of the ongoing mimicry
observed among CPCs, the overall expectation was that cognitive legitimacy would be
relatively stable over time and that it would increase only slightly, if at all. In the
following, the individual steps of the analysis are outlined in more detail.
As a first step, sampling procedures were designed with which stories about CPCs
could be captured. Similar to the difficulties encountered in the process of identifying
potential organizations, the detection of stories about CPCs needed to take into account
that they were operating under a highly differentiated and ambiguous range of
156
organization labels. As the purpose of the analysis was not to determine the degree to
which the label "crisis pregnancy center" had stabilized, but rather the degree to which
CPCs "clothed" in all kinds of their usual labels were recognized as a distinct category, it
was deemed necessary to cast a wide net. Just like some CPCs could not be differentiated
from RHPs by examining their labels alone, there had to be some newspaper stories in
which the demarcation between the two forms was left unspecified as well. It is exactly in
the nature of ambiguous organization labeling that readers of stories will find it difficult
to differentiate between a CPC and a RHP. Accordingly, five different search expressions
were designed with could identify articles featuring a variety of common CPC
organization labels. The exact expressions for the searches are listed in Appendix B. The
searches were conducted to cover yearly segments from 1970 to 2009 and were combined
afterwards in order to circumvent restrictions on the number of possible retrievals per
search. A typical search term included variants of the word pregnancy in vicinity to
common components of a CPC label. Search 1 listed in Appendix B was, for example,
able to capture stories featuring distinct labels such as "Birthright of Kokomo" and
"SAV-A-LIFE of the Pearl River Area." Search 4 was intended to identify articles
referring to mimicry labels such as "New Dawn Women's Clinic" and "Women's
Resource Centers of Jacksonville."
As a next step, duplicate entries were removed from the search results, which
constituted between 5% and 10% of the search results. Additionally, obituaries and
calendar items were excluded. Obituaries usually only referred to CPCs in brief without
supplying any substantive information about the nature of the organizations. Typical
mentions of CPCs in obituaries included that the deceased had served as a volunteer
157
at a CPC or that donations should be made to such an organization. Social, church, and
community calendars were excluded due to the fact that they tended to be atypically long.
They usually listed a variety of activities offered by many organizations, but they only
featured references to CPCs in very small segments. Also, calendars were not listed at all
in earlier years, but their occurrence exploded in frequency during later years, occupying
up to 40% of a year's search results, which threatened to dilute the analysis. Opinion
pieces such as letters to the editor, however, were retained as they were particularly
valuable in capturing the attitudes of individual audience members about these
organizations.
After the exclusion of articles as described above, coverage from the years 1970
to 1980 yielded none or only a very low number of articles, so these years were removed
from further analysis. As demonstrated in previous research (Abrahamson, 1997;
Lounsbury, et al., 2003), the absence of coverage of a cognitive template such as an
organizational form implies that its cognitive legitimacy levels are low, which was not a
surprising result for the first decade of the existence of CPCs. As an intermediate step,
the content analysis software package WordStat was used to determine the frequency
with which articles referred to the four major types of CPC subforms. For this purpose,
categories were created which contained the various label strings and frequency analyses
were run. The results are displayed in Figure 7 in chapter 5 and demonstrate the
fluctuations in the number of articles featuring a specific type of label.
The bulk of the remaining analysis was conducted in AutoMap, which is a
network analysis application with which mental models can be extracted from texts
(Carley, 2010a). Articles grouped by year were fed separately into AutoMap and
158
subjected to a pre-processing routine. First, symbols such as asterisks, apostrophes, and
percentage signs were removed automatically. After the removal of punctuation and
numbers, all words were converted to lowercase, which is a common data cleaning
procedure in AutoMap (Carley, 2010a). In order to reduce the number of typical high-
frequency concepts occurring in the texts, a standard delete list available in AutoMap was
applied. All word strings listed in a delete list are removed from the text files, which
purges the texts from words such as "a," "and," and "our." As a next step, K-stemming
was applied, which is a procedure that converts plural forms into singulars and past
tenses into present tenses and removes "ing" endings from verbs (Carley, 2010a).
In order to reduce the texts to concepts deemed important to features, structures,
and identity-related elements that were either distinct for CPCs or overlapped with those
of RHPs, a generalization thesaurus was constructed next. A thesaurus file always
requires two columns, one listing the strings that are supposed to be combined or
converted, and the other designating the resulting concept. If applied exclusively, a
generalization thesaurus purges texts from all concepts not listed in it and converts those
specified in its left column into higher-level concepts. In order to design the
generalization thesaurus, literature about and from CPCs and RHPs was examined
extensively. Additionally, frequency analyses were run and all words and phrases
occurring at least three times in a given year were inspected in terms of their relevance to
describing CPC and RHP organizational forms. A total of 302 concepts and phrases were
identified and used to create an exclusive generalization thesaurus. These concepts and
phrases represented feature-, structure-, and identity-based references that either served to
distinguish a CPC or highlight its similarity with RHPs.
159
An important feature of a generalization thesaurus is that its individual rows are
processed sequentially, so the ordering of rows plays a role. The original phrase "doesn't
refer for abortion," for example, will be turned into "doesn t refer abortion" in the process
of pre-processing. The row specifying the substitution "doesn t" with "does not" in the
generalization thesaurus will then convert "doesn t" into "does not." As a last step, the
phrase "not refer abortion" is transformed into "not_refer_abortion" by a row further
below. Due to this sequential processing of rows in a generalization thesaurus, it is
possible to transform expressions such as "doesn't refer for abortion" or "won't refer to
abortion providers" into the phrase "not_refer_abortion." Appendix C provides the
generalization thesaurus used.
Due to the strong semantic reduction of texts that is inherent when they are
condensed to a 300-item dictionary, only those years were retained for analysis that had
at least 100 individual articles. The first year to meet this requirement was 1987, so the
remainder of the analysis focuses on the period between 1987 and 2009. An important
strategy for the construction of the thesaurus was to include phrases that foregrounded
differences between RHPs and CPCs by using negations. Such explicitly articulated
"negative" or absence links (Carley & Kaufer, 1993; Mueller, 2008) are very
consequential for articulating form distinctions. For example, all phrases in texts that
mentioned explicitly that organization X was not an abortion provider were combined
into the phrase "no_abortion_provider," which thus exemplifies an expression that serves
to emphasize an identity distinct from RHPs. Semantically, it is of fundamental
importance to differentiate between the mere absence of a linkage or an articulated "non-
linkage" (Mueller, 2008).
160
As a next step, a meta thesaurus was constructed in order to associate the concepts
and phrases specified in the generalization thesaurus with the higher-order ideas
represented by them. A meta thesaurus helps to cluster linguistic expressions according to
their membership in larger semantic categories. In the context of the project, the premise
of constructing such a textual inventory was to identify the most salient features,
structural elements, and identity-based words and phrases used to describe the CPC
organizational form in contrast to the one of RHPs. As mentioned previously, all three
dimensions capture aspects about a form that matter to audiences. Three clusters of the
meta thesaurus referred to distinct features, structures, and identity claims of CPCs, and
another three clusters were based on features, structures, and identity claims typical for
RHPs. The differentiation into the six clusters was also helpful as it already guided the
search for words and phrases with potential relevance for the external audiences of CPCs
(Hannan, et al., 2007). Thus, the meta thesaurus, which in turn was based on frequently
occurring words and phrases in the articles as well as other literature about CPCs and
RHPs, served to sort salient concepts and phrases into the clusters, which can be thought
of as semantic "buckets." In Appendix D, an overview is provided in which the concepts
and phrases generated by the generalization thesaurus are listed in terms of their
"membership" in six semantic clusters.
The fact that audiences may differ widely in terms of what they consider relevant
components of form communication (Glynn & Marquis, 2006) posed a major challenge
for selecting salient communication about form. In the absence of questioning audiences
directly about their form concepts as derived from news exposure about CPCs, it
appeared preferable to select and categorize concepts a priori and through an iterative,
161
albeit highly subjective process. Computer-aided clustering would have yielded a meta
thesaurus based on occurrence frequencies alone and the resulting clusters would have
relied on frequent co-occurrence of concepts. The advantage of determining a priori
which concepts constructed a distinct CPC form versus those that might have as well
been associated with the RHP form was that this technique allowed for the inclusion of at
least some information about the substantive meaning conveyed by concepts in a cluster.
A priori classification allowed for examining quantitatively how the connections between
clusters evolved over time.
A quick example may serve to illustrate the process. Literature and websites often
referred to the provision of free maternity clothes at CPCs, so the phrase specifying it was
captured in the concept "free_maternity_clothes" as it appeared to provide salient and
recurring information about the CPC form. It was classified as referring to features as it
described a service provided by CPCs to their clients. As professional health care
providers such as RHPs were considered unlikely to give away free clothing to their
patients, it was decided that information about this feature was distinctive and unique for
CPCs rather than leading to them being confused with an RHP. In contrast, the concept
"treat_std" included phrases describing treatments provided to clients for curing STDs.
This appeared to be an organizational feature commonly associated with RHPs rather
than CPCs, so it was categorized as an ambiguous feature.
Together with a second judge, the researcher decided where every item should be
placed. The level of agreement between coders was not assessed formally as the whole
process was necessarily somewhat subjective and did not follow a strict coding scheme. It
was particularly difficult to determine for some items if they constituted features,
162
structures, or identity-based form aspects. However, emphasis was placed on coming to
an agreement about the inclusion as well as the placement of words and phrases as
distinct from RHPs versus ambiguous as this distinction was relevant for the ensuing
network analysis. Accordingly, all items were discussed individually before they were
assigned to a specific cluster and these discussions resulted in agreement in all cases.
Underlying the idea of a network conceptualization of texts is the notion that texts
do not appear in isolation, but have to be appreciated as elements of an intertextual
network (Cerulo, 2002; Monge & Poole, 2008). By examining the embeddedness
structure of concepts making up such an intertextual network, the stability of
argumentative networks can be inferred and the discursive underpinnings of
organizational forms can be examined (Weyer, 1989). The evolution of the co-occurrence
connections between higher-order meta concepts provides information about the
changing discursive structures describing a form. After the generalization thesaurus was
applied, an undirected meta network was extracted that was based on the application of
the meta thesaurus described above. In a meta network, concepts and phrases constitute
the nodes and their co-occurrence is regarded as a link. Meta nodes are clusters of
concepts and phrases associated with the same meta concept. For the extraction of the
meta network, the window size for detecting co-occurrences was set to 10 and allowed to
slide over an entire article. In other words, two concepts or phrases were considered as
co-occurring if they were listed in the same article and not separated by more than 8 other
concepts or phrases.
Based on the application of the pre-clustered meta thesaurus, individual meta
networks were produced that represented the discursive constellations of a given year.
163
The structure of the co-occurrence network between individual concepts and phrases is
left intact when a meta thesaurus is applied. The networks were analyzed in the software
*ORA, which is designed for the purpose of gathering network measures associated with
meta networks generated in AutoMap (Carley, 2010b). Due to the rising number of
articles per year, the meta networks tended to feature an increasing number of co-
occurrence links and the overall number of individual concepts in a given year also
increased slightly over time. However, the overall network size rose only moderately,
which suggests that the discourse about CPCs had not gained much in complexity. In the
year 1987, for example, 134 individual concepts were present, and in the year 2009, the
number had merely increased to 195.
Table 6 shows that the overall density of co-occurrence linkages increased as
well, but not equally throughout the network. For the purpose of determining cognitive
legitimacy, the main comparison focused on differences between the distinct versus the
mimicry-related co-occurrences. The specific classification of a concept as feature,
structure, or identity-based form aspect did not flow into the calculation of densities
reported in Table 6. A high density of linkages within a semantic cluster indicates that its
concepts co-occur frequently, which aids in the development of that cluster as a well-
recognized cognitive category. Therefore, an increase in the density of the distinct
semantic cluster, which constitutes a collection of concepts and phrases that emphasize
the distinct identity, features, and embeddedness patterns of CPCs, indicates a rise in the
cognitive legitimacy of the CPC organizational form. Stories that explain the special
network connections, activities, and characteristics of a crisis pregnancy ministry are
164
typical examples for coverage that causes the density within the distinct cluster to
increase.
Table 6
Descriptive Statistics of Meta Networks Derived From News Coverage About CPCs
Network density
Year
Total
number of
articles
Number of
concept co-
occurrences
Mimicry
cluster
Distinct
cluster
Intercluster
1987 127 3322 .253 .211 .141
1988 126 3530 .199 .238 .136
1989 295 6606 .271 .258 .196
1990 195 4088 .231 .214 .149
1991 258 5646 .252 .213 .172
1992 350 6678 .274 .261 .182
1993 412 7218 .269 .288 .196
1994 557 9318 .302 .315 .234
1995 679 9564 .319 .370 .237
1996 643 9090 .294 .362 .225
1997 669 9326 .281 .353 .215
1998 762 11778 .342 .370 .267
1999 830 11054 .308 .356 .245
2000 887 11286 .318 .406 .262
2001 857 11380 .311 .366 .250
2002 844 12004 .328 .370 .286
2003 689 11494 .333 .355 .276
2004 737 10158 .277 .380 .239
2005 854 11986 .323 .389 .279
2006 845 13032 .313 .405 .297
2007 731 11594 .348 .329 .263
2008 698 10964 .300 .357 .253
2009 725 10956 .290 .371 .251
Note. Prior to 1987, the annual number of articles was below 100.
165
Further, a higher density of cross-linkages between two semantic clusters suggests
that theorization (Rao, et al., 2003) is taking place more frequently. Cross-theorization is
consequential as it aids in the articulation of the boundaries between two semantic
clusters. It is based on the notion that the boundaries between social forms are enacted
relationally, which is one idea that was formulated by Tilly (2004, 2006) and others.
Viewed from this perspective, the distinctive form qualities encapsulated in a cognitive
cluster are enhanced by contrasting them with their semantic "neighbors." This
juxtaposition is achieved by stories that establish cross-cluster connections. In the context
of the cognitive legitimacy of the CPC organizational form, an increase in the density
between the distinct cluster and the mimicry cluster signals that frequent discursive
connections between distinct pro-life concepts and mimicry features, structures, and
identity elements are created. Reports about CPCs posing as mimics of health care
providers exemplify such cross-theorization if they simultaneously identify them as
members of the pro-life community. The result of these articulations is that the
differences between the distinct and mimicry clusters are pronounced. Thus, they are a
vital discursive indicator for an increase in the recognizability of an organizational form.
A visual inspection of the columns displaying the density values associated with
various clusters in Table 6 suggests that both the densities within the distinct cluster and
the intercluster densities connecting distinct and mimicry clusters appeared to have
increased over time. In contrast, the densities in the mimicry cluster remained fairly
stable, which indicates that CPCs employing a variety of labels indeed gained in
recognizability. The same patterns were detected when the six smaller clusters were
examined, which lends support to the assertion that the recognizability of the CPC
166
form has been increasing. The densification of cross-theorization becomes visible when
the meta concept networks are plotted over time. Figure 8 shows how the linkages
between pro-life specific elements and those associated with RHPs tended to increase in
later years. Most noticeably, a bulk of these connections was created as distinct pro-life
identity markers tended to co-occur more frequently with RHP-typical features such as
STD testing or ultrasound. This means that over time, the discursive connections between
pro-life identities and the increasingly frequent practice of CPCs to offer medical services
did in fact become stronger.
Figure 8: Meta network representations of form discourse about CPCs over time.
Note. Node size is scaled by number of concepts. Visible links represent > 10% of total links.
167
In order to further examine the relationship between cognitive legitimacy and the
vital rates of CPCs, the densities within the distinct cluster and between the two clusters,
which represent the two related aspects of cognitive legitimacy, were combined into a
single summary indicator. Based on correlation analyses that showed satisfactory
correlation (Pearson's r = .91, t(21) = 10.07, p < .001 [two-tailed]) between the two
density values throughout the years 1987 to 2009, a reliability test for the combination
measure was conducted, which yielded a Cronbach's α of .93. A high level of Cronbach's
α is desirable as it suggests that the internal consistency of the summary indicator is high
and that it constitutes a reliable composite of individual indicator variables measuring
dimensions of cognitive legitimacy.
Sociopolitical Legitimacy Operationalization
As a number of studies on public opinion and ideological variation within the
United States have shown, there is considerable variation between individual states when
it comes to their policy climates (Brace, Arceneaux, Johnson, & Ulbig, 2004).
Additionally, W. D. Berry et al.'s (1998) longitudinal measures of political ideology
suggest that there is some temporal variation that should be taken into account. Thus, a
measure of the sociopolitical legitimacy climate towards CPCs needs to differentiate
between states and reflect change over time. A second requirement was that the
instrument be specific enough to serve as an indicator for the policy climate in respect to
legal access to abortion. In other words, it should, at a minimum, inform about the overall
state climate regarding abortion rights given that state-level policies unfavorable towards
abortion signify a policy environment favorable to CPCs. The NARAL abortion policy
index is a combination indicator that includes a variety of such regulatory measures,
168
which made it an ideal measure of the sociopolitical legitimacy environment surrounding
CPCs.
For the purposes of this study, Nicholson-Crotty's (2007) strategy was followed
and the NARAL policy index was used for the years in which it was available. Starting in
the year 1999, the NARAL policy index could be obtained from the annual Who Decides
reports published by NARAL. For the years 1989 to 1999, NARAL provided only the
state-level information about legislative enactments necessary to construct the index
based on manual coding. As the longitudinal analysis required measurement of
sociopolitical legitimacy for earlier years than provided by NARAL, manual coding of
legislative information for periods before the NARAL index was available became
necessary. The coding procedures of NARAL were followed as closely as possible in
order to ensure high compatibility levels between the manually coded years and the years
provided by NARAL. As NARAL information did not go back to years prior to 1989, an
abortion policy measure constructed by Halva-Neubauer (1993) was used to supplement
the yearly measurements for that period. The Halva-Neubauer (1993) legislative
measures count index captures the number of abortion laws enacted in a state. Due to its
similarity with the NARAL index, it exhibited high compatibility with measures collected
for other years. Sufficient compatibility was concluded based on the pairwise correlation
between the Halva-Neubauer measure prior to 1989 and the 1989 measure constructed
based on NARAL data. Pearson's r equaled .67 with t(48) = 9.59 and p < .001 (two-
tailed).
The NARAL policy index is based on counting legislative measures that indicate
the state-level reproductive rights climate in currently up to 14 different categories.
169
These include restrictive laws designating comprehensive pre- and post-Roe abortion
bans, bans on abortion procedures, conscience-based exemption, counseling bans,
husband consent, waiting periods, mandatory informed consent, insurance coverage
prohibitions, restrictions to minors' access to abortions, physician-only requirements,
post-viability bans, provision restrictions for public employees and/or facilities, and laws
restricting public funding of abortions. States with low scores on the index display very
few policy restrictions when it comes to accessing abortion services whereas states with
high scores exhibit a complex and restrictive policy environment. Regulations aiding
access to abortions included in the measure are laws prohibiting clinic violence, which
are reverse-coded in order to reflect their positive effects on RHPs.
Over the years, additional legal categories have been gradually added to the index
as the variety of laws with which reproductive rights are regulated has increased. These
fluctuations are reflected in alterations to the calculation of the NARAL policy index. In
the year 1989, for example, the only laws available and coded for the index included
post-viability bans, minors' access restrictions, and public funding regulations. Between
1989 and 1997, the number of legal dimensions increased from 3 to 14, which meant a
considerable extension of the number of categories and at the same time, a large increase
in the overall score calculated based on them. Between 1999 and 2004, the scoring
system changed only very little, allowing for a total of 150 points and including all 14
indicators. This consistency alleviated the adjustment of manual coding prior to 1999. In
2005, larger changes were introduced as contraception coverage and so-called TRAP
laws were added. TRAP is an acronym designating the targeted regulation of abortion
providers. Trap laws include regulations that pose specific restrictions on the way in
170
which abortion services may be performed. Due to the gradual addition of categories,
average NARAL policy index scores and their spread overall have been increasing over
time. While the mean state score in 1989 equaled 23.76 with a standard deviation of
10.61, it had climbed to 50.41 points (SD = 91.01) in 2009. In order to correct for this
increase in variability, the spread of scores below the value of zero, and fluctuations due
to combining indices from different sources, the final annual state scores for
sociopolitical legitimacy were obtained in a two-step procedure. First, all scores were
transformed into z-scores. Second, based on their percentiles, these z-scores were
transformed into a ranked variable with four categories.
Measures
The founding rates of CPCs, their label transformations towards mimicry, and
their disbandings served as dependent variables for several statistical models. In order to
differentiate CPCs pursuing mimicry strategies from those selecting distinct labels, two
dummy variables were created that were used to stratify the CPC population accordingly.
Organizational founding. The founding date of a CPC was determined based on
the organization's ruling date, which constitutes the year in which the organization
registered with the IRS as a tax-exempt entity. Foundings of mimics versus non-mimics
were differentiated based on the name choices at first appearance. Founding events were
categorized as mimic foundings if organizations entering the population had selected
labels of the types "strong mimic" or "weak mimic." They were classified as non-mimic
foundings if the names organizations had chosen upon entry were of the types "strong
distinct" or "weak distinct." The first two types correspond to label categories 3 and 4
whereas the last two types pertain to label categories 1 and 2.
171
Label transformation. Organizational name changes towards mimicry were
registered for all label transformations in which an organization switched for the first
time from the name categories 1, 2, or 3 into a higher category, regardless of the degree
of change. Subsequent additional category switches into a higher category were very rare
and were not considered for further analysis.
Organizational failure. The year following the CPC's last appearance in the
BMFs or Core files was considered its year of dissolution. As previously outlined, failure
determinations were based on a variety of conditions accounting for the differences
between the file types in terms of their sensitivity to organizational failures. For example,
as it took several years until defunct organizations disappeared from the BMFs, in certain
cases, their disappearance from Core files was used to determine their year of
disbandment.
CPC mimic. For Hypothesis 4g, which examines the failure rate of mimics only,
mimics were operationalized as organizations that had either been born into a mimicry
name category (types 3 and 4) or gone through a mimicry name transformation at some
point after their founding, which involved switching from types 1, 2, or 3 into a higher
category. A dummy variable was used to distinguish the group of CPCs comprising
"born" and transformed mimics (CPC mimic = 1) from those CPCs pursuing distinct
labeling strategies (CPC mimic = 0). The same dummy coding was applied to calculate
yearly density variables describing the numbers of CPC mimics that existed on the
national, state, and county levels.
CPC "born" mimic. An additional variable was created to further distinguish
specifically between the two ways in which organizations were engaging in mimicry.
172
Thus, similar to Usher and Evans's (1996) specification of form speciation among
gasoline alleys, selective as well as adaptive pressures towards mimicry could be
captured. "Born" mimics, organizations that entered the CPC populations by choosing
mimicry names of label types 3 and 4 (CPC "born" mimic = 1) were differentiated from
CPCs that actively pursued a mimicry labeling strategy at some point in their existence
by switching into a higher label category (CPC "born" mimic = 0). Hypotheses about the
founding rates of CPC mimics considered only "born" mimics in order to avoid a
misspecification of these models (Yamaguchi, 1991). It would be incorrect to use
mimicry transitions in later life to distinguish between organizations at their time of
founding.
Independent variables used for analysis included several legitimacy indicators,
which were measured annually on the national or state levels. Another group of
predictors comprised measures of environmental resource availability, which were
measured on the levels of the county or the nation. County-level indicators were time-
invariant measures whereas the country-level indicator varied per year. As county-level
resource indicators were based on county population numbers, they depended on data
collected in the context of the decennial U.S. Census in 1990 (church membership) or
2000 (women's population). In contrast, the data on charitable giving to NGOs are
provided annually (Giving USA Foundation, 2007).
A third type of annually changing independent variables was based on the
organizational densities of CPCs and CPC subpopulations measured on the national,
state, and county levels. Additionally, densities of RHPs on the national and county levels
as well as a proximity measure indicating the co-presence of an RHP in the same
173
county were included in some analyses. The distributions of all density measures were
checked for skewness, but as they appeared normal, no log transformations were
considered necessary. Incorporation dates were used to stratify organizations into four
founding cohorts and a measure of size based on organizational assets was constructed as
a second organization-level attribute. Two time-invariant variables were constructed on
the county-level in order to control for the location of CPCs in metropolitan regions or
counties characterized by a high degree of urbanization.
Cognitive form legitimacy. As described previously in more detail, the measure of
the legitimacy of the CPC form was calculated annually and consisted of a summary
index combining two semantic network density measures. In concordance with all other
exogenous indicators, cognitive form legitimacy was lagged by two years.
Sociopolitical legitimacy. Sociopolitical legitimacy was measured yearly for
individual U.S. States and the District of Columbia and lagged by two years. Categorized
z-scores of NARAL abortion policy index scores were used to differentiate between
states along four degrees of legitimacy. Higher numbers denote a more favorable policy
environment for CPCs as they correspond to a state's degree of abortion policy
restrictiveness. For years prior to 1989, the values for D.C. were missing as they are
based on the Halva-Neubauer policy index (Halva-Neubauer, 1993), which does not
include data on D.C.
Presidential abortion stance. Similar to the operationalization introduced by
Nownes (2004), who examined sociopolitical factors influencing homosexual social
movements, a dummy variable was constructed indicating a president's declared position
on the issue of abortion. During the observation period, all presidents had given
174
public statements about their position. Pro-life abortion stances received values of 1
whereas pro-choice abortion stances were coded with 0. The variable was lagged by two
years.
Human resource availability. Data from the ARDA were used to determine a
county's fraction of the adult population 14 years and older with declared membership in
a religious congregation. As described previously, this operationalization was based on
the consistent finding in previous research that CPCs were drawing heavily on support
from religious networks originating in local congregations. The calculation of the
percentage of church members was based on the 1990 ARDA study combined with
county population data collected for the 1990 U.S. Census. The variable was entered as a
time-invariant covariate on the level of a CPC's county.
Client availability. The construction of an indicator for the potential demand for
RHP and CPC services was derived from female population statistics collected for the
2000 U.S. Census. A county's total number of females ranging in age from 15 to 45 was
log-transformed to control for skewness and served as a time-constant variable on the
county level.
NGO financial resources. In order to measure the influx of financial resources in
to the NGO sector overall, the same strategy as employed by Minkoff (1999) was
pursued and annual Giving USA statistics about donations to charitable organizations
were used (Giving USA Foundation, 2007). These statistics were available on the country
level and provided in billions of current U.S. Dollars. The yearly values were lagged by
two years.
175
National, state, and county CPC densities. Given that competitive processes may
differ depending upon which geographic region is selected for density measures (Cattani,
Pennings, & Wezel, 2003), it seemed appropriate to examine density effects on several
levels. Based on the NCCS data, yearly count variables were created that reflect the
density of CPCs at the level of the country, the state, and the county.
National, state, and county CPC mimic densities. The variable CPC mimic was
used to calculate annual counts of "born" and transformed mimics on the national, state,
and county levels.
County CPC non-mimic density. The measure for the density of non-mimic CPCs
on the county level, which was used to test hypotheses about organizational failures, was
restricted to organizations not pursuing a mimicry strategy by means of adaptation or
selection (CPC mimic = 0).
National, state, and county mimic foundings. All variables describing the number
of mimic foundings were based on tallying up those founding events in which a new
population entrant chose a label of types 3 or 4. Similar to all other count variables
indicating the density of CPCs, they were entered as annually changing covariates lagged
by two years.
Weighted national and county RHP densities. Both density measures of RHPs
were derived from the AGI database and weighted according to the organization sizes of
RHPs. For years in which no data were available about the number of RHPs, one-
dimensional piecewise linear interpolation methods were used to fill the gaps. AGI count
data were grouped according to five different size categories based on the yearly number
of abortions performed by RHPs, These categories differentiated between RHPs
176
providing 1 to 24 abortions, 25 to 94 abortions, 95 to 394 abortions, 395 to 994 abortions,
and more than 995 abortions per year. Weighting was computed linearly, so a factor of
five was assigned to the biggest organizations and a factor of one to the smallest
organizations. In other words, whereas the smallest RHPs providing between 1 and 24
abortions per year were counted once, the largest RHPs were counted five times in all
computations conducted for calculating weighted RHP measures. The same principle
applied to the remaining three categories, which were counted twice, three times, or four
times depending on their rank.
RHP proximity. The proximity of an RHP was measured with a dummy variable
set to 1 for counties with at least one abortion provider. In transformation and failure
analyses, which focused on organizations as the units of analysis, it referred to
organizations founded or active in a county with at least one abortion provider. It was set
to 0 for counties without an RHP as well as for organizations located in these counties
and lagged by two years.
CPC cohorts founded 1980-1989, 1990-1999, and 2000-2009. CPCs were
stratified according to their incorporation decade in order to construct dummy variables
assigning organizations to one of four founding cohorts. Organizations founded between
the years 1970 and 1979 received a value of 0 for all three decade dummies, thus serving
as the reference category for various statistical analyses.
Organization size. Following prior research that has used information about
organizational assets as a proxy for the size of NGOs (Twombly, 2003), annual size
measures of CPCs were calculated based on a variable provided by the NCCS database
denoting the organization's total assets at the end of the year. Assets were listed in
177
currency, so their logged version was used to correct for overdispersion. However, as
values for many organizations were missing, it was not possible to use this variable in
any of the models, particularly as preliminary analyses showed that missing values
appeared not to occur randomly but rather reflected a group of organizations that was on
average smaller in size. Given that very small organizations are not required to disclose
their financial information and often do not file IRS Form 990, there seemed to be a
patterned relationship between the absence of asset records and organization size. In
order to conduct at least a descriptive analysis comparing large and small organizations in
terms of their survival rates, an organization's asset values were averaged over all years
with data and CPCs were classified as large (organization size = 1) or small
(organization size = 0) based on a median split.
Metropolitan area CPC and county. As abortion providers are mainly
concentrated in metropolitan areas (Doan, 2007), a control variable was created in order
to determine if the location of a county (for founding analyses) or an organization (for
transformation and failure analyses) fell within a metropolitan region. Metropolitan
regions are characterized by high population densities and typically refer to areas with
many cities of moderate to large size. The determination was derived from the
metropolitan state area (MSA) classification system used by many U.S. governmental
agencies. Counties and organizations located within an MSA were coded with values of 1
and all others received a value of 0.
Urban population in CPC's county. Urbanism is considered an influential
demographic predictor for abortion demand and the location choices of RHPs (Doan,
2007), so a second location-dependent control variable was created. Following the
178
operationalization pursued by J. D. McCarthy et al. (1988) in their study of local groups
opposing drunken driving, a variable was constructed based on the percentage of a
county's population classified as urban. This percentage value was assigned as an
attribute to organizations or counties, with higher values denoting higher percentages.
Analyses
Hypotheses examining CPC foundings and CPC mimic foundings conditional on
a vector of covariates were tested with a number of negative binomial regression models.
Hypotheses focusing on transformations and failures of organizations were estimated
using transitional hazard models. All statistical modeling was performed with version 11
of the statistical software package Stata.
Founding Rate Analyses
Negative binomial regression models constitute a type of count model that is a
special case of the Poisson regression, a modeling approach commonly recommended for
the estimation of founding events (Minkoff, 2002). Founding processes are typically
studied at the population level, but it is also possible to choose geographical entities such
as states (Swaminathan, 1998), regions (Simons & Ingram, 2003), metropolitan areas
(Okamoto, 2006), or counties (Sommerfeld, 2008) as units of analysis at risk for
population entries (Hannan & Freeman, 1989; Twombly, 2003). In all cases, individual
organizations' arrivals are counted as events in a continuous counting process. Depending
upon the resolution of the data, organizational ecologists often examine population entry
counts per analysis year and assume "a constant rate of founding with log-linear
dependence on covariates" (Swaminathan, 1998, p. 395), even though the exact order of
foundings during an episode is unknown.
179
The disadvantage of selecting the population as the unit at risk for experiencing
foundings as an event is that it is very difficult to include a variety of both temporal and
spatial predictors into the resulting models. As mimicry is a coevolutionary tactic which
assumes some spatial proximity between model and mimic, the added modeling
flexibility afforded by choosing U.S. counties as the entities at risk for experiencing
founding events seemed preferable even though it required the assembly of a very large
data file. Accordingly, the hypotheses examining CPC foundings conditional on a vector
of covariates were tested on the county level. Given the small size and local reach of
CPCs, grounding the analysis at the level of the county proved very beneficial for
capturing local effects and allowed for the inclusion of a range of county-specific
covariates.
The Federal Information Processing Standard (FIPS) classification of counties
facilitates longitudinal observations of county-level processes over time as it is a fairly
stable geographical categorization system (Sommerfeld, 2008). Both data on CPCs as
well as RHPs were available categorized by a county's FIPS code, which made it possible
to compute density indicators for the analyses that represented the national, state, and
local density counts. All counties in the 50 states of the U.S. and the District of Columbia
were entered into the risk set for experiencing the founding of a CPC. Due to the
governmental restructuring of county boundaries in the late 1990s, the number of
counties changed slightly during the study period, which included annual observations
from 1989 to 2009. As both CPCs and RHPs existed in some counties that were no longer
captured by the geographic county codes determined according to the FIPS, the total
180
number of counties included in the analyses pertains to 3144 counties rather than the
3142 counties currently indexed by FIPS.
As mentioned previously, the likelihood of the occurrence of annual counts may
be modeled as following a Poisson process. The corresponding probability function
describing it can be stated as follows:
€
Pr Y
t
= y
t
( )
=exp −λ
( )
λ
t
y
t/y
t
!
The number of founding events occurring in year t is specified with Y
t
. The
relationship between the founding rate λ
t
and the covariate vector x
t
takes the following
functional form:
€
lnλ
t
=α +βx
t
However, Poisson regression may only be applied without the risk of
misspecification if the conditional variance of the number of event counts does not
exceed its mean. If Var(Y
t
) > E(Y
t
), overdispersion exists and the assumption underlying
Poisson regression is violated. The use of Poisson regression for the analysis of
overdispersed data leads to the underestimation of standard errors, which means that
significant effects may be identified erroneously (Dobbin & Dowd, 1997). If count data
are overdispersed, which happens frequently due to "contagion across events within
spells" (Dobbin & Dowd, 1997, p. 519), negative binomial regression should be used
instead of Poisson regression as it specifically includes an overdispersion paramenter
which follows a gamma distribution (Swaminathan, 1998). In a negative binomial
regression, the adjusted specification of the relationship between the founding rate and
the vector of covariates can be written like this:
181
€
lnλ
t
=α +βx
t
+ε
The overdispersion parameter now included follows a gamma distribution. While
there are several ways to determine if overdispersion is a problem and the mean number
of foundings exeeds the variance of foundings per observation period, the choice of
negative binomial regression for the purpose of this study was based on an inspection of
the overdispersion parameter α provided by Stata. First, both Poisson regression and
negative binomial regression estimates were calculated for any given model. Second, a
likelihood ratio test statistic was examined in order to determine if the overdispersion
parameter associated with the model was significantly different from zero. Nested model
likelihood ratio tests are suitable for this purpose as a negative binomial regression
reduces to a Poisson regression if the overdispersion parameter equals zero (Kuilman &
Li, 2009). As these tests were significant for all models examined, it was concluded that
the data were in fact exhibiting overdispersion and that the comparison negative binomial
variant with the added parameter fitted the data significantly better than the Poisson
model.
It is also common to observe the occurrence of excessive zero-count observations
when smaller geographic locations are chosen as units for analysis in Poisson or negative
binomial regression models (see, for example, Greve, et al., 2006; Okamoto, 2006). If the
number of counts per unit often equals 0 and rarely exceeds 1, it is possible that there is a
separate underlying process responsible for such zero counts, and the statistical methods
chosen should account for it. While zero-inflated models have been used for this purpose
as they include a logit model predicting the likelihood of event occurrence overall
together with a count model estimating the number of events likely to occur, such
182
approaches should only be used if they fit the data significantly better than the baseline
models. Within Stata, the Vuong test statistic can be requested with which this fit
comparison between a standard negative binomial model and its zero-inflated version is
assessed. As Vuong tests evaluating the fit of zero-inflated negative binomial regression
variants in comparison to the baseline models proved insignificant, it was concluded that
negative binomial regression modeling without added specifications was adequate for the
estimation of the hypothesized effects.
Transformation and Mortality Rate Analyses
As hazard models are deemed an appropriate method to study the mortality
effects of organizational change (Carroll & Hannan, 2000), hypotheses about the failures
of CPCs and their transformations were tested using semi-parametric Cox regression and
parametric piecewise constant exponential regression models, respectively. Most
generally, hazard functions describe the effects of a covariate vector on the duration of
time until an event such as organizational change or death occurs. A hazard function
determining this distribution of survival time can be stated as follows:
€
h t,x,β
( )
= h
0
t
( )
r x,β
( )
As described by Hosmer, Lemeshow, and May (2008), this specification
combines two functions. h
0
(t) describes functional changes in dependence of survival
time whereas r(x,β) specifies the dependence of the hazard rate on a covariate vector. In a
Cox proportional hazard model, r(x,β) is set to equal exp(xβ). Accordingly, the hazard
function takes this form:
€
h t,x,β
( )
= h
0
t
( )
e
xβ
183
Cox regression models are popular for studying failure processes as they allow for
an inclusion of time-varying covariates and do not make specific assumptions about the
shape the distribution of survival time will take. Thus, the functional shape of the
baseline hazard rate h
0
(t) is left unparameterized and Cox regression models do not have
an intercept. However, they may only be used without risking misspecification if their
hazard rates are proportional. According to Blossfeld and Rohwer (2002), for the
transition rates r(t) and r'(t) corresponding with the values A
i
and A
i
' of the ith covariate,
hazard rate proportionality can be expressed as follows:
€
r t
( )
′ r t
( )
=exp A
i
− ′ A
i
( )
α
{ }
One common method to verify that the parameter specification of a Cox model
does not violate the proportionality assumption involves estimating Schoenfeld residuals
for all covariates, which "can essentially be thought of as the observed minus the
expected values of the covariates at each failure time" (Box-Steffensmeier & Jones, 2004,
p. 121). Proportionality can be depicted graphically by plotting Schoenfeld residuals
versus a function of analysis time, which should result in a curve with a slope of zero. In
Stata, a function based on Schoenfeld residuals returns results of a test estimating "the
null hypothesis of zero slope" (Cleves, Gould, Gutierrez, & Marchenko, 2008, p. 201).
As the results of proportionality tests relying on this function suggested that the failure
models fitted to the data on CPCs did not violate the assumption of the Cox regression
model, this estimation technique was used for examining mortality hypotheses. In
contrast, Schoenfeld residual tests indicated that hazard models fitted to data on the label
transformation of CPCs could not be estimated with Cox hazard regression methods
184
as it was detected that they were violating the proportionality requirement associated with
Cox modeling.
In cases where Cox models are not applicable as there are doubts as to the
proportionality of the hazard rates, piecewise constant exponential models can be used
instead. The advantage of these parametric models over a semi-parametric Cox regression
model is that their hazard rate is only fixed to a constant rate during an individual period,
but allowed to vary freely between periods (Blossfeld & Rohwer, 2002). The total
number of time pieces and their length can be specified a priori. Thus, the models
estimating covariate effects on the label transformation rates of CPCs were calculated in
Stata with a piecewise constant hazard rate routine created by Sørensen (1999).
As a dichotomous covariate is added for every time piece, Blossfeld (2007)
recommends limiting the number of periods estimated in order to avoid inflating the
model with an excessive number of parameters. However, their number should also
suffice to appropriately capture fluctuations in the hazard rate over time (Blossfeld, et al.,
2007). Additionally, it is important to verify that the event of interest actually occurs
during every individual time piece specified in piecewise constant models. While the
observation period for transformations, which spanned the years 1989 to 2009, was split
into annual episodes in order to accommodate time-varying covariates which were
updated yearly, preliminary tests suggested that the hazard rate variations during the 21
annual episodes were best controlled for by creating seven time pieces comprising three
years each.
185
Additional Aspects of Statistical Modeling
As the founding dates for all organizations in the sample could be established, all
left-truncated observations used for the mortality analysis had known start times. Left
truncation occurs when organizations have been at risk for an event sometime before they
enter observation. This means that organizations in the sample founded before 1989, but
still alive at that point, have been unobserved before they entered observation.
Conversely, organizations with founding and dissolution dates before 1989 never entered
into the risk set. Thus, left truncation can cause problems in the sense that of
organizations founded prior to 1989, more resilient ones are overrepresented in the
sample as some of their less fit cohort members dropped out before they came under
observation (Guo, 1993). Cox regression is well suited to accommodate left-truncated
data of this nature. Likewise, piecewise constant modeling is considered an especially
effective method of dealing with left-truncated subjects, particularly if their start times
are known (Guo, 1993). Similar to Cox hazard models, piecewise constant procedures are
also able to accommodate right-censored data well. Observations are right-censored when
the analysis time ends before the event of interest occurred, and as many CPCs had
neither disbanded nor changed their organization labels when observation ended, a large
number of cases exhibited right-censoring.
Organizational age effects on mortality were modeled indirectly by examining
contrast-coded cohort effects as Cox regression models do not allow for an inclusion of
time-varying covariates directly correlated with analysis time (Yamaguchi, 1991). In the
piecewise exponential regression models, preliminary models also included cohort-period
interactions to check for systematic time-dependent factors related to organization
186
age. However, these interactions did not appear to play a large role, so the addition of
such a high number of controls inflated the models unnecessarily without adding much
explanatory value. As multicollinearity problems were aggravated (Ruef, 2004a) by
including the cohort-period interactions, they were not entered into the final set of models
presented.
All time-varying predictor variables used for estimating foundings, label
transformations, and failures were lagged by two years in order to circumvent
endogeneity problems. Lagging organization counts, which provided the basis for many
density variables used in the models, also reduces the problem of autocorrelation created
by contagion across spells (Barron, 1992). As CPCs are small organizations, their
"gestation periods" (Hannan & Freeman, 1989, p. 121) are short, but their incorporation
as a nonprofit entity adds several months to the duration between initial planning and full
operation of a CPC. Lagging takes into account that founding processes prior to operation
include planning, acquisition of resources and office space, legal incorporation, hiring,
and training of staff (Hannan & Freeman, 1989).
Log transformations of variables such as population density or organization size
based on currency are recommended if distributions are skewed (Boslaugh & Watters,
2008; Weinberg & Abramowitz, 2002). As the U.S. Census population data exhibited
such skewness, the variable measuring client availability was calculated by logging the
absolute population number. All other count and density variables were inspected to
determine if additional log transformations were necessary, but as this was not the case,
they were retained in their original forms. Typically, population ecologists also
recommend that researchers "include a measure of density squared to test for the
187
second-order effects of competition on organizational specialization" (Soule & King,
2008, p. 1585). It is common to include quadratic density terms that are rescaled by 100
or 1000 (Baum & Singh, 1994c; Delacroix, et al., 1989; Dobrev, 2001). However,
multicollinearity problems often result from fitting models that include both main density
effects and their squared terms. One strategy of preventing the fitting of models with
collinear terms is described by Galaskiewicz and Bielefeld (1998), who suggest that
density variables should be entered together with their quadratic forms and to drop all
second-order effects if they prove insignificant. This recommendation was followed
throughout the process of hypothesis testing.
In order to compensate for occasional gaps in annually changing covariates, linear
interpolation methods were used. These are common techniques based on estimating
missing values as located on a straight geometric line between two neighboring data
points. Linear interpolation was particularly useful to complete count data on RHPs,
which was missing for some years. All models used for testing hypotheses used robust
estimation techniques. As recommended by Box-Steffensmeier and Jones (2004), robust
estimators are particularly useful in the presence of time-varying variables as researchers
can never be sure that they have not introduced some time dependence into their models.
Using robust estimators does not affect the coefficients, but it avoids underestimating the
standard errors associated with them (Box-Steffensmeier & Jones, 2004). Thus, they
represent a more conservative estimation technique that helps to prevent an overstatement
of model significance. They are preferable when it is suspected that unobserved
heterogeneity is introduced due to some clustering in the data, which is often the case for
population vital rates.
188
Overall, the construction of the models for all statistical analyses followed a
parsimonious modeling strategy. The inclusion of variables into the models concentrated
on hypothesized effects and the number and kind of variables entered into the individual
analyses varied based on the predicted relationships. A small number of control variables
were added if their influence was suspected to affect the estimation of the coefficients.
Nested models were created for the analyses of CPC foundings and transformations and
the stability of effects across models was examined. Likelihood ratio tests were used as
they allow for the comparison of two nested models in order to determine if one fits the
data significantly better than the other. While the model fit of Cox regression models
cannot be compared with likelihood ratio tests, the same additive method was used to
estimate models predicting CPC failures. In both founding analyses, Model 3 was
compared with Model 2 and Model 2 was compared with Model 1. In the piecewise
constant exponential hazard rate analyses of organizational label transformations, Model
3 was compared with Model 2 and Model 2 was compared with Model 1. Model 1 was
compared with an interval-effects only baseline model.
In all analyses, independent variables were grouped according to their types. They
were entered into a given model in the order listed in the results tables, which did not
vary between the different models belonging to the same analysis. The order of entry of
individual groups of variables was determined by the author based on their suspected
importance, with the most important groups of predictors entered first. It was not based
on specific theoretical expectations and slightly varied between founding, transformation,
and survival analyses. These variable groups included density variables representing
189
CPCs, RHPs, or both. Additional groups represented legitimacy indicators, resource
predictors, CPC cohort indicators, and one or two control variables.
190
CHAPTER 7: RESULTS
Descriptive Analyses
Since 1970, the total number of organizations within the CPC population as a
whole has been growing steadily as can be seen in Figure 9. This increase in overall
population density has been mainly due to a consistent influx of new organizations
coupled with an overall very low mortality rate. Peaks in population entries occurred in
the mid-1980s as well as from 1990 to 1995. Within the observation frame of the study,
which included the years between 1989 and 2009, 2045 unique organizations were
identified. During the same time, only 316 organizational failures occurred, which
represents approximately 15% of all organizations that underwent observation. The
incidence rate of organizational failure throughout the observation period equaled .011.
Figure 9: Total CPC population vital rates from 1970 to 2009.
0
200
400
600
800
1000
1200
1400
1600
1800
2000
1970 1975 1980 1985 1990 1995 2000 2005
Number of Organizations
Number of Incorporated CPCs
Exits
Entries
191
The spatial distribution of CPCs listed in the NCCS database between 1989 and
2009 was found to differ widely between individual states. In general, the highest
concentration of incorporated CPCs can be found in the Northern and Midwestern U.S.
states, which is shown in Figure 10.
Figure 10: Number of CPCs by state per 100,000 women of age 14 to 44.
While the NCCS data do not allow for an analysis of organizational failures prior
to 1989, it is possible based on the survivors still present in that year to trace the
development of individual subforms within the CPC population. Strong and weak
distincts, which are subpopulations composed of organizations with labels differentiating
them from RHPs, constituted the two most common CPC forms until the late 1990s.
From that time on, the growth rate of strong distincts decelerated and weak distincts
192
declined in numbers, as is depicted in Figure 11. At the same time, weak mimics, which
had been gaining in numbers from the early 1980s on, continued to grow. Strong mimics,
which represent the smallest CPC subgroup, have also expanded in numbers. Thus, the
overall proportion of strong and weak mimics among CPCs has increased, surpassing the
50% mark in the mid-2000s.
Figure 11: CPC population count stratified by subform from 1970 to 2009.
A closer look at the subpopulation entry rates in Figure 12 reveals that subform
preferences of new entrants have been changing over time and could be roughly
classified as occurring in four subsequent waves. Prior to the 1980s, organizations were
typically founded as strong distincts. During the 1980s, the numbers of organizations
entering as weak distincts peaked. Since the late 1990s, entries into the subpopulation
0
200
400
600
800
1000
1200
1400
1600
1800
2000
1970 1975 1980 1985 1990 1995 2000 2005
Number of Organizations
Strong Mimics
Weak Mimics
Weak Distincts
Strong Distincts
193
of weak mimics have surpassed those into the other subforms. Strong CPC mimics,
which commonly acquire clinic licensures to approximate RHPs more closely, constitute
the latest wave of population entrants. Their overall entry percentage is low and weak
mimic entries are much more numerous, but they are about to surpass the entry numbers
of weak and strong distincts.
Figure 12: Subpopulation entries of CPCs from 1970 to 2009.
Figure 13 depicts CPC subpopulation exits during the observation period, which
have typically unfolded in bursts. Prior to the mid-1990s, overall exit numbers were
negligible. The first exit wave occurred in the second half of the 1990s and included
strong and weak distincts as well as weak mimics. A second exit wave from the same
three subpopulations occurred around the year 2007, but this time, the exit numbers of
0
10
20
30
40
50
60
70
80
1970 1975 1980 1985 1990 1995 2000 2005
Number of Subform Entrants
Strong Distincts
Weak Distincts
Weak Mimics
Strong Mimics
194
weak mimics surpassed those of strong and weak distincts. While overall, strong distincts
represent the subpopulation with the highest number of exits throughout the decades, exit
numbers for both weak distincts and weak mimics have been rising since the turn of the
millennium.
Figure 13: Subpopulation exits of CPCs from 1989 to 2009.
At approximately the same time, the number of organizational name
transformations peaked, as is shown in Figure 14. CPCs have consistently engaged in
mimicry changes more frequently than in modifications that involved the acquisition of a
more distinct name or a label similar to their previous name. This lends support to the
claim that CPCs overall exhibit a noticeable tendency towards pursuing mimic strategies.
0
2
4
6
8
10
12
14
16
18
20
1989 1994 1999 2004 2009
Number of Subform Exits
Strong Distincts
Weak Distincts
Weak Mimics
Strong Mimics
195
Figure 14: Label transformations among CPCs from 1990 to 2009.
Life table techniques are a nonparametric descriptive method particularly well
suited for an initial examination of survival data (Blossfeld, et al., 2007). Thus, they were
used to create plots comparing the survivor functions of CPC mimics and non-mimics in
order to explore differences between the two groups. As shown in Figure 15, both
functions resemble each other at the beginning but are starting to diverge after
approximately three to five years. After roughly twenty years, the percentage of survivors
among mimics equals approximately 90% whereas only 80% of non-mimics are still in
existence. The analysis time plotted in Figure 15 exceeds the twenty-year observation as
it includes left-truncated observations, so organizations were at risk for failure once they
were founded.
0
10
20
30
40
50
60
1990 1995 2000 2005
Number of Name Changes
Distinctive Change
Change within Category
Mimicry Change
196
Figure 15: Mortality rate comparison between CPC mimics and non-mimics.
The log-rank, Wilcoxon, Tarone-Ware, and Peto-Peto-Prentice tests (Cleves, et
al., 2008) are common hypothesis tests for the equality of survivor functions. If such tests
achieve significance, it can be concluded that the survivor functions that are compared
differ significantly from each other (Blossfeld & Rohwer, 2002). As they were all highly
significant (p < .001, with χ
2
values ranging from 31.48 to 52.14), descriptive
nonparametric evidence suggested that the survival rate of mimics significantly exceeded
the one of non-mimics. This indicated that mimics were in fact fitter than non-mimics. In
a similar fashion, the relationship between organizational size and mortality was
explored. As mentioned previously, organization size could not be entered into any of the
statistical analyses due to the large number of missing values and the systematic
patterning of gaps in the availability of asset data. However, among those
0 .2 .4 .6 .8 1
Proportion Surviving
0 5 10 15 20 25 30 35 40 45 50
Analysis Time
CPC Non-Mimics
CPC Mimics
197
organizations for which size estimates could be made, noticeable size-based differences
between survivor functions were identified. As shown in Figure 16, the survivor function
plots revealed that large organizations had superior survival chances when compared to
small organizations, which attests to "the liability of smallness thesis" (Brüderl &
Schüssler, 1990, p. 534). Again, the results from log-rank, Wilcoxon, Tarone-Ware, and
Peto-Peto-Prentice tests confirmed this finding (p < .001, with χ
2
values ranging from
45.09 to 53.57).
Figure 16: Mortality rate comparison between large and small CPCs.
Table 7 provides summary descriptives and pairwise correlations between the
main effects of predictors hypothesized to affect CPC founding rates overall and those of
CPC mimics in particular. In Table 8, the same information is listed for all predictors
0 .2 .4 .6 .8 1
Proportion Surviving
0 5 10 15 20 25 30 35 40 45 50
Analysis Time
Small CPCs
Large CPCs
198
entered into the analyses of CPC label transformations. The descriptive statistics for
covariates used to predict CPC failures are presented in Table 9.
199
Table 7
Descriptive Statistics and Correlations for Founding Rate Analyses
Founding rates of CPC population (n = 3,061) Mean SD 1 2 3 4 5 6 7 8 9
1. Cognitive legitimacy
.02 .11
2. Sociopolitical legitimacy
.55 1.01 .03
3. Presidential abortion stance
.62 .49 -.27 .01
4. Human resource availability
41.96 21.57 0 .23 0
5. Client availability
8.64 1.52 0 -.13 0 -.29
6. NGO financial resources
175.95 71.47 .83 .04 .19 0 0
7. Metropolitan area county
.19 .4 0 -.12 0 -.26 .66 0
8. National CPC density
1,379.71 313.67 .94 .03 -.14 0 0 .91 0
9. State CPC density
36.48 26.45 .32 -.16 -.04 -.06 .20 .32 .17 .34
10. County CPC density
.44 .93 .10 -.14 -.02 -.25 .61 .10 .58 .11 .22
Founding rates of CPC mimics (n = 3,144) Mean SD 1 2 3 4 5 6 7 8 9 10
1. Cognitive legitimacy
.55 .11
2. Sociopolitical legitimacy
2.81 1.01 .03
3. Metropolitan area county
.19 .40 0 -.12
4. National CPC mimic entries
29.81 6.61 .24 0 0
5. State CPC mimic entries
.84 1.16 .07 -.14 .06 .16
6. County CPC mimic entries
.01 .10 .01 -.03 .10 .02 .12
7. National CPC mimic density
717.05 213.73 .92 .04 0 .37 .10 .01
8. State CPC mimic density
19.14 16.2 .34 -.15 .15 .14 .54 .04 .37
9. County CPC mimic density
.23 .60 .11 -.10 .47 .04 .08 .25 .11 .22
10. Weighted national RHP density
5,999.91 804.96 -.91 -.04 0 -.29 -.08 -.01 -.97 -.36 -.11
11. RHP proximity
.14 .35 -.04 -.23 .54 -.01 .04 .08 -.04 .07 .41 .05
200
Table 8
Descriptive Statistics and Correlations for Label Transformation Rate Analyses
Label transformations (n = 2,019) Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1. Cognitive legitimacy .57 .10
2. Sociopolitical legitimacy 2.54 1.10 .01
3. Presidential abortion stance .61 .49 -.15 .02
4. Human resource availability 30.8 17.97 .06 .35 0
5. Client availability 10.54 1.53 -.06 -.17 .01 -.37
6. NGO financial resources 188.87 70.85 .81 .02 .32 .06 -.05
7. Metropolitan area CPC .67 .47 -.06 -.09 .01 -.27 .70 -.06
8. Urban population in CPC's county 71.29 24.24 -.07 -.13 0 -.33 .79 -.07 .66
9. CPC cohort founded 1980-1989 .46 .50 -.23 .01 .01 -.06 .07 -.22 .08 .07
10. CPC cohort founded 1990-1999 .34 .47 .19 -.04 -.13 .09 -.14 .11 -.13 -.15 -.66
11. CPC cohort founded 2000-2009 .08 .28 .22 .03 .20 .08 -.09 .32 -.09 -.11 -.28 -.22
12. National CPC mimic density 759.28 199.68 .91 .01 .09 .06 -.06 .94 -.06 -.08 -.24 .17 .28
13. State CPC mimic density 25.92 18.53 .33 -.09 .04 -.11 .27 .34 .17 .15 -.08 .08 .09 .36
14. County CPC mimic density 1.21 1.64 .13 -.11 .02 -.20 .58 .14 .28 .37 -.04 .03 .04 .15 .41
15. Weighted national RHP density 5,845.41 748.26 -.89 -.01 .04 -.06 .06 -.89 .06 .08 .23 -.17 -.25 -.96 -.35 -.14
16. RHP proximity .52 .50 -.09 -.27 .01 -.37 .70 -.09 .55 .64 .09 -.14 -.09 -.10 .11 .35 .10
201
Table 9
Descriptive Statistics and Correlations for CPC Failure Rate Analyses
CPC failures (n = 2,019) Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1. Cognitive legitimacy .57 .10
2. Sociopolitical legitimacy 2.54 1.10 .01
3. Presidential abortion stance .61 .49 -.15 .02
4. Human resource availability 30.8 17.97 .06 .35 0
5. Client availability 10.54 1.53 -.06 -.17 .01 -.37
6. NGO financial resources 188.87 70.85 .81 .02 .32 .06 -.05
7. Metropolitan area CPC .67 .47 -.06 -.09 .01 -.27 .70 -.06
8. Urban population in CPC's
county 71.29 24.24 -.07 -.13 0 -.33 .79 -.07 .66
9. CPC cohort founded 1980-1989 .46 .50 -.23 .01 .01 -.06 .07 -.22 .08 .07
10. CPC cohort founded 1990-1999 .34 .47 .19 -.04 -.13 .09 -.14 .11 -.13 -.15 -.66
11. CPC cohort founded 2000-2009 .08 .28 .22 .03 .20 .08 -.09 .32 -.09 -.11 -.28 -.22
12. County CPC non-mimic density 1.55 1.56 .03 -.11 -.02 -.23 .64 .02 .35 .46 .03 -.08 -.08
13. Weighted national RHP
density 5,845.41 748.26 -.89 -.01 .04 -.06 .06 -.89 .06 .08 .23 -.17 -.25 .53
14. Weighted county RHP density 13.45 34.85 -.04 -.22 0 -.24 .59 -.04 .25 .37 -.05 .01 -.01 -.14 .04
15. CPC mimic .53 .50 .07 .01 .01 -.02 .01 .08 0 -.01 0 .10 .09 -.19 -.08 .04
16. CPC "born" mimic .60 .49 .11 .01 .04 0 .01 .13 .04 .02 -.27 .10 .25 -.25 -.12 .04 -
Note. Pairwise correlations with CPC "born" mimic were only calculated for CPC mimics; n = 1,115.
202
Hypothesis Tests
Founding Rates of CPCs
The first set of hypotheses tested focused on exploring the effects of density
dependence, legitimacy levels, and environmental resource availability on the founding
rates of all CPCs in the United States. As overdispersion precluded the use of Poisson
regression for modeling count data on CPC foundings, three negative binomial regression
models were estimated to examine these effects. The units at risk for experiencing CPC
founding events were U.S. counties and the analysis period included annual observations
between the years 1989 and 2009. Vuong tests were conducted and proved insignificant
for all models, which confirmed that zero-inflated negative binomial models were not
providing a superior fit to the data due to an excessive number of zero counts. The
model-fitting strategy was based on the additive selection of indicators, so groups of
variables were added in a stepwise fashion and the resulting nested models were
compared with likelihood ratio tests. Every subsequent model was found to fit the data
significantly better than the previous model. Table 10 provides an overview of coefficient
estimates with robust standard errors, fit comparison statistics, and values for the
overdispersion parameter α.
A dummy variable distinguishing between counties located in metropolitan areas
(metropolitan area county = 1) and those with locations outside of them (metropolitan
area county = 0) was added in order to control for the locational preferences of CPC
founders. In all three models, the coefficients for this variable were positive and highly
significant (b = 1.461, p < .001 in Model 1; b = 1.446, p < .001 in Model 2), which
suggested that CPCs are indeed founded at a higher rate in metropolitan settings.
203
Based on the estimated coefficient in Model 3, which fit the data best, the founding rate
of CPCs in metropolitan areas equaled e
(.5685)
= 1.77. Thus, the rate of metropolitan-area
CPC openings exceeded the rate of CPC emergence in non-metropolitan counties by
approximately 77%.
Table 10
Negative Binomial Regression Estimates of CPC Foundings, 1989-2009
Variables Model 1 Model 2 Model 3
CPC densities
National CPC density .0023 (.0009) ** .0044 (.0012) *** .00823 (.0017) ***
(National CPC density)
2
/ 100 -.0001 (.00003) *** -.0023 (.0005) *** -.0004 (.00008) ***
State CPC density .0043 (.0012) *** .0037 (.0011) ** .0026 (.0011) *
County CPC density .1420 (.0191) *** .1383 (.0192) *** -.0317 (.0274)
Legitimacy
Cognitive legitimacy 1.215 (.8616) 1.058 (.8484)
Sociopolitical legitimacy -.0853 (.0275) ** -.0464 (.0273)
Presidential abortion stance .2837 (.0930) ** .1263 (.1039)
Resources
Human resource availability .0011 (.0017)
Client availability .4691 (.0281) ***
NGO financial resources .0086 (.0026) **
Metropolitan area county 1.461 (.0667) *** 1.446 (.0671) *** .5685 (.0908) ***
Constant -5.372 (.5194) *** -7.018 (.8404) *** -13.47 (1.115) ***
Degrees of freedom 5 8 11
Log likelihood -5,591.73 -5,581.91 -5,456.93
Wald test χ
2
992.32 1,011.34 1,491.75
Δ df 3 3
Δ χ
2
9.82 * 124.98 ***
Number of foundings 1,227 1,227 1,227
Number of counties 3,061 3,061 3,061
Number of county-years 64,279 64,279 64,279
Overdispersion parameter α .529 (.245) ** .495 (.237) ** .320 (.207) *
Note. Model 3 was compared with Model 2; Model 2 was compared with Model 1.
* p < .05. ** p <. 01. *** p < .001 (two-tailed). Numbers in parentheses are robust standard errors.
Alpha significance levels: * p < .05 that α = 0, ** p < .01 that α = 0.
204
Hypothesis 1a stated that there would be a quadratic relationship between the
overall density of CPCs and the founding rate of CPCs. Such density-dependent effects
can be estimated by adding both main effect and squared density term to the estimation
and retaining the second-order effect if the coefficients for both are significant.
Accordingly, as the main effect of the density of CPCs on the national level as well as its
squared term appeared significant, they were entered into Model 1 together with the
metropolitan area control variable. The sign of the main effect was positive and
significant (b = .002, p < .01) as expected. The coefficient estimate for the squared term
was negative and highly significant (b = -.0001, p < .001).
Both findings together suggested that the hypothesized relationship between the
population density of CPCs and their founding rate indeed assumed the typical ∩-shaped
form. Given that this pattern remained stable and even increased in strength in
subsequent, better-fitting models, with p values for both estimates attaining significance
at the .001 level, the data provided strong support for Hypothesis 1a. Variables
representing CPC densities on the state (b = .004 in Model 1, b = .004 in Model 2) and
county levels (b = .142 in Model 1, b = .138 in Model 2) were also added into all three
models and appeared positive and strongly significant in the first two. However, as can be
seen in Model 3, the effects of county density disappear entirely (b = -.032, p = .25) and
only a small significant positive effect of state CPC density remains (b = .003, p < .05)
when resource effects are added. These results suggest that resource effects may be
mediating the effects of lower-level densities on the founding rates of CPCs. The
nonmonotonic relationship between national CPC density rates and CPC foundings
appeared unaffected by the addition of resource indicators.
205
Hypothesis 1b stated that there would be a positive relationship between
sociopolitical legitimacy levels and the founding rates of CPCs. Sociopolitical legitimacy
effects were estimated together with Hypothesis 1c, which predicted that CPCs would be
founded at a higher rate during the terms of pro-life presidents. The estimate for
sociopolitical legitimacy in Model 2 was significant (b = -.085, p < .01), but negative, so
the direction of the effect was not as hypothesized and Hypothesis 1b was not supported.
The coefficient for the dummy variable stratifying foundings according to the presidential
attitude towards abortion was found to be positive and significant (b = .284, p < .01) in
Model 2. Based on the parameter value, the founding rate of CPCs during pro-life
presidencies equaled e
(.2837)
= 1.33, which means that it surpassed the CPC emergence
rate during pro-choice presidencies by approximately 33%. However, in Model 3, which
was found to fit the data best, this effect was no longer significant. Thus, while there was
partial evidence in support of Hypothesis 1c, it was not confirmed conclusively. Again,
the disappearance of significant effects of sociopolitical legitimacy indicators in Model 3
hints at the possibility that resource availability mediates their influence on CPC
founding rates.
The remaining three hypotheses examined the influence of resource availability
on the founding rate of CPCs and were tested by adding resource indicators in Model 3.
Hypothesis 1d suggested a positive relationship between the influx of financial resources
into the NGO sector and the founding rates of CPCs. As the estimate of this effect was
positive and significant (b = .009, p < .01), Hypothesis 1d was supported. Hypothesis 1e
stated that there would be a positive relationship between the availability of human
resources and the founding rates of CPCs. Since the estimate failed to achieve
206
significance (b =.001, p = .52), Hypothesis 1e was not supported. Hypothesis 1f asserted
that there would be a positive relationship between client availability and the rate at
which CPCs are founded. The coefficient estimated for this predictor in Model 3 was
found to be positive and highly significant (b = .469, p < .001). Thus, Hypothesis 1f was
confirmed.
Founding Rates of CPC Mimics
The second group of hypotheses focused specifically on density and legitimacy
effects on the founding rates of CPC mimics, which were defined as organizations
entering the CPC population with organizational labels of types 3 and 4. These categories
correspond to weak and strong mimics, respectively. Therefore, the analysis was
restricted to a smaller subset of founding events. The model-fitting strategy for the
estimation of three negative binomial regression models resembled the approach
described in the previous section, but different independent variables were emphasized.
Again, the sequential addition of sets of predictors resulted in significantly better-fitting
models. Table 11 provides a summary of all models with coefficient estimates, robust
standard errors, and likelihood ratio test results, which constitute fit improvement
statistics. It also includes the values for the overdispersion parameter α, which was found
to be significantly different from zero for all three models.
Based on the significant results of these tests, it was concluded that fitting
negative binomial models to the data was more appropriate than Poisson models.
Additionally, Vuong tests were conducted for all models and did not yield significance.
Accordingly, it was inferred that zero-inflated model specifications would not provide a
significantly better fit for the data, which further confirmed the selection of negative
207
binomial regression modeling as appropriate for estimating the coefficients. Similar to the
examination of CPC foundings, the variable controlling for the metropolitan location of a
county was entered in Models 2 (b = 1.547) and 3 (b = 1.363) and its estimate was again
found to be significant (p < .001), which suggests that CPC mimics also tended to be
founded at a faster rate in metropolitan areas.
Table 11
Negative Binomial Regression Estimates of CPC Mimic Foundings, 1989-2009
Variables Model 1 Model 2 Model 3
CPC mimic densities
National CPC mimic foundings -.0029 (.0068) -.0016 (.0069) -.0033 (.0072)
State CPC mimic foundings -.0221 (.0378) -.0195 (.0369) -.0194 (.0367)
County CPC mimic foundings .1465 (.2393) .2426 (.2408) .2607 (.2427)
National CPC mimic density -.0006 (.0002) * -.0006 (.0005) .0001 (.0009)
State CPC mimic density .0144 (.0028) *** .0089 (.0028) ** .0094 (.0028) **
County CPC mimic density .4238 (.0448) *** .1482 (.0331) *** .1166 (.0353) **
Legitimacy
Cognitive legitimacy .6815 (.9763) .7706 (.9835)
Sociopolitical legitimacy -.1311 (.0385) ** -.0940 (.0392) *
Weighted national RHP density .0002 (.0002)
RHP proximity .4068 (.1082) ***
Metropolitan area county 1.547 (.0886) *** 1.363 (.1068) ***
Constant -4.629 (.2072) *** -5.065 (.3419) *** -6.767 (1.805) ***
Degrees of freedom 6 9 11
Log likelihood -3,433.71 -3,280.96 -3,273.38
Wald test χ2 195.37 555.10 590.62
Δ df 3 2
Δ χ2 152.74 *** 7.58 *
Number of foundings 626 626 626
Number of counties 3,144 3,144 3,144
Number of county-years 66,022 66,022 66,022
Overdispersion parameter α 3.779 (.926) ** 1.824 (.703) ** 1.794 (.690) **
Note. Model 3 was compared with Model 2; Model 2 was compared with Model 1.
* p < .05. ** p <. 01. *** p < .001 (two-tailed). Numbers in parentheses are robust standard errors.
Alpha significance levels: * p < .05 that α = 0, ** p < .01 that α = 0.
208
Hypothesis 2a stated that with increasing density levels of CPC mimics,
additional CPC mimics would enter the CPC population at a higher rate. In order to test
this hypothesis, density indicators representing the number of CPC mimics on the
national, state, and county levels were added in Model 1. Research suggests that
subpopulation founding events often spark additional entries at a later point in time,
which has been termed "frequency dependence" (Hannan & Carroll, 1992, p. 205).
Therefore, founding counts of mimics on the same three levels of analysis were included
in the model as controls. While none of the mimic foundings appeared to have effects on
the rate of subsequent foundings, the coefficients for the density of CPC mimics on the
state and county levels were found to be strong and positive in all three models. The
effect of the density of CPC mimics on the national level was negative and significant (b
= -.001, p < .05) in Model 1, but did not approach significance in subsequently fitted
models. This finding is in line with Lomi and Larsen's (1996) observation that
organizations with a mainly local activity radius may be powerfully affected by local
density rates but not by high-level densities at the national level. In Model 3, which fitted
the data best, the parameters for county (b = .117) and state densities (b = .009) of
mimics were found to exhibit significance at the .01 level, so Hypothesis 2a was
confirmed.
The next two hypotheses specified the relationship between legitimacy and the
founding rate of mimics. Hypothesis 2b asserted that there would be a negative
relationship between the cognitive legitimacy of the CPC form and the founding rate of
CPC mimics. As the direction of the parameters estimated in Models 2 (b = .681, p = .49)
and 3 (b = .771, p = .43) was positive and the estimates did not achieve statistical
209
significance, the results did not provide support for Hypothesis 2b. Hypothesis 2c stated
that there would be a positive relationship between the sociopolitical legitimacy of the
CPC form and the founding rate of CPC mimics. The coefficient estimate for this
variable was negative and significant (p < .05) when entered into Model 2 (b = -.131) and
Model 3 (b = -.093), but its direction did not concur with the prediction, so Hypothesis 2c
was not confirmed. The negative coefficient suggested that with sinking levels of
sociopolitical legitimacy, the founding rates of mimics increased rather than dropped.
The last two hypotheses about CPC mimic foundings concentrated on exploring
the role of the model population as driver of selection into a subpopulation of mimics.
Hypothesis 2d asserted that with increasing density of RHPs, CPC mimics would be
founded at a higher rate. While the estimate of the indicator for the weighted national
RHP density was positive when included in Model 3, it failed to achieve significance (b =
.0002, p = .39), so Hypothesis 2d was not confirmed. Hypothesis 2e stated that CPC
mimics would be founded at a higher rate in close proximity to an RHP. It was tested
with a dummy variable set to a value of 1 for those counties in which at least one RHP
was present and to a value of 0 for all others. As shown in the summary for Model 3, the
coefficient was positive and highly significant (b = .407, p < .001), which provided
strong evidence for Hypothesis 2e. Based on the RHP proximity estimate in this model,
CPC mimics were founded in counties with an RHP at a rate of e
(.4068)
= 1.50, which
means that the presence of an RHP was responsible for a mimic founding rate increase of
approximately 50% as compared to foundings in counties without an RHP.
210
Aggressive Mimicry Label Transformations
A third set of hypotheses examining label transformation events was introduced in
order to explore the conditions under which CPCs were pursuing mimicry strategies
through adaptation. Three piecewise constant exponential hazard rate models were
estimated to test these hypotheses. The estimation results, which include fit comparison
statistics and robust standard errors, are summarized in Table 12. Organizations were
considered at risk for a mimicry label transformation if their names were of types 1, 2, or
3, which correspond to strong and weak distincts or weak mimics, as only these
categories allowed for a switch to a name type more similar to RHP labeling conventions.
Groups of estimators were added in a stepwise fashion, so models were nested and
likelihood ratio tests could be conducted to determine improvements in overall model fit.
Model 1 was compared to a model containing only the seven period-specific effects (df =
7, n = 22,475, log likelihood = -1,059.28) and its fit was significantly (p < .001) better
than the baseline model. While the fit improvement of Model 2 over Model 1 did not
approach statistical significance, Model 3 exhibited a significant improvement (p < .001)
over Model 2 and constituted the best-fitting model. The estimates of most predictors
appeared significant across all models, which provides some evidence for the robustness
of the identified effects. The coefficients for the seven period controls were highly
significant for Models 1 and 2, which is not unusual for piecewise constant exponential
hazard models. They were no longer significant in Model 3 and were omitted in Table 12
for clarity of presentation.
211
Table 12
Piecewise Constant Exponential Hazard Rate Estimates of CPC Mimicry Label
Transformations, 1989-2009
Variables Model 1 Model 2 Model 3
CPC cohorts
Cohort founded 1980-1989 1.470 (.3055) *** 1.488 (.3061) *** 1.436 (.3082) ***
Cohort founded 1990-1999 1.421 (.3230) *** 1.676 (.3316) *** 1.714 (.3317) ***
Cohort founded 2000-2009 .6110 (.4791) 1.313 (.5471) * 1.644 (.5425) **
Legitimacy
Cognitive legitimacy .1846 (.9817) 3.025 (1.423) * 6.074 (1.606) ***
Sociopolitical legitimacy .0220 (.0509) .0461 (.0520) .0540 (.0537)
Presidential abortion stance -.3373 (.1399) * -.0019 (.1887) .1578 (.2127)
Resources
Human resource availability -.0026 (.0034) -.0023 (.0035)
Client availability .0968 (.0644) -.1489 (.0763)
NGO financial resources -.0082 (.0027) ** -.0057 (.0038)
CPC mimic densities
National CPC mimic density -.0068 (.0017) ***
State CPC mimic density -.0028 (.0040)
County CPC mimic density .2158 (.0344) ***
Weighted national RHP density -.0008 (.0036) *
RHP proximity -.0426 (.1720)
Control variables
Metropolitan area CPC -.0716 (.1512) -.1719 (.1713) .0537 (.1730)
Urban population in CPC's county .0019 (.0030) -.0022 (.0037) .0006 (.0037)
Number of transformations 335 335 335
Number of organizations 1,873 1,873 1,873
Number of organization-years 22,475 22,475 22,475
Degrees of freedom 15 18 23
Log likelihood -1,031.06 -1,025.36 -1,002.09
Wald test χ
2
5,816.48 5,795.09 5,818.66
Δ df 8 3 5
Δ χ
2
28.22 *** 5.70 23.27 ***
Note. Model 3 was compared with Model 2, Model 2 with Model 1, Model 1 with interval-effects-only baseline model.
Reference category for cohort effects: Cohort founded 1970-1979.
* p < .05. ** p <. 01. *** p < .001 (two-tailed). Numbers in parentheses are robust standard errors.
Variables controlling for a CPC's location in a metropolitan area or an urbanized
county were entered into all models, but proved insignificant. Three variables controlling
for the effects of the availability of resources were included in Models 2 and 3, but their
212
effects varied and disappeared entirely in the best-fitting model. Additionally, a dummy
variable examining the influence of the presidential abortion stance, which took on values
of 1 during the terms of pro-life presidents, was included as a control variable, but it was
only significant in Model 1 (b = -.337, p < .05) and the sign of the parameter changed in
Model 3 (b = .158, p = .458), which suggests that its effect overall was not only weak, but
also lacked consistency.
Hypothesis 3a stated that the first CPC founding cohort would engage in
aggressive mimicry at a lower rate than subsequent cohorts. It was tested by including
three dummy variables designating the 1980s, 1990s, and 2000s as founding decades,
with the founding period from 1970 to 1979 serving as a reference category. As shown in
Model 3, the coefficients for CPC cohorts founded during the 1980s, 1990s, and 2000s
were highly significant, so this hypothesis was supported. For cohorts from the 1980s and
1990s, effects were consistently strong (p < .001) across all models, suggesting that they
employed label mimicry during the observation period from 1989 to 2009 at a much
higher rate than organizations founded in the 1970s. Effects for the 2000s cohorts gained
in significance in better-fitting models (p < .01), thus lending further support to
Hypothesis 3a. Based on Model 3, which fitted the data best, CPCs founded during the
1980s transformed at a rate of e
(1.4360)
= 4.20, which means that their rate of change
exceeded the rate of the earliest founding cohort by approximately 320%. The estimated
transformation rate of CPCs founded during the 1990s equals e
(1.7141)
= 5.55, which
represents an increase of 455% compared to CPCs founded between 1970 and 1979. In a
similar fashion, CPCs founded during the 2000s transformed at a rate of e
(1.6440)
= 5.18,
213
which exceeds the earliest founding cohort's mimicry transition rate by approximately
418%.
Records for organizations founded prior to the year 1989, which marks the
beginning of the observation frame, are left-truncated, which affects both the 1970s and
1980s cohorts. Therefore, it is possible that these groups engaged in mimicry changes
before they underwent observation. However, given the high overall proportion of CPCs
with strongly distinct names among the oldest cohort, it appears that if they decided to
change their labels before 1989, it would not have been a mimicry transformation, but a
within-category change. In summary, the strength and consistency of the observed
effects, which extends to all cohorts throughout the observation period of 20 years, lend
strong support to the relationship articulated in Hypothesis 3a.
Hypothesis 3b stated that there would be a negative relationship between the
cognitive legitimacy of the CPC form and the rate of aggressive mimicry adoption among
CPCs. The estimates in Model 3 indicate that the effect of cognitive legitimacy on label
transformations was highly significant, but contrary to the expected direction. According
to the significance level (p < .001) and positive sign of the estimated coefficient (b =
6.074), CPCs engaged in mimicry transitions at an increasing rate with rising levels of
cognitive form legitimacy. Instead of being deterred by an increase in form
recognizability, which makes it more difficult to sustain a mimicry strategy, CPC
entrepreneurs appeared enticed to adopt labels that increased their organization's
similarity with RHPs. Thus, hypothesis 3b was not confirmed. Hypothesis 3c tested for a
positive relationship between the sociopolitical legitimacy of the CPC form and the rate
of aggressive mimicry adoption among CPCs. As none of the estimates for
214
sociopolitical legitimacy effects approached significance, this hypothesis was also not
supported by any of the models.
According to Hypothesis 3d, CPC mimics would engage in aggressive mimicry at
a higher rate with rising densities of mimics within the CPC population. For the purpose
of testing this hypothesis, three types of density variables were entered into Model 3.
While the national density of CPC mimics had a strong negative effect on the mimicry
transformation rate (b = -.007, p < .001), the coefficient estimating the effect of mimic
density on the county level had the predicted direction and was highly significant (b =
.216, p < .001). The indicator for state-level density of mimics did not emerge as a
significant predictor (b = -.003, p = .48) for CPC mimicry changes. In summary,
increasing overall numbers of mimics depressed the rate of further mimicry transitions
whereas an elevation of the local density of mimics was associated with higher transition
rates towards mimicry. Thus, Hypothesis 3d was fully supported in the context of
counties, but not confirmed on the national level; this mixed evidence leads to the
conclusion that Hypothesis 3d was only partially supported.
Hypothesis 3e suggested that with increasing RHP densities, CPC mimics would
engage in aggressive mimicry at a higher rate. However, as shown in Model 3, a higher
national density of RHP was associated with significantly (b = -.009, p < .05) lower
transformation rates towards mimicry, so Hypothesis 3e was not confirmed. It appeared
that an overall rise in the numbers of RHPs did not serve as an incentive for engaging in
aggressive mimicry, but rather a deterrent. Similarly, in Hypothesis 3f, it was
hypothesized that an organization's co-presence with a RHP in the same county would be
associated with increases in the transformation towards aggressive mimicry. The
215
coefficient was negative and insignificant (b = -.043, p = .80), thus failing to support
Hypothesis 3f.
Organizational Failures
The last group of hypotheses examined the influence of a variety of predictors on
the organizational demise of CPCs. Four Cox regression models were estimated to test
Hypotheses 4a through 4f. All incorporated CPCs in existence during 1989 to 2009 were
considered at risk for a failure event. The results of the analyses are presented in Table
13. The risk set for a fifth model was restricted to CPC mimics to allow for testing
Hypothesis 4g. As this hypothesis differentiated between organizations that entered the
mimicry population through selection and those that became mimics through adaptation,
non-mimics were excluded from the analysis in Model 5, which is presented in Table 14.
All models were estimated with three dummy variables controlling for the influence of
cohort effects on the failure rate of CPCs. The estimate for the variable referring to the
cohort founded between 2000 and 2009 was positive and significant in Model 1 (b =
.905, p < .001), suggesting that this group failed at a higher rate than the cohort founded
during the 1970s, but this effect disappeared in all other models with more covariates. In
Model 4, a slight negative cohort effect for organizations founded between 1990 and
1999 was detected (b = -.472, p < .05), which indicated that this cohort failed at a lower
rate than the earliest founding cohort.
216
Table 13
Cox Regression Estimates of CPC Failures, 1989-2009
Variables Model 1 Model 2 Model 3 Model 4
CPC mimic -.7100 (.1173) *** -.7104 (.1167) *** -.7128 (.1177) *** -.6198 (.1226) ***
CPC cohorts
Cohort founded 1980-1989 -.2233 (.1772) -.2489 (.1800) -.2975 (.1838) -.3250 (.1867)
Cohort founded 1990-1999 .2152 (.1742) -.2810 (.2158) -.3870 (.2314) -.4716 (.2321) *
Cohort founded 2000-2009 .9052 (.2299) *** -.1732 (.3545) -.3728 (.3936) -.5424 (.3859)
Legitimacy
Cognitive legitimacy
8.1359 (1.265) *** 6.7563 (1.4912) *** 6.2316 (1.580)
***
Sociopolitical legitimacy
-.0332 (.0505) -.0040 (.0549) .0146 (.0564)
Presidential abortion stance
-.1644 (.1580) -.3169 (.1771) -.2153 (.2232)
Resources
Human resource availability
-.0086 (.0038) * -.0087 (.0038)
*
Client availability
-.0869 (.0444) -.2757 (.0939)
**
NGO financial resources
.0029 (.0022) .0004 (.0037)
RHP densities
Weighted national RHP density -.0004 (.0004)
Weighted county RHP density .0011 (.0022)
County CPC non-mimic density
.1427 (.0516)
**
Control variables
Metropolitan area county
.0836 (.1826)
Urban population in CPC's county
.0038 (.0039)
Number of failures 316 316 314 306
Number of organizations 2,045 2,045 2,019 1,988
Number of organization-years 29,586 29,582 29,185 27,700
Degrees of freedom 4 7 10 15
Log likelihood -2,271.50 -2,249.17 -2,226.30 -2,157.36
Wald test χ
2
77.61 134.42 137.11 141.04
Note. * p < .05. ** p <. 01. *** p < .001 (two-tailed). Numbers in parentheses are robust standard errors.
217
Other controls included the weighted national and county densities of RHPs as
well as variables controlling for the metropolitan location of CPCs and their county's
percentage of urban population. These were entered in Models 4 and 5, but were not
found to be significant. Hypothesis 4a stated that CPC mimics would have a lower failure
rate than CPCs pursuing a distinctive identity strategy. CPC mimics include both
organizations "born" into label types 3 and 4 and those that changed from a lower
category into a higher one during their existence. A dummy variable distinguishing
between mimics (CPC mimic = 1) and non-mimics (CPC mimic = 0) was included in
Models 1 through 4 to test the first hypothesis. Its coefficients were consistently negative
and highly significant (p < .001) in all models, which provided strong support for
Hypothesis 4a. Based on the coefficient for the dummy variable in Model 4, the estimated
failure rate of CPC mimics equals e
(-.61980)
= .5381, which means that in comparison to
the failure rate of CPC non-mimics, these organization's failure rate was reduced by
approximately 46%.
The next two hypotheses focused on the role of legitimacy effects on the rate of
CPC dissolution. Hypothesis 4b claimed that there would be a positive relationship
between the cognitive legitimacy of the CPC form and the rate of failure among CPCs.
As the coefficient for cognitive legitimacy was positive and highly significant (p = .0001)
in Models 2 (b = 8.136), 3 (b = 6.757), and 4 (b = 6.232), this hypothesis was supported.
None of the models provided support for Hypothesis 4c, which suggested a negative
relationship between the sociopolitical legitimacy of the CPC form and the rate of failure
among CPCs. The estimated coefficient for this variable varied in its direction across
models and failed to achieve significance in any of them.
218
Two hypotheses investigated environmental resource effects on the mortality rates
of CPCs. Hypothesis 4d stated that there would be a negative relationship between the
availability of human resources and the rate of failure among CPCs. As can be seen in
Models 3 (b = -.009) and 4 (b = -.009), the estimates were negative and significant at α =
.05, which lends support to Hypothesis 4d. Hypothesis 4e examined if the availability of
clients would be negatively related to CPC failure rates. The variable was entered into
Models 3 and 4 together with the indicator for human resource availability. The p value
associated with its coefficient (b = -.087) in Model 3 equaled .05, thus approaching
significance, and it emerged as a significant predictor in Model 4 (b = -.276, p < .01).
These results provide evidence in support of Hypothesis 4e.
Hypothesis 4f stated that there would be a positive relationship between the
density of CPC non-mimics and the rate of failure among CPCs. The effects of the
density of CPCs pursuing a distinct identity strategy were estimated in Model 4 and the
associated coefficient appeared positive and significant (b = .143, p < .01). Thus,
Hypothesis 4f was supported. Table 14 provides the results for Model 5, which examined
the survival implications of mimicry via selection versus mimicry versus adaptation. As
mentioned previously, a smaller number of organizations were examined in order to
investigate Hypothesis 4g. It stated that "born" mimics would have a lower failure rate
than CPCs that entered the mimic subpopulation via label transformation. The effects of
the same covariates included in Model 4 were also estimated in Model 5, with similar
results.
219
Table 14
Cox Regression Estimates of CPC Mimic Failures, 1989-2009
Variables Model 5
CPC cohorts
Cohort founded 1980-1989 -.5309 (.4597)
Cohort founded 1990-1999 -.4043 (.4808)
Cohort founded 2000-2009 -.4473 (.6646)
Legitimacy
Cognitive legitimacy 9.2352 (2.759) **
Sociopolitical legitimacy -.0287 (.0978)
Presidential abortion stance .3156 (.3659)
Resources
Human resource availability -.0061 (.0061)
Client availability -.3234 (.1516) *
NGO financial resources -.0061 (.0058)
RHP densities
Weighted national RHP density -.0008 (.0005)
Weighted county RHP density .0007 (.0028)
CPC "born" mimic 1.0078 (.2581) ***
County CPC non-mimic density .1997 (.0806) *
Control variables
Metropolitan area county .1313 (.2878)
Urban population in CPC's county .0073 (.0069)
Number of failures 111
Number of organizations 1,084
Number of organization-years 14,137
Degrees of freedom 15
Log pseudo-likelihood -698.21
Wald test χ
2
72.24
Note. The risk set only includes CPC mimics.
* p < .05. ** p <. 01. *** p < .001 (two-tailed).
Numbers in parentheses are robust standard errors.
The coefficient estimated in Model 5 for the dummy variable differentiating
between adaptive and selective mimics was positive and highly significant (b = 1.008,
p < .001). Based on this coefficient, the estimated failure rate of "born" mimics equaled
e
(1.0078)
= 2.74, which exceeds the failure rate of organizations engaging in mimicry
220
through transformation by approximately 174%. While significant differences in the
survival rates of the two groups of mimics were discovered, the predicted direction of the
effect was not supported by the estimation results. Thus, Hypothesis 4g was not
confirmed. Table 15 provides a summary of all independent variables examined together
with their hypothesized effects and the corresponding results based on the statistical
analyses conducted.
221
Table 15
Summary of Hypothesized Effects and Corresponding Results From Hypothesis Tests
Founding rates of CPCs Hypothesized effects Result
Hypothesis 1a:
National CPC density, (national CPC density)
2
/ 100
(+), (-) Supported
Hypothesis 1b: Sociopolitical legitimacy (+) Not supported
Hypothesis 1c: Presidential abortion stance (+) Partially supported
Hypothesis 1d: NGO financial resources (+) Supported
Hypothesis 1e: Human resource availability (+) Not supported
Hypothesis 1f: Client availability (+) Supported
Founding rates of mimics within the CPC population
Hypothesis 2a:
State CPC mimic density, county CPC mimic density
(+), (+) Supported
Hypothesis 2b: Cognitive legitimacy (-) Not supported
Hypothesis 2c: Sociopolitical legitimacy (+) Not supported
Hypothesis 2d: Weighted national RHP density (+) Not supported
Hypothesis 2e: RHP proximity (+) Supported
Aggressive mimicry label transformations
Hypothesis 3a: Cohort founded 1980-1989,
cohort founded 1990-1999, cohort founded 2000-2009
(+), (+), (+) Supported
Hypothesis 3b: Cognitive legitimacy (-) Not supported
Hypothesis 3c: Sociopolitical legitimacy (+) Not supported
Hypothesis 3d:
National CPC mimic density, county CPC mimic density
(+), (+) Partially supported
Hypothesis 3e: Weighted national RHP density (+) Not supported
Hypothesis 3f: RHP proximity (+) Not supported
Organizational failures among CPCs
Hypothesis 4a: CPC mimic (-) Supported
Hypothesis 4b: Cognitive legitimacy (+) Supported
Hypothesis 4c: Sociopolitical legitimacy (-) Not supported
Hypothesis 4d: Human resource availability (-) Supported
Hypothesis 4e: Client availability (-) Supported
Hypothesis 4f: County CPC non-mimic density (+) Supported
Hypothesis 4g: CPC "born" mimic (-) Not supported
222
CHAPTER 8: CONCLUSION
Discussion and Implications
Even in the life sciences, the notion of mimicry is contested (Wickler, 1968) due
to the difficulty of establishing boundary conditions for the range of phenomena the
concept comprises. Prior to this study, ecologists have not empirically studied the
conditions for and effects of aggressive mimicry in organizational communities. In order
to determine the usefulness of this concept for the study of organizations, it is necessary
to take a step back and critically examine its potential for stimulating institutional
ecology theorizing. Therefore, the results of the analyses have to be revisited and
scrutinized in terms of their implications for existing research on organizations.
As described previously, the notion of aggressive mimicry challenges three
assumptions commonly built into ecological models of organizational communities. First,
it questions the expectation that low levels of cognitive legitimacy must be associated
with higher mortality rates. As mimics are believed to thrive on form ambiguity, it is
beneficial for them if their form remains unrecognized by audiences. In turn, a high
degree of cognitive legitimacy should have adverse effects on the survival rate of mimics.
The study results provide support for this qualification of ecological models as they
suggested that the relationship between cognitive legitimacy and CPC mortality was
indeed positive and highly significant. Accordingly, organizational scholars may have to
consider that for those organizations pursuing mimicry strategy, traditional assumptions
about the beneficial effects of cognitive legitimacy do not hold.
It has been a long-standing tradition among communication scholars to
acknowledge the strategic value of communicative ambiguity. The theoretical
223
appreciation of its significance marks a transition from systems approaches to those
explicitly including cultural aspects of communication (Eisenberg, 1984, 2007). The
current tendency among institutional ecologists to emphasize the beneficial effects of
form recognizability corresponds closely with early discussions about the need for
reducing equivocality in order to improve the effectiveness of organizational systems.
The concept of aggressive mimicry foregrounds strategic ambiguity, and since it
originated in the life sciences, it is highly compatible with ecological modeling. Thus, it
may prove fruitful for institutional ecologists to investigate the modeling implications of
the occurrence of aggressive mimicry within organizational communities in greater
detail.
A second expectation derived from organizational ecology and challenged by the
notion of aggressive mimicry is that competitive pressures due to rising density levels
will result in attempts by populations to carve out distinct niches in order to reduce these
pressures (Carroll, et al., 2002; Dobrev, Kim, & Hannan, 2001; Soule & King, 2008).
Instead, higher densities of mimics should be associated with additional foundings of
mimics regardless of increasing competitive pressures. The results of the current study
support this revised expectation. While the overall founding rates of CPCs followed the
typical density-dependent nonmonotonic pattern, foundings of CPC mimics were found
to rise strongly with the number of existing mimics in the same state as well as in the
same county. Likewise, analysis results confirmed that the rate of entry into the CPC
mimic population via label transformations was positively related to the density of
mimics on the county level. None of the second-order effects were detected that give the
∩-shaped curve its downward slope (Carroll & Hannan, 1989b). These findings
224
indicate that the availability of local intrapopulation role models pursuing a mimicry
strategy constitutes an important incentive for other organizations to adopt the same
strategy regardless of the density pressures exerted by their growing numbers.
A third assumption common among organizational ecologists is that organizations
switching form identities are penalized with exacerbated mortality risks. These two-fold
risks stem from the general dangers inherent in organizational transformations on the one
hand (Hannan, et al., 2006) and from the adverse effects of signaling an ambiguous
identity on the other hand (Hsu, 2006a, 2006b; Hsu, et al., 2009; McKendrick, et al.,
2003). According to theorizing about aggressive mimicry, successful mimics should
derive survival benefits from mimicking the form of a different population instead of
incurring higher mortality rates. The empirical evidence collected for this study supports
this revised assumption as it suggests that adaptive mimics strongly improve their
survival chances by selecting labels more similar to those of RHPs. As mimics seemingly
do not depend on developing stable and distinct identities towards various constituencies,
they appear to thrive on label transformations rather than risking demise when engaging
in them.
However, while the study results confirmed the superior survival chances
associated with adaptive mimicry, they were also puzzling as they suggested that entries
into the mimicry subpopulation through adaptation entailed larger survival-enhancing
benefits than being "born" into mimicry via selection. Both types of entry are
advantageous as CPC mimics were found to be significantly fitter than subpopulations
engaging in distinct identity strategies, but the expectation that Darwinian entries would
be associated with lower mortality rates than Lamarckian entries was not confirmed.
225
Given that Hannan et al. (2006) were not able to discover any circumstances under which
identity-related change proved beneficial to organizations and concluded that such
transformations present great destabilization risks for organizations, it appears surprising
just how valuable the choice of mimicry as an adaptive strategy proved for CPCs.
In the following, empirical results regarding the influence of legitimacy variables,
environmental resource factors, CPC and RHP density effects, and the role of cohort and
size will be discussed in further detail. The discussion of these effects departs from the
presentational structure of the statistical models in order to highlight the variegated
impact of these groups of predictors across outcomes. Tracing the differential impact of a
predictor on CPC vital rates allows for comparing its effects across different situations.
As will be seen, their impact on foundings, transformations, and failures does not always
fully comply with the theoretical expectations on which the hypotheses were based.
In general, cognitive legitimacy was not discovered to play a noteworthy role in
the founding of mimics, even though it had been assumed that entries into the mimicry
subpopulation would be accelerated under conditions of low cognitive legitimacy.
Contrary to similar expectations regarding adaptive mimicry via label transformations,
cognitive legitimacy was found to increase transformation rates, not depress them. These
results suggest that CPC entrepreneurs may become more susceptible to experimenting
with mimicry if increasing amounts of form information are starting to circulate. Instead
of staying away from mimicry, which works best when cognitive legitimacy levels are
low, CPCs adopt mimicry labels at an increasing rate when cognitive legitimacy is high.
Several explanations may apply to make sense of this finding. Either CPC
entrepreneurs do not actually ponder heightened risks associated with aggressive
226
mimicry when form recognition is high or they are unable to accurately evaluate its
potential adversarial effects (Strang & Macy, 2001). In any event, the diffusion of
information about the mimicry form spurs mimicry adaptations. In a similar vein, the
results of Conell and Cohn's (1995) research on coal miners' strikes suggest that news
about successful strikes provides incentives for additional protest. Aldrich and Fiol
(1994) point out that high levels of cognitive legitimacy are conducive to industry entry
as influential third-party actors may spread knowledge about the new venture, which is
another explanation for the reversed relationship discovered. In the case of CPCs, which
can rely on well-connected national umbrella organizations for expertise on how to
operate as mimics, this might explain their preferences for switching into the CPC mimic
subpopulation. As Strang and Macy (2001) have shown, entrepreneurs emulate business
models that are touted as successful by the media. Incidentally, this means that dupes
may learn about the practices of CPC mimics through the same coverage that informs
CPC entrepreneurs about how to successfully implement a mimicry strategy.
It is even more challenging to interpret the size and direction of the effects of
sociopolitical legitimacy on the CPC population and the mimics among it. Overall, the
influence of a favorable policy environment on CPC foundings, transformations, and
survival was not found to be positive as originally expected, and in most cases, there
were no discernable effects at all. Some evidence suggested that especially CPC mimics,
but also some non-mimics are preferentially founded in states providing a less favorable
sociopolitical environment. Again, it appears that CPC entrepreneurs do not shy away
from entering the mimic subpopulation just because the potential sociopolitical sanctions
associated with a mimicry strategy may be higher. This indicates that even though
227
penalties for mimics may be higher under conditions of low sociopolitical form
legitimacy, mimicry strategists among CPCs are quite likely to exploit viability
combinations that involve both low sociopolitical legitimacy and, occasionally, high
cognitive legitimacy. A similar and related finding is that high sociopolitical legitimacy
levels have not been found to make mimicry labeling strategies more attractive to
existing CPCs.
Several potential explanations may be applied to make sense of these results. For
one, the penalties associated with selecting a mimicry strategy could be so small that they
are of little concern to CPC entrepreneurs, even if they are operating in comparatively
"unfriendly" policy environments. One indication that this may be the case is the
enduring preference of a many CPCs to advertise under "abortion services" in the Yellow
Pages, even though this section is reserved for RHPs and their actions can be sanctioned
as fraudulent advertising in a number of states. Additionally, it is plausible that in states
with a consistently favorable policy environment, mimicry has simply not appeared to be
as attractive an evolutionary strategy as in other states. As mentioned previously, label
choices such as "Women's Choice Pregnancy Center" were lamented as disastrously
ineffective center titles for rural CPCs in politically conservative counties without a
proximately located RHP (Monahan, 2007). Seemingly, CPCs do not incur positive
returns for mimicking RHPs under such conditions.
In the context of this study, the role of three types of environmental resources for
the evolution of the CPC population was examined. The overall influx of funding into the
NGO sector appeared important for foundings, but it had no impact on CPC failures. This
finding suggests that the resource mobilization patterns of CPCs follow similar
228
dynamics as those common in other social movement sectors, but once they are founded,
their dependence upon general resource availability declines. However, as the availability
of financial resources was only measured for the entire nonprofit sector rather than the
population of CPCs specifically, these findings have to be interpreted accordingly and
with caution. They constitute a preliminary examination of the relationship between
resources and CPC vital rates and additional research is necessary that operationalizes
resource availability by drawing on data about the actual fundraising habits of CPCs.
The impact of the local availability of human resources in terms of the size of a
county's religious volunteer base was found to be just the opposite. It turned out to be
significant for averting CPC failures, but it did not constitute an influential factor in the
CPC founding process. As many CPCs start out as projects of dedicated individuals that
are acquiring volunteer staff over time, it is possible that start-ups seem less reliant upon
the accessibility of human resources prior to founding. Finally, the availability of clients
proved a significant resource for CPCs both in terms of founding and disbanding
dynamics. So, not only do CPCs exhibit some dependence upon demand-side
evolutionary pressures that affect their survival rates, but they also respond to resource
opportunities with increased founding activities.
As the study results suggest, the tight relationship between the dynamics of RHPs,
the model population, and CPCs, the mimic population, was rarely shown to be as strong
as previously expected. A rise in the overall RHP density levels appeared to have no
impact on mimic foundings and it was not confirmed as an influential factor for CPC
transformations towards mimicry. On the contrary, a density increase within the total
population of RHPs seemed to somewhat depress the rate of mimicry change among
229
CPCs. Additionally, RHP densities on the national and local levels, which were entered
as controls into the failure models, did not emerge as important determinants of CPC
mortality, which could have been expected given their competitive coevolutionary
relationship. The local proximity of RHPs proved to be an influential factor for mimic
foundings, but not for transformations towards mimicry. This provides some evidence
that CPCs monitor their models, but raises the question of why the presence of models
constitutes a driver for selective entries into the mimicry subpopulation, but not for
adaptive entries.
Given the strong and consistent finding that CPCs will readily copy
intrapopulation role models engaging in mimicry strategies, it appears that mimetic
isomorphic tendencies among CPCs (Barreto & Baden-Fuller, 2006) are a much stronger
motor for mimicry evolution than their direct coevolutionary linkage with RHPs. As
discussed previously, a large proportion of CPCs is associated with national umbrella
organizations, and higher-level strategic visions pursued by associations such as NIFLA
and Heartbeat may be ultimately responsible for the evolutionary trajectories detected
among individual population members (Lee & Pennings, 2002). The ultrasound initiative
promoted by NIFLA since the early 1990s (Glessner, 2001), for example, constitutes a
population-wide incentive to engage in a strong version of mimicry as it enables CPCs to
acquire clinic capabilities. CPCs that adopt the clinic form in response to incentives
provided by national umbrella organizations are engaging in transformations that
exemplify "coercive isomorphism" pressures as described by DiMaggio and Powell
(1983, p. 150). While the outcomes on the interpopulation level are the same in terms of a
lateral niche migration of CPCs towards RHPs, the empirical findings imply that
230
mimicry dynamics may be fueled more forcefully by diffuse intrapopulation mimetic
isomorphism rather than the dynamics of direct interpopulation competition with
individual RHPs.
Cohort and size effects did not occupy prominent roles in terms of hypothesis
testing, but they deserve mentioning as such organization-level characteristics aid in
gathering more knowledge about CPCs as a relatively understudied population. While no
differences were detected between CPC founding cohorts in terms of their failure rates,
expectations informed by theorizing about organizational inertia (Hannan & Freeman,
1984) were confirmed when cohort effects were entered into the transformation rate
models. CPCs founded in the 1970s were found to engage in label changes towards
mimicry at a significantly lower rate than all other founding cohorts. As label
transformations represent identity-related change that is suppressed by cognitive inertia
(G. P. Hodgkinson, 1997), this is a finding supported by current theories about
organizational identities and form change. Likewise, the positive relationship between the
organization size of CPCs and their survival rate, which was demonstrated in univariate
descriptive analyses based on life table estimates, concurs with expectations derived from
existing organizational ecology.
In summary, the results of the study verified some aspects already known about
organizational population dynamics in the context of a new population of NGOs. They
hinted at the appropriateness of revisiting extant theories and adjusting them in order to
incorporate novel concepts such as aggressive mimicry as several of the revised
assumptions associated with this line of theorizing were confirmed. Additionally, some of
the findings were surprising and challenged expectations, thus offering the potential
231
to further stimulate theory development among organizational ecologists seeking to find
apt interpretations for them. Hopefully, the study will also generate advances in terms of
modeling the evolution of cognitive legitimacy. As the method of tracking the meaning
stabilization processes of form communication by employing meta network analysis has
not been applied previously, it invites scrutiny and may encourage further developments
in this area.
Limitations
While some of the limitations of the study have already been pointed out
previously, they have to be addressed again explicitly so that their impact on the
interpretation of the study results can be better assessed. First, the paradigmatic
underpinnings of the study, which strongly rely on principles from organizational ecology
(Hannan & Freeman, 1989), do not constitute a limitation per se, but they certainly lay
out its trajectory in terms of theorizing, method selection, data collection techniques, and
the interpretation of analyses (Porac, 1994). Second, ecological principles also inform the
evaluation of limitations as they pertain to the data sources used, which are numerous and
deserve careful scrutiny. Potential weaknesses of the measurements employed are a third
aspect that delineates boundaries for the usefulness of the study. Fourth, the study tests
aggressive mimicry only indirectly, and its heavy reliance on such inferences constitutes
a possible vulnerability that has to be examined in greater detail.
As is common in macrolevel modeling of population dynamics based on
ecological theorizing, challenging questions such as when exactly organizations appear,
change, and expire must be addressed by measures that are not always as nuanced as
desirable. McKendrick and Carroll (2001) note that transient and emergent
232
populations tend to be understudied and that the emphasis is typically placed on
established organizations. This limitation also applies to the current study given that it
relies on the tax information available about incorporated CPCs. As described before, the
flurry of CPC listings online suggests that the true population size and its labeling
volatility and locational migration patterns are much more expansive than indicated by
the data set. CPCs were frequently found to have multiple locations and many different
names and the true percentage of CPCs with tax-exempt status is unknown.
The current study focuses exclusively on the examination of vital rates among
members of the CPC population, which also attests to its theoretical rootedness in
ecological modeling. However, the question arises if measures such as organizational
survival are able to capture the activities of NGOs and other types of organizations
appropriately (M. W. Meyer & Zucker, 1989). Particularly, concerns about the suitability
of the dependent variables examined arise given that aggressive mimics may be mainly
found among organizational populations that can hardly be thought of as thriving as their
success is temporary and fluctuates with rising levels of cognitive legitimacy. Fraudulent
business schemes wax and wane and impostors come and go in waves depending upon
possible sanctions keeping their activities in check. M. W. Meyer and Zucker (1989)
formulate a variety of critical questions about the measurement of organizational
performance in terms of mere persistence and their critique certainly applies to this study
as well as it employed many techniques common among ecological analyses.
An assessment of the quality of the data used for this study in light of the criteria
applied by organizational ecology standards points to a number of weaknesses that may
raise criticism even among those that firmly subscribe to the merits of organizational
233
ecology. For one, the data available for CPCs does not span the whole history of the
population of interest, which is desirable for the construction of a full ecological model
(Hannan & Carroll, 1992; Hannan & Freeman, 1989). Associated limitations such as left-
truncation (Allison, 1984) and the reliance on a restricted-inclusion database have been
outlined by Carroll and Hannan (2000). While methodologists recommend to consider
birth and death processes simultaneously in ecological modeling (Carroll & Hannan,
1989b), no data sources are available that span the entire history of CPCs. Therefore, the
limitations of the data restrict the interpretability of the study results.
Many of the independent variables included in the statistical analyses are concepts
measured only with a single indicator, which has been criticized as insufficient for tracing
complex variable interrelationships (Young, 1988). The variable capturing the
sociopolitical environment of CPCs illustrates that this criticism is justified. Existing
research has shown that sociopolitical legitimacy should be measured on multiple levels
and is a multidimensional construct that manifests itself in a host of possible indicators
(Ruef & Scott, 1998; E. T. Walker & McCarthy, 2010) . While in the context of CPCs,
the state level seemed appropriate as state laws may affect CPCs most strongly, there are
more refined ways of measuring sociopolitical legitimacy that rely less heavily on its
policy dimensions (W. D. Berry, et al., 1998; D. S. Meyer & Staggenborg, 1996).
A second concern about the sociopolitical legitimacy measure emerged from the
extraordinary transformation of abortion policies in the United States over time. The
difficulty of measuring a policy environment in flux was reflected by the tremendous
alterations of the NARAL policy index over time, which was updated regularly to capture
the rising number of laws and regulations governing abortion access. Additionally,
234
the variable measuring the amount of financial resources available to CPCs consisted of a
single indicator which included the total amount of annual charitable giving within the
U.S. As mentioned previously, this is a very global measure that is unable to capture the
specific fluctuations associated with financial resources flowing into the pro-life
movement in general and the CPC sector in particular.
The interpretability of the effects of cognitive legitimacy is severely restricted as
its operationalization for the purpose of this study suffers from a variety of flaws. The
limitations of this particular measure will be briefly outlined in the next paragraphs. As
the descriptions of the theory about form emergence as well as the methodology used to
track form communication suggest, cognitive legitimacy is a multi-faceted construct that
is shaped by the interactions of multiple audiences over time (Hsu, 2006a). Instead, the
measure was restricted to media coverage as a single source for textual data and the
resulting indicator distilled it into a single-variable measure. Ideally, audiences should be
questioned directly about how they are conceptualizing an organizational form (Glynn &
Marquis, 2006). As mentioned by Glynn and Marquis (2006) in reference to form
communication, "the recognition of those signals will vary depending on individuals'
receptivity to those cultural accounts" (p. 119), which essentially means that people differ
when it comes to their conclusions drawn about a form even if they were exposed to the
same media content.
News coverage over time constitutes a collection of stories as complex
communicative artifacts which document the evolution of an argument over time (Green,
et al., 2009). Computer-aided analyses of a large quantity of textual data cannot examine
the substantive individual contributions of the stories. They necessarily omit a large
235
semantic portion of any individual story even though it might have been potentially
relevant for the discursive development of a form. While the measure goes beyond the
common quantification of cognitive legitimacy as tantamount to the number of articles
per year, it employs methods that are problematic in their own way.
First, as was described in chapter 6, the indicator for cognitive legitimacy was
based on the combination of two network density coefficients that were measured every
year. Density measures vary based on the number of nodes and the number of links.
Textual networks such as the ones used for the purpose of this analysis have special
properties due to the way the number of nodes and the number of ties grow over time. As
the number of nodes in the network was based on the occurrence of concepts drawn from
a finite thesaurus, it grew comparatively little over time. However, the number of
linkages between these nodes was based on the number of concept co-occurrences and
grew very strongly. This growth was due to a considerable increase in the number of
yearly articles, which resulted in a strong increase of the number of co-occurrences. Thus,
the analyzed textual networks were found to exhibit densification over time (Leskovec,
Kleinberg, & Faloutsos, 2007). Ideally, it might improve the measure if the densification
of the networks over time could be controlled for. This would mean that the observed
density changes between the years could be interpreted without being affected by the
exponential growth of the links in comparison to the nodes. In summary, densities are a
useful network measure, but they can be also influenced unduly by increases in the article
counts, so they should be interpreted with caution.
Second, the clustering method used to determine the density measures was based
on the researcher's subjective categorization of recurring concepts. Different
236
audiences would likely arrive at different conclusions if a given concept contributes to
the discursive distinction of CPCs as an organizational form or if it rather signals their
resemblance with RHPs. As there is no objective basis for conducting such a
classification, the resulting differentiation between mimicry and ambiguous clusters must
be considered exploratory. It might be possible in future research to formalize the process
of cluster creation in order to establish the reliability of the cognitive legitimacy measures
derived from meta network analysis. However, the underlying idea informing the method
presented was that a measure of cognitive legitimacy has to fit its specific purpose (in this
case, CPC form recognizability in relationship to the RHP form. Accordingly, the
measure developed always requires customization with the help of human coders and
cannot be generalized without the risk of further misspecification.
Another limitation of the study is that it depends on archival data to draw
inferences about the conditions and outcomes of mimicry. It relies on the appropriateness
of anecdotal evidence about the intentionality of labeling transformations as it is based on
the claim that the name changes of CPCs are in fact strategic adaptations. Additionally,
there is only secondary evidence that the newly chosen names are indeed successful in
sending ambiguous signals to potential clients of RHPs. CPC staff was not questioned
about internal identity claims or about explicit strategies for engaging in some name
changes, but rarely others. In order to determine the degree of ambiguity inherent in an
organization name, groups of clients would have to be surveyed, which was also not a
focus of this study.
Drawing on detailed data about the adoption intentions of individual
entrepreneurs in the health industry, Kennedy and Fiss (2009) were able to show that
237
their perceptions in fact influenced their decisions in favor of organizational adaptation.
However, this very fact makes their study truly exceptional as intentions about
managerial decision-making examined on the population level are usually inferred. In the
context of the current study, it would have been desirable, but not feasible to confirm the
motivations of organization members within the CPC population. As staff members of
individual CPCs as well as of national umbrella organizations are unlikely to participate
in any study that would trace their intentions to mimic RHPs, surveying them about their
strategies would have probably yielded results reflecting social desirability rather than
actual behaviors.
While it is consequential that many of the assumptions are derived from
inferential evidence, it can be argued that for the purpose of an ecological study, the
outcomes observed in terms of the name transitions do give some weight to the hunch
that they are goal-oriented and intentional. The examination of the relationship between
the observed CPC population vital rates and a variety of covariates informs about some of
the conditions for label transformations as well as the effects of such name changes. As
evolutionary processes do not depend on variations being purposive or blind as long as
there are systematic differences in entry, exit, and direction of name transition, the
reliance upon inferences can still yield results that remain valid in the context of
ecological modeling. While it would have been advantageous to study the motivations of
CPC entrepreneurs to pursue mimicry strategies and to confirm the ambiguous effects of
CPC labels on their clients, this study may nevertheless provide insights into the overall
population dynamics within which mimicry unfolds.
238
Suggestions for Future Research
Especially the unexpected findings and study limitations discussed previously
represent opportunities for further inquiry, which is why they also inform the suggestions
for future research described in the following. Apart from many other possible avenues,
the study of mimicry opens up space for further investigations of legitimation processes,
model-mimic interaction dynamics, and examinations of the ethical dimensions
associated with mimicry strategies.
First, much work remains to be done in terms of capturing the evolution of
cognitive legitimacy over time. Recognizability is a central aspect studied in the context
of aggressive mimicry, and mimics themselves may play a very active role in
communicating differential identities to various audiences. As mentioned by Hannan
(2005), "organizations interact in important ways with multiple audiences, and thus to
some extent might be defined by several possibly interrelated identities" (p. 66). This
combination of the strategic form communication of individual mimics within the context
of higher-level audience negotiations about the meaning of a form has not been fully
captured theoretically and only facets of it have been studied empirically. CPCs engage in
diversified communication strategies, for example, when they are hosting two separate
organizational websites. One is intended for communication with pro-life donors whereas
the other is designed to attract clients by mimicking the appearance of a health care
professional's Internet presence.
As described by Jacobson (2005), "such a site would communicate very different
messages than would be given to volunteers and supporters" (p. 27). The donor website
of a CPC called "Pregnancy Center of Pinellas County" (online at
239
http://www.newlifesolutions.org), for example, identifies the Christian orientation of the
organization as a pregnancy ministry whereas the client website (online at
http://www.pregctr.net) appears entirely secular and is unconnected to the donor website.
Similarly, it is possible to get from the donor website of a CPC called "Blue Ridge
Women's Center" (online at http://www.supportblueridge.org) to the client website
(online at hptt://www.blueridgewoman.org), but not vice versa. The website for clients of
the CPC is also strictly secular in appearance and reminiscent of a professional health
care provider's website. In contrast, donors are informed about the activities of the
pregnancy ministry on the donor website of the CPC, which contains ample references to
the religious nature of the enterprise. By simultaneously running two disconnected
websites, CPCs may seek to garner support from Christian congregations while
simultaneously presenting themselves as secular organizations (Glessner & O'Connor,
2007) to potential clients perusing their websites. These routine switches (Mische &
White, 1998) between foregrounding differential organizational identities unfold in a
communicative space bounded by the overall evolution of form legitimation that is
dominated by larger social movement dynamics.
Ideally, VSR principles could be applied to the evolution of the textual networks
that constitute the fabric out of which organizational forms are spun. The application of
methods infused by evolutionary theorizing would enable researchers to empirically
investigate how the social structures of organizational communities coevolve with the
cultural forms describing them (Luhmann, 2008). Monge and Poole (2008) describe the
benefits of such evolutionary approaches as follows:
The ecological perspective offers a theoretical basis for understanding how
discourses develop both within and across populations of organizations.
240
Ecological theory would represent these multiple conversations and the discourses
they constitute as a community ecology of texts and conversations engaged in
competitive or cooperative relationships for scarce resources and driven by
variation, selection, and retention processes. (p. 687)
Textual analysis in combination with analysis techniques based on ecological
principles can help to further scrutinize the legitimation dynamics associated with form
negotiations. Communication scholarship provides advanced textual analysis methods
with which theorizing derived from new institutionalism (Powell & DiMaggio, 1991) can
be better empirically verified. At the same time, these methods are rigorous enough to
create measures of form communication that are suitable for the kinds of quantitative
analyses typical for organizational ecology. Communication scholars may continue in the
future to provide important contributions to the investigation of the effects of form
negotiation struggles.
In addition to differential external communication strategies, investigations could
also focus on internal legitimation processes as outlined by Aldrich and Ruef (2006). At
the heart of mimicry lies the ability of organizations to engage in deception in the sense
that their members are consciously aware of differences between what they construct as
their internal identity versus their identity constructed for external audiences. Based on
what is known about them, organization members within CPCs have very strong and
committed ties to their organizations and a very clear sense of their purpose (Conway,
2000). If mimicry strategies are pursued for ideological reasons, it does not apply that
"the organization name [is] a succinct marker of the organization's identity to both
internal and external audiences" (Glynn & Marquis, 2007, p. 19). Sahlin and Wedlin
(2008) argue that "imitation shapes identity" (p. 223), and as aggressive mimics do not
241
select their models based on identifying with them as positively valenced organizational
exemplars, research is needed that examines the effects of mimicry on the evolution of
their internal identities. The exact dynamics of how members of organizations engaging
in mimicry are dealing with the incongruences between their internal and external
identities remain largely unexplored even though they constitute a vital aspect of
theorizing about organizational deception dynamics.
As mentioned previously, mimicry strategies may involve feature-based as well as
structure-based aspects of organizational forms in addition to focusing on identity-related
aspects. More cases of mimicry have to be studied in more detail in order to enhance
insights about the extent and prevalence of such tactics. In the context of CPC
populations, the acquisition of clinic capabilities constitutes an organizational innovation
much more profound than organizational relabeling, which could be considered largely
cosmetic. Most ecological analyses use only one definition of organizational form when
investigating change without acknowledging that there is often a close correspondence
between an identity change, a feature change, and a structural change. To understand how
these changes may be connected and contingent upon each other when it comes to
designing a mimicry strategy, analyses may profit from employing all three. However,
the current study may contribute to a better understanding of the importance of
considering all three dimensions even though its focus was restricted to label
transformations.
Another aspect of mimicry that invites additional research revolves around tracing
its effects on the model population. As described previously, the model-mimic-dupe triad
creates an interrelated space in which the evolution of one population affects the
242
others (Pasteur, 1982; Wickler, 1968). The current study emphasized studying the mimic
population without examining the effects of its actions on the fate of the models. Previous
research suggests that legitimacy effects transcend population boundaries (Dobrev, et al.,
2006) and evidence from aggressive mimicry in biology attests to the adversarial effects
of mimicry for the models. The consequences of astroturfing, for example, entail that
genuine grassroots movements lose authenticity if key stakeholders repeatedly find that
allegedly authentic members of the movement have coopted its practices and
appearances. The continuing decline of the total density of RHPs indicates that the
growing CPC population may play a part in their demise, but the exact interdependencies
remain to be investigated empirically. Likewise, little is known about the relationship
between mimics and dupes and how mimics are able to sustain them over extended
periods of time.
Given that mimicry is not a static strategy but depends on sequential moves that
are dictated by the actions of both mimics and models, the evolutionary arms race
characteristic for mimicry dynamics could be studied fruitfully in the context of
organizational communities. Mimics and models may both engage in definitional
struggles about their forms (McCaffrey & Keys, 2000), models are likely to seek
differentiation strategies in order to distance themselves from the impostors, and mimics
could be found to focus on establishing common ground and drawing on the legitimacy
bestowed on their models. In the context of CPCs, several large campaigns have been
launched by pro-choice organizations in the past in order to expose CPCs as "fake
clinics," but so far, such attempts have not been successful. However, Rao and Giorgi
(2006) describe a case in which the efforts of the model to expose the mimic in fact
243
thwarted their strategies. As they explain, "many of the authentic microbrewers rejected
the practices of contract brewers and labeled them as 'marketing people' and as faux,
fraudulent brewers" (Rao & Giorgi, 2006, p. 286), and their awareness-raising campaign
ultimately led to the demise of contract brewers (Carroll & Swaminathan, 2000).
The ethical implications of aggressive mimicry strategies represent the last area
for future research mentioned. However, they constitute maybe the most important aspect
of studying mimicry. Ultimately, the prevalence of varying degrees of aggressive
mimicry within social movements suggests that a critical examination of the strategies of
many NGOs is warranted. All organizations have agendas, and while the goals of
companies and lobbyists are understood as profit-oriented, the agendas of charitable
organizations are often evaluated based on the assumption that they are motivated by
contributing to a greater good (Steinberg, 2006). Ambiguous strategies such as mimicry
are just one type of social movement tactic with which NGOs are pursuing particularistic
goals through questionable means. Social movement actors have, for example, been
found to employ illegitimate methods such as tree-spiking and road-blocking in order to
increase their organizational legitimacy towards various stakeholder groups (Elsbach &
Sutton, 1992).
Most observers would agree that in the spirit of a participatory Tocquevillian civil
society, the pro-life advocacy exhibited by CPCs is acceptable and must be tolerated.
However, the strategies of CPCs also appear to lack in appropriateness and raise serious
questions about the ethical implications of such strategic choices as they entail assuming
the identity of a reproductive health care provider such as Planned Parenthood. The
discussion of the ethical boundaries of mimicry is connected with a larger ongoing
244
debate about the boundaries of civil society and the "dark side" of social movements
(Edwards & Foley, 1996). Again, CPCs constitute a prime example of the challenges that
arise when the ethical aspects of mimicry strategies are evaluated. In many regions of the
United States, CPC clinics are increasingly gaining core positions when it comes to
providing prenatal health care services to low-income women. As RHPs face restrictions
that do not allow them to extend their services without fees, CPCs often represent the
only option for poor clients acquiring ultrasounds and some material aid, even though
their range of services is restricted to those in line with pro-life ideology. In light of these
developments, it is certainly difficult to envision regulatory penalties for CPCs that
would bar them from acquiring RHP-related features. On the other hand, the deceptive
elements typical for aggressive mimicry are themselves reasons for concerns that require
a careful and balanced ethical examination.
In general, preliminary research suggests that the large-scale entry of faith-based
organizations such as CPCs into the provision of health-care related services (DeHaven,
Hunter, Wilder, Walton, & Berry, 2004; Vanderwoerd, 2004; Wilson, 2003) is associated
with significant transformations in the mix of services offered (Cadge & Wuthnow, 2006;
Esparza, 2007). Tucker (1994) notes that the results of ecological studies can and should
inform policy-makers who construct the regulatory environment of organizations.
Additional research in this important area is needed, as it might thus be able to guide the
creation of policies governing the nonprofit sector in general and the field of reproductive
health care in particular.
In summary, the current study contributes to the field of institutional ecology,
which aims at creating "a more sophisticated theoretical understanding of the co-
245
evolving nature of cultural understandings, organizational forms, and resource
constraints" (Baum & Powell, 1995, p. 536). The introduction of the concept of
aggressive mimicry into organizational ecology aided in a critical exploration of several
important theoretical assumptions derived from this approach. The findings provide
preliminary evidence that some theoretical revisions might be justified as they suggest
that the occurrence of aggressive mimicry strategies changes some of the commonly
observed ecological dynamics in organizational communities. Additionally, the study
contributes to research on American social movements. It examined the evolution of
CPCs and RHPs, which constitute an understudied, but prominent segment of the
American pro-life and reproductive health care communities. The questions raised by a
number of its results may give rise to future inquiry into many other aspects associated
with aggressive mimicry as a viable strategy of organizational populations.
246
BIBLIOGRAPHY
Abortions: Webster's quotations, facts and phrases (2008). San Diego, CA: ICON Group
International.
Abrahamson, E. (1997). The emergence and prevalence of employee management
rhetorics: The effects of long waves, labor unions, and turnover, 1875 to 1992
Academy of Management Journal, 40, 491-533.
Abrahamson, E., & Fairchild, G. (1999). Management fashion: Lifecycles, triggers, and
collective learning processes. Administrative Science Quarterly, 44, 708-740.
Abrahamson, E., & Fairchild, G. (2001). Knowledge industries and idea entrepreneurs:
New dimensions of innovative products, services, and organizations. In C. B.
Schoonhoven & E. Romanelli (Eds.), The entrepreneurship dynamic: Origins of
entrepreneurship and the evolution of industries (pp. 147-177). Stanford, CA:
Stanford University Press.
Afifi, W. A., & Metts, S. (1998). Characteristics and consequences of expectation
violations in close relationships. Journal of Social and Personal Relationships,
15, 365-392.
Aldrich, H. E. (1999). Organizations evolving. Thousand Oaks, CA: Sage.
Aldrich, H. E. (2008). Organizations and environments. Stanford, CA: Stanford
University Press.
Aldrich, H. E., & Baker, W. E. (2001). Learning and legitimacy: Entrepreneurial
responses to constraints on the emergence of new populations and organizations.
In C. B. Schoonhoven & E. Romanelli (Eds.), The entrepreneurship dynamic:
Origins of entrepreneurship and the evolution of industries (pp. 207-235).
Stanford, CA: Stanford University Press.
Aldrich, H. E., & Fiol, C. M. (1994). Fools rush in? The institutional context of industry
creation. Academy of Management Review, 19, 645-670.
247
Aldrich, H. E., & Pfeffer, J. (1976). Environments of organizations. Annual Review of
Sociology, 2, 79-105.
Aldrich, H. E., & Ruef, M. (2006). Organizations evolving (2nd ed.). Thousand Oaks,
CA: Sage.
Alexander, C. (1964). Notes on the synthesis of form. Cambridge, MA: Harvard
University Press.
Allaire, G., & Wolf, S. A. (2004). Cognitive representations and institutional hybridity in
agrofood innovation. Science, Technology, and Human Values, 29, 431-458.
Allison, P. (1984). Event history analysis: Regression for longitudinal event data.
Newbury Park, CA: Sage.
Amburgey, T. L., Dacin, T., & Kelly, D. (1994). Disruptive selection and population
segmentation: Interpopulation competition as a segregating process. In J. A. C.
Baum & J. V. Singh (Eds.), Evolutionary dynamics of organizations (pp. 240-
254). New York: Oxford University Press.
Amburgey, T. L., & Singh, J. V. (2002). Organizational evolution. In J. A. C. Baum
(Ed.), The Blackwell companion to organizations (pp. 327-343). Malden, MA:
Blackwell.
Anderson, P., & Tushman, M. L. (2001). Organizational environments and industry exit:
The effects of uncertainty, munificence and complexity. Industrial and Corporate
Change, 10, 675-711.
Anheier, H. K. (2005). Nonprofit organizations: Theory, management, policy. New York:
Routledge.
Arruñada, B., González-Diaz, M., & Fernández, A. (2004). Determinants of
organizational form: Transaction costs and institutions in the European trucking
industry. Industrial and Corporate Change, 13, 867-882.
Ashforth, B. E., & Humphrey, R. H. (1997). The ubiquity and potency of labeling in
organizations. Organization Science, 8, 43-58.
248
Astley, W. G. (1985). The two ecologies: Population and community perspectives on
organizational evolution. Administrative Science Quarterly, 30, 224-241.
Astley, W. G., & Fombrun, C. J. (1983). Collective strategy: Social ecology of
organizational environments. Academy of Management Review, 8, 576-587.
Audia, P. G., Freeman, J. H., & Reynolds, P. D. (2006). Organizational foundings in
community context: Instruments manufacturers and their interrelationships with
other organizations. Administrative Science Quarterly, 51, 381-419.
Baird-Windle, P., & Bader, E. J. (2001). Targets of hatred: Anti-abortion terrorism. New
York: Palgrave St. Martin's.
Baker, A. (1995). Abortion and options counseling: A comprehensive reference (Rev. and
expanded ed.). Granite City, IL: Hope Clinic for Women.
Baker, A. J. (1991). A model of competition and monopoly in the record industry.
Journal of Cultural Economics, 15, 29-53.
Barnett, M. L., & Carroll, G. R. (1987). Competition and mutualism among early
telephone companies. Administrative Science Quarterly, 32, 400-421.
Barnett, W. P. (1997). The dynamics of competitive intensity. Administrative Science
Quarterly, 42, 128-160.
Barnett, W. P., Swanson, A.-N., & Sorenson, O. (2003). Asymmetric selection among
organizations. Industrial and Corporate Change, 12, 673-695.
Barnett, W. P., & Woywode, M. (2004). From Red Vienna to the Anschluss: Ideological
competition among Viennese newspapers during the rise of National Socialism.
American Journal of Sociology, 109, 1452-1499.
Baron, J. N. (2004). Employing identities in organizational ecology. Industrial and
Corporate Change, 13, 3-32.
249
Barreto, I., & Baden-Fuller, C. (2006). To conform or to perform? Mimetic behaviour,
legitimacy-based groups and performance consequences. Journal of Management
Studies, 43, 1595-1581.
Barron, D. N. (1992). The analysis of count data: Overdispersion and autocorrelation.
Sociological Methodology, 22, 179-220.
Bates, H. W. (1862). Contributions to an insect fauna of the Amazon Valley.
Lepidoptera: Heliconidae. Transactions of the Linnean Society London, 23, 495-
566.
Baum, J. A. C., & Haveman, H. A. (1997). Love thy neighbor? Differentiation and
agglomeration in the Manhattan hotel industry, 1898-1990. Administrative
Science Quarterly, 42, 304-338.
Baum, J. A. C., & Oliver, C. (1991). Institutional linkages and organizational mortality.
Administrative Science Quarterly, 36, 187-218.
Baum, J. A. C., & Powell, W. W. (1995). Cultivating an institutional ecology of
organizations: Comment on Hannan, Carroll, Dundon, and Torres. American
Sociological Review, 60, 529-538.
Baum, J. A. C., & Singh, J. V. (1994a). Evolutionary dynamics of organizations. New
York: Oxford University Press.
Baum, J. A. C., & Singh, J. V. (1994b). Organizational hierarchies and evolutionary
processes: Some reflections on a theory of organizational evolution. In J. A. C.
Baum & J. V. Singh (Eds.), Evolutionary dynamics of organizations (pp. 3-20).
New York: Oxford University Press.
Baum, J. A. C., & Singh, J. V. (1994c). Organizational niches and the dynamics of
organizational founding. Organization Science, 5, 483-501.
Baum, J. A. C., & Singh, J. V. (1996). Dynamics of organizational responses to
competition. Social Forces, 74, 1261-1297.
250
Berry, J. T. (2002). Licensing a choice: “Choose life” specialty license plate and their
constitutional implications. Emory Law Journal, 51, 1605-1644.
Berry, W. D., Ringquist, E. J., Fording, R. C., & Hanson, R. L. (1998). Measuring citizen
and government ideology in the American States, 1960-93. American Journal of
Political Science, 42, 327-348.
Blanchard, D. A. (1994). The anti-abortion movement and the rise of the religious right:
From polite to fiery protest. New York: Twayne.
Blossfeld, H.-P., Golsch, K., & Rohwer, G. (2007). Event history analysis with Stata.
Mahwah, NJ: Erlbaum.
Blossfeld, H.-P., & Rohwer, G. (2002). Techniques of event history modeling: New
approaches to causal analysis (2nd ed.). Mahwah, NJ: Erlbaum.
Bookstaber, R. (2007). A demon of our own design: Markets, hedge funds, and the perils
of financial innovation. Hoboken, NJ: Wiley.
Boone, C., van Witteloostuijn, A., & Carroll, G. R. (2002). Resource distributions and
market partitioning: Dutch daily newspapers, 1968 to 1994. American
Sociological Review, 67, 408-431.
Boris, E. T., & Steuerle, C. E. (2006). Scope and dimensions of the nonprofit sector. In
W. W. Powell & R. Steinberg (Eds.), The nonprofit sector: A research handbook
(2nd ed., pp. 66-88). New Haven, CT: Yale University Press.
Boslaugh, S., & Watters, P. A. (2008). Statistics in a nutshell. Sebastopol, CA: O'Reilly
Media.
Box-Steffensmeier, J. M., & Jones, B. S. (2004). Event history modeling: A guide for
social scientists. Cambridge, NY: Cambridge University Press.
Brace, P., Arceneaux, K., Johnson, M., & Ulbig, S. G. (2004). Does state political
ideology change over time? Political Research Quarterly, 57, 529-540.
251
Breiger, R. L., & Mohr, J. W. (2004). Institutional logics from the aggregation of
organizational networks: Operational procedures for the analysis of counted data.
Computational and Mathematical Organization Theory, 10, 17-43.
Brower, L. P., Brower, J. V. Z., & Westcott, P. W. (1960). Experimental studies of
mimicry. 5. The reactions of toads (Bufo terrestris) to bumblebees (Bombus
americanorum) and their robberfly mimics (Mallophora bomboides), with a
discussion of aggressive mimicry. American Naturalist, 94, 343-355.
Brown, E., & Slivinski, A. (2006). Nonprofit organizations and the market. In W. W.
Powell & R. Steinberg (Eds.), The nonprofit sector: A research handbook (2nd
ed., pp. 140-158). New Haven, CT: Yale University Press.
Brüderl, J., & Schüssler, R. (1990). Organizational mortality: The liabilities of newness
and adolescence. Administrative Science Quarterly, 35, 530-547.
Brummans, B. H. J. M., Putnam, L. L., Gray, B., Hanke, R., Lewicki, R. J., & Wiethoff,
C. (2008). Making sense of intractable multiparty conflict: A study of framing in
four environmental disputes. Communication Monographs, 75, 25-51.
Bryant, J. A., & Monge, P. R. (2008). The evolution of the children’s television
community, 1953-2003 International Journal of Communication, 2, 160-192.
Bryce, D. J., & Singh, J. V. (2001). The future of the firm from an evolutionary
perspective. In P. J. DiMaggio (Ed.), The twenty-first-century firm: Changing
economic organization in international perspective (pp. 160-185). Princeton, NJ:
Princeton University Press.
Burgelman, R. A. (1990). Strategy-making and organizational ecology: A conceptual
integration. In J. V. Singh (Ed.), Organizational evolution: New directions (pp.
164-181). Newbury Park, CA: Sage.
Burgoon, J. K., & Hale, J. L. (1988). Nonverbal expectancy violations: Model elaboration
and application to immediacy behaviors. Communication Monographs, 55, 58-79.
252
Burt, R. S., & Talmud, I. (1993). Market niche. Social Networks, 15, 133-149.
Buttenweiser, S., & Levine, R. (1990). Breaking silences: A post-abortion support model.
In M. G. Fried (Ed.), From abortion to reproductive freedom: Transforming a
movement (pp. 121-130). Boston, MA: South End Press.
Cadge, W., & Wuthnow, R. (2006). Religion and the nonprofit sector. In W. W. Powell
& R. Steinberg (Eds.), The nonprofit sector: A research handbook (2nd ed., pp.
485-505). New Haven, CT: Yale University Press.
Calhoon, M. (2007, Summer). Federally funded lies: Legislation aims to curb deceptive
advertising by crisis pregnancy centers. Ms. Magazine, 35, 18.
Calori, R., Johnson, G., & Sarnin, P. (1992). French and British top managers'
understanding of the structure and the dynamics of their industries: A cognitive
analysis and comparison. British Journal of Management, 3(2), 61-78.
Campbell, D. T. (1965). Variation and selective retention in socio-cultural evolution. In
H. R. Barringer, G. I. Blanksten & R. W. Mack (Eds.), Social change in
developing areas: A reinterpretation of evolutionary theory (pp. 19-48).
Cambridge, MA: Schenkman.
Carley, K. M. (2010a). AutoMap (Version 3.0.3) [Computer software and manual].
Retrieved May 12, 2010, from
http://www.casos.cs.cmu.edu/projects/automap/software.php
Carley, K. M. (2010b). *ORA (Version 2.0.8) [Computer software]. Retrieved June 8,
2010, from http://www.casos.cs.cmu.edu/projects/ora/software.php
Carpenter, G. D. H. (1933). Mimicry. London: Methuen.
Carroll, G. R. (1988). Ecological models of organizations. Cambridge, MA: Ballinger.
Carroll, G. R. (1997). Long-term evolutionary change in organizational populations:
Theory, models and empirical findings in industrial demography. Industrial and
Corporate Change, 6, 119-143.
253
Carroll, G. R., Bigelow, L. S., Seidel, M.-D. L., & Tsai, L. (1996). The fates of de novo
and de alio producers in the American automobile industry 1885-1981. Strategic
Management Journal, 17, 117-137.
Carroll, G. R., Delacroix, J., & Goodstein, J. (1990). The political environment of
organizations: An ecological view. In L. L. Cummings & B. M. Staw (Eds.), The
evolution and adaptation of organizations (pp. 67-101). Greenwich, CT: JAI.
Carroll, G. R., Dobrev, S. D., & Swaminathan, A. (2002). Organizational processes of
resource partitioning. Research in Organizational Behavior, 24, 1-40.
Carroll, G. R., & Hannan, M. T. (1989a). Density delay in the evolution of organizational
populations: A model and five empirical tests. In J. V. Singh (Ed.),
Organizational evolution: New directions (pp. 103-123). Newbury Park, CA:
Sage.
Carroll, G. R., & Hannan, M. T. (1989b). Density dependence in the evolution of
newspaper organizations. American Sociological Review, 54, 524-541.
Carroll, G. R., & Hannan, M. T. (2000). The demography of corporations and industries.
Princeton, NJ: Princeton University Press.
Carroll, G. R., & Swaminathan, A. (2000). Why the microbrewery movement?
Organizational dynamics of resource partitioning in the U.S. brewing industry.
American Journal of Sociology, 106, 715-762.
Cattani, G., Pennings, J. M., & Wezel, F. C. (2003). Spatial and temporal heterogeneity
in founding patterns. Organization Science, 14, 670-685.
Cerulo, K. A. (2002). Culture in mind: Toward a sociology of culture and cognition. New
York: Routledge.
Chandler, M. A. (2006, September 9). Antiabortion centers offer sonograms to further
cause. The Washington Post. Retrieved May 1, 2010, from
http://www.washingtonpost.com/wp-dyn/content/article/2006/09/08/
AR2006090801967.html
254
Child, J. (1972). Organizational structure, environment and performance: The role of
strategic choice. Sociology, 6, 1-22.
Clarke, A. E. (1991). Social worlds/arenas theory as organizational theory. In D. R.
Maines (Ed.), Social organization and social process: Essays in honor of Anselm
Strauss (pp. 119-157). New York: Aldine de Gruyter.
Cleves, M., Gould, W. W., Gutierrez, R. G., & Marchenko, Y. U. (2008). An introduction
to survival analysis using Stata. College Station, TX: Stata Press.
Conell, C., & Cohn, S. (1995). Learning from other people's actions: Environmental
variation and diffusion in French coal mining strikes, 1890-1935. American
Journal of Sociology, 101, 366-403.
Contractor, N. S., Wasserman, S., & Faust, K. (2006). Testing multitheoretical, multilevel
hypotheses about organizational networks: An analytic framework and empirical
example. Academy of Management Review, 31, 681-703.
Conway, L. A. (2000). Directors of crisis pregnancy centers: A look at job satisfaction
and burnout. Unpublished master's thesis, East Carolina University, Greenville,
North Carolina.
Csikszentmihalyi, M., & Rochberg-Halton, E. (1981). The meaning of things: Domestic
symbols and the self. New York: Cambridge University Press.
D'Aunno, T., Succi, M. J., & Alexander, J. A. (2000). The role of institutional and market
forces in divergent organizational change. Administrative Science Quarterly, 45,
679-703.
Dahrendorf, R. (1964). Homo sociologicus (5th ed.). Cologne, Germany: Westdeutscher
Verlag.
Davis, F. (1991). Identity ambivalence in clothing: The dialectic of the erotic and the
chaste. In D. R. Maines (Ed.), Social organization and social process: Essays in
honor of Anselm Strauss (pp. 105-117). New York: Aldine de Gruyter.
255
de Chernatony, L., Daniels, K., & Johnson, G. (1993). A cognitive perspective on
managers' perceptions of competition. Journal of Marketing Management, 9, 373-
381.
Deephouse, D. L. (1996). Does isomorphism legitimate? Academy of Management
Journal, 39, 1024-1039.
DeHaven, M. J., Hunter, I. B., Wilder, L., Walton, J. W., & Berry, J. (2004). Health
programs in faith-based organizations: Are they effective? American Journal of
Public Health, 94, 1030-1036.
Delacroix, J., Swaminathan, A., & Solt, M. E. (1989). Density dependence versus
population dynamics: An ecological study of failings in the California wine
industry. American Sociological Review, 54, 245-262.
Diamond, S. (1989). Spiritual warfare: The politics of the Christian right. Boston, MA:
South End.
DiMaggio, P. J. (1986). Structural analysis of organizational fields: A blockmodel
approach. Research in Organizational Behavior, 8, 335-370.
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional
isomorphism and collective rationality in organizational fields. American
Sociological Review, 48, 147-160.
Dimmick, J. W. (2003). Media competition and coexistence: The theory of the niche.
Mahwah, NJ: Erlbaum.
Djelic, M.-L., & Ainamo, A. (1999). The coevolution of new organizational forms in the
fashion industry: A historical and comparative study of France, Italy, and the
United States. Organization Science, 10, 622-637.
Doan, A. E. (2007). Opposition and intimidation: The abortion wars and strategies of
political harassment. Ann Arbor, MA: University of Michigan Press.
Dobbin, F., & Dowd, T. J. (1997). How policy shapes competition: Early railroad
foundings in Massachusetts. Administrative Science Quarterly, 42, 501-529.
256
Dobrev, S. D. (2001). Revisiting organizational legitimation: Cognitive diffusion and
sociopolitical factors in the evolution of Bulgarian newspaper enterprises,
1846-1992. Organization Studies, 22, 419-444.
Dobrev, S. D., Kim, T.-Y., & Carroll, G. R. (2003). Shifting gears, shifting niches:
Organizational inertia and change in the evolution of the U.S. automobile
industry, 1885-1981. Organization Science, 14, 264-282.
Dobrev, S. D., Kim, T.-Y., & Hannan, M. T. (2001). Dynamics of niche width and
resource partitioning. American Journal of Sociology, 106, 1299-1337.
Dobrev, S. D., Ozdemir, S. Z., & Teo, A. C. (2006). The ecological interdependence of
emergent and established organizational populations: Legitimacy transfer,
violation by comparison, and unstable identities. Organization Science, 17,
577-597.
Durkheim, É., & Mauss, M. (1963). Primitive classification (R. Needham, Trans.).
Chicago, IL: University of Chicago Press. (Original work published 1903).
Ebaugh, H. R. F., Chafetz, J. S., & Pipes, P. F. (2005). Faith-based social service
organizations and government funding: Data from a national survey. Social
Science Quarterly, 86, 273-292.
Edwards, B., & Foley, M. W. (1996). The paradox of civil society. Journal of
Democracy, 7, 38-52.
Eisenberg, E. M. (1984). Ambiguity as strategy in organizational communication.
Communication Monographs, 51, 227-242.
Eisenberg, E. M. (2007). Strategic ambiguities: Essays on communication, organization,
and identity. Thousand Oaks, CA: Sage.
Elsbach, K. D., & Sutton, R. I. (1992). Acquiring organizational legitimacy through
illegitimate actions: A marriage of institutional and impression management
theories. Academy of Management Journal, 35, 699-738.
257
Entsminger, L. (2005). 2005 Care Net client marketing research. The Center of
Tomorrow Journal, 5-23. (Available from Care Net, 44180 Riverside Parkway,
Suite 44200, Lansdowne, VA 20176).
Esau, J. C. (2005). Start and grow your faith-based nonprofit: Answering your call in the
service of others. San Francisco, CA: Jossey-Bass.
Esparza, N. E. (2007). Shelters, soup kitchens, and supportive housing: An open systems
analysis of the field of homeless assistance organizations (Doctoral dissertation,
Princeton University, 2007). Dissertation Abstracts International, 68, Publication
No. AAT 3271634.
Fang, L. (2009, July 28). Exclusive: Infamous astroturf lobbying firm behind new anti-
health reform group. Retrieved August 5, 2009, from http://thinkprogress.org/
2009/07/28/cppr-dci-astroturf/
Feldman, M. S. (2003). A performative perspective on stability and change in
organizational routines. Industrial and Corporate Change, 12, 727-752.
Fiol, C. M., & O'Connor, E. J. (2006). Stuff matters: Artifacts, social identity, and
legitimacy in the U.S. medical profession. In A. Rafaeli & M. G. Pratt (Eds.),
Artifacts and organizations: Beyond mere symbolism (pp. 241-257). Mahwah, NJ:
Erlbaum.
Fombrun, C. J. (1988). Crafting an institutionally informed ecology of organizations. In
G. R. Carroll (Ed.), Ecological models of organizations (pp. 223-239).
Cambridge, MA: Ballinger.
Freeman, J. H. (1995). Business strategy from the population level. In C. A. Montgomery
(Ed.), Resource-based and evolutionary theories of the firm: Towards a synthesis
(pp. 219-250). London: Kluwer.
Freeman, J. H., & Audia, P. G. (2006). Community ecology and the sociology of
organizations. Annual Review of Sociology, 32, 145-169.
Freeman, J. H., & Hannan, M. T. (1983). Niche width and the dynamics of organizational
populations. American Journal of Sociology, 88, 1116-1145.
258
Fung, W., & Hsieh, D. A. (1999). A primer on hedge funds. Journal of Empirical
Finance, 6, 309-331.
Galaskiewicz, J., & Bielefeld, W. (1998). Nonprofit organizations in an age of
uncertainty: A study of organizational change. New York: Aldine de Gruyter.
Gamson, W. A. (1995). Constructing social protest. In H. Johnston & B. Klandermans
(Eds.), Social movements and culture (pp. 85-106). Minneapolis, MN: University
of Minnesota Press.
Gerschlager, C. (2005). Introduction. In C. Gerschlager (Ed.), Deception in markets: An
economic analysis (pp. 1-22). New York, N.Y.: Palgrave Macmillan.
Ginsburg, F. D. (1989). Contested lives: The abortion debate in an American community.
Berkeley, CA: University of California Press.
Giving USA Foundation (2007). Giving USA 2007: The annual report on philanthropy
for the year 2006. Glenview, IL: Giving USA Foundation.
Glessner, T. A. (2001, Spring). The language of life vs. the language of choice: A
comment on the power of words in the abortion debate. At the Center, 2(2).
Retrieved September 22, 2007, from http://www.atcmag.com/v2n2/article4.asp
Glessner, T. A., & O'Connor, A. J. (2007, Fall). Evangelism: The core of the pregnancy
resource center/clinic ministry. At the Center, 8(3). Retrieved September 22,
2007, from http://www.atcmag.com/v8n3/article11.asp
Glynn, M. A., & Abzug, R. (2002). Institutionalizing identity: Symbolic isomorphism
and organizational names. Academy of Management Journal, 45, 267-280.
Glynn, M. A., & Marquis, C. (2004). When good names go bad: Symbolic illegitimacy in
organizations. Research in the Sociology of Organizations, 22, 147-170.
Glynn, M. A., & Marquis, C. (2006). Fred's Bank: How institutional norms and
individual preferences legitimate organizational names. In A. Rafaeli & M. G.
Pratt (Eds.), Artifacts and organizations: Beyond mere symbolism (pp. 223-239).
Mahwah, NJ: Erlbaum.
259
Glynn, M. A., & Marquis, C. (2007). Legitimating identities: How institutional logics
motivate organizational name choices. In C. Bartel, S. L. Blader & A.
Wrzesniewski (Eds.), Identity and the modern organization (pp. 17-33). Mahwah,
NJ: Erlbaum.
Gordon, T. P., Greenlee, J. S., & Nitterhouse, D. (1999). Tax-exempt organization
financial data: Availability and limitations. American Horizons, 13(2), 113-128.
Green, S. E., Li, Y., & Nohria, N. (2009). Suspended in self-spun webs of significance: A
rhetorical model of institutionalization and institutionally embedded agency.
Academy of Management Journal, 52, 11-36.
Greif, A. (1996). The study of organizations and evolving organizational forms through
history: Reflections from the late medieval family firm. Industrial and Corporate
Change, 5, 473-502.
Greve, H. R. (2002). Interorganizational evolution. In J. A. C. Baum (Ed.), The Blackwell
companion to organizations (pp. 557-578). Malden, MA: Blackwell.
Greve, H. R., Pozner, J.-E., & Rao, H. (2006). Vox populi: Resource partitioning,
organizational proliferation, and the cultural impact of the insurgent microradio
movement. American Journal of Sociology, 112, 802-837.
Griebel, U., & Oller, D. K. (2008). Evolutionary forces favoring communicative
flexibility. In D. K. Oller & U. Griebel (Eds.), Evolution of communicative
flexibility: Complexity, creativity, and adaptability in human and animal
communication (pp. 9-40). Cambridge, MA: MIT Press.
Guo, G. (1993). Event-history analysis for left-truncated data. Sociological Methodology,
23, 217-243.
Halva-Neubauer, G. A. (1993). The states after Roe: No “paper tigers”. In M. L. Goggin
(Ed.), Understanding the new politics of abortion (pp. 167-189). Newbury Park,
CA: Sage.
260
Hamilton, E. A. (2006). An exploration of the relationship between loss of legitimacy and
the sudden death of organizations. Group and Organization Management, 31,
327-358.
Hannan, M. T. (1997). Inertia, density and the structure of organizational populations:
Entries in European automobile industries, 1886-1981. Organization Studies, 18,
193-228.
Hannan, M. T. (2005). Ecologies of organizations: Diversity and identity. Journal of
Economic Perspectives, 19(1), 51-70.
Hannan, M. T., Baron, J. N., Hsu, G., & Koçak, Ö. (2006). Organizational identities and
the hazard of change. Industrial and Corporate Change, 15, 755-784.
Hannan, M. T., & Carroll, G. R. (1992). Dynamics of organizational populations. New
York: Oxford University Press.
Hannan, M. T., Carroll, G. R., & Pólos, L. (2003). The organizational niche. Sociological
Theory, 21, 309-340.
Hannan, M. T., & Freeman, J. H. (1977). The population ecology of organizations.
American Journal of Sociology, 82, 929-964.
Hannan, M. T., & Freeman, J. H. (1984). Structural inertia and organizational change.
American Sociological Review, 49, 149-164.
Hannan, M. T., & Freeman, J. H. (1989). Organizational ecology. Cambridge, MA:
Harvard University Press.
Hannan, M. T., Pólos, L., & Carroll, G. R. (2007). Logics of organization theory:
Audiences, codes, and ecologies. Princeton, NJ: Princeton University Press.
Hartshorn, M. (2003). Pregnancy help centers, abstinence, STDs, and healing: Putting it
all together. In T. R. Wagner (Ed.), Back to the drawing board: The future of the
pro-life movement (pp. 106-122). South Bend, IN: St. Augustine's Press.
261
Haveman, H. A. (1993). Follow the leader: Mimetic isomorphism and entry into new
markets. Administrative Science Quarterly, 38, 593-627.
Hawley, A. H. (1986). Human ecology: A theoretical essay. Chicago, IL: University of
Chicago Press.
Hayes, A. F. (2005). Statistical methods for communication science. Mahwah, NJ:
Erlbaum.
Hensmans, M. (2003). Social movement organizations: A metaphor for strategic actors in
institutional fields. Organization Studies, 23, 355-381.
Hodgkinson, G. P. (1997). Cognitive inertia in a turbulent market: The case of UK
residential estate agents. Journal of Management Studies, 34, 921-945.
Hodgkinson, V. A., Weitzman, M. S., Abrahams, J. A., Crutchfield, E. A., & Stevenson,
D. R. (1997). Nonprofit Almanac 1996-1997: Dimensions of the independent
sector (5th ed.). San Francisco: Jossey-Bass.
Hosmer, D. W., Lemeshow, S., & May, S. (2008). Applied survival anaysis: Regression
modeling of time-to-event data. Hoboken, NJ: Wiley-Interscience.
Howard, J. A. (1994). A social cognitive conception of social structure. Social
Psychology Quarterly, 57, 210-227.
Hsu, G. (2006a). Evaluative schemas and the attention of critics in the U.S. film industry.
Industrial and Corporate Change, 15, 467-496.
Hsu, G. (2006b). Jacks of all trades and masters of none: Audiences’ reactions to
spanning genres in feature film production. Administrative Science Quarterly, 51,
420-450.
Hsu, G., & Hannan, M. T. (2005). Identities, genres, and organizational forms.
Organization Science, 16, 474-490.
262
Hsu, G., Hannan, M. T., & Kocak, Ö. (2009). Multiple category memberships in markets:
An integrative theory and two empirical tests. American Sociological Review, 74,
150-169.
Hsu, G., & Podolny, J. M. (2005). Critiquing the critics: An approach for the comparative
evaluation of critical schemas. Social Science Research, 34, 189-214.
Hudson, B. A. (2008). Against all odds: A consideration of core-stigmatized
organizations. Academy of Management Review, 33, 252-266.
Hudson, B. A., & Okhuysen, G. A. (2009). Not with a ten-foot pole: Core stigma, stigma
transfer, and improbable persistence of men's bathhouses. Organization Science,
20, 134-153.
Huheey, J. E. (1988). Mathematical models of mimicry. The American Naturalist,
131(Suppl.), 22-41.
Hunt, C. S., & Aldrich, H. E. (1998). The second ecology: Creation and evolution of
organizational communities. Research in Organizational Behavior, 20, 267-301.
Ingram, P., & Rao, H. (2004). Store wars: The enactment and repeal of anti-chain-store
legislation in America. American Journal of Sociology, 110, 446-487.
Jackson, R., & Wilcox, R. C. F. (1990). Aggressive mimicry, prey-specific predatory
behaviour and predator-recognition in the predator-prey interactions of Portia
fimbriata and Euryattus sp., jumping spiders from Queensland. Behavioral
Ecology and Sociobiology, 26, 111-119.
Jacobson, L. (2005). Commentary on Care Net client marketing research. The Center of
Tomorrow Journal, 24-30. (Available from Care Net, 44180 Riverside Parkway,
Suite 44200, Lansdowne, VA 20176).
Jepperson, R. L. (1991). Institutions, institutional effects, and institutionalism. In W. W.
Powell & P. J. DiMaggio (Eds.), The new institutionalism in organizational
analysis (pp. 146-163). Chicago, IL: University of Chicago Press.
263
Kahane, L. H. (2000). Anti-abortion activities and the market for abortion services:
Protest as a disincentive. American Journal of Economics and Sociology, 59, 463-
485.
Karafillidis, A. (2009). Entkopplung und Kopplung: Wie die Netzwerktheorie zur
Bestimmung sozialer Grenzen beitragen kann [Decoupling and coupling:
Contributions of network theory to the determination of social boundaries]. In R.
Häußling (Ed.), Grenzen von Netzwerken (pp. 105-131). Wiesbaden, Germany:
VS Verlag.
Kaufmann, K. (1997). The abortion resource handbook. New York: Simon & Schuster.
Kennedy, M. T. (2008). Getting counted: Markets, media, and reality. American
Sociological Review, 73, 270-295.
Kennedy, M. T., & Fiss, P. C. (2009). Institutionalization, framing, and diffusion: The
logic of TQM adoption and implementation decisions among U.S. hospitals.
Academy of Management Journal, 52, 897-918.
Kim, C.-W., & Lee, K. (2003). Innovation, technological regimes and organizational
selection in industry evolution: A 'history friendly model' of the DRAM industry.
Industrial and Corporate Change, 12, 1195-1221.
King, A. A., Lenox, M. J., & Barnett, M. L. (2002). Strategic responses to the reputation
commons problem. In A. J. Hoffman & M. J. Ventresca (Eds.), Organizations,
policy and the natural environment: Institutional and strategic perspectives (pp.
393-406). Stanford, CA: Stanford University Press.
Kloock, C. T. (2000). Aggressive mimicry in the pirate spider, Mimetus notius (Aranae;
Mimetidae) (Doctoral Dissertation, Michigan State University, 2000).
Dissertation Abstracts International, 61, 2340.
Klotz, R. J. (2007). Internet campaigning for grassroots and astroturf support. Social
Science Computer Review, 25(1), 3-12.
Kogut, B., & Zander, U. (1996). What firms do? Coordination, identity, and learning.
Organization Science, 7, 502-518.
264
Kosmin, B. A., Mayer, E., & Keysar, A. (2001, December). American religious
identification survey 2001 (The Graduate Center of the University of New York).
Retrieved May 22, 2010, from
http://www.gc.cuny.edu/faculty/research_briefs/aris.pdf
Krippendorff, K. (2004a). Content analysis: An introduction to its methodology (2nd ed.).
Thousand Oaks, CA: Sage.
Krippendorff, K. (2004b). Reliability in content analysis. Human Communication
Research, 30, 411-433.
Kuilman, J. G., & Li, J. (2009). Grades of membership and legitimacy spillovers: Foreign
banks in Shanghai, 1847-1935. Academy of Management Journal, 52, 229-245.
Lamont, M., & Molnár, V. (2002). The study of boundaries in the social sciences. Annual
Review of Sociology, 28, 167-195.
Lampkin, L. M., & Boris, E. T. (2002). Nonprofit organization data: What we have and
what we need. American Behavioral Scientist, 45, 1675-1715.
Lee, K., & Pennings, J. M. (2002). Mimicry and the market: Adoption of a new
organizational form. Academy of Management Journal, 45, 144-162.
Leskovec, J., Kleinberg, J. M., & Faloutsos, C. (2007). Graph evolution: Densification
and shrinking diameters. ACM Transactions on Knowledge Discovery from Data,
1(1), Article 2.
Levins, R. (1968). Evolution in changing environments: Some theoretical explorations.
Princeton, NJ: Princeton University Press.
Lewin, A. Y., Long, C. P., & Carroll, T. N. (1999). The coevolution of new
organizational forms. Organization Science, 10, 535-550.
Lin, V., & Dailard, C. (2002). Crisis pregnancy centers seek to increase political clout,
secure government subsidy. The Guttmacher Report on Public Policy, 5(2), 4-6.
265
Lloyd, J. E. (1965, August 6). Aggressive mimicry in Photuris: Firefly femmes fatales.
Science, 149, 653-654.
Lloyd, J. E. (1984). Occurrence of aggressive mimicry in fireflies. Florida Entomologist,
67, 368-376.
Lomi, A., & Larsen, E. R. (1996). Interacting locally and evolving globally: A
computational approach to the dynamics of organizational populations. Academy
of Management Journal, 39, 1287-1321.
Lomi, A., Larsen, E. R., & Freeman, J. H. (2005). Things change: Dynamic resource
constraints and system-dependent selection in the evolution of organizational
populations. Management Science, 51, 882-903.
Lounsbury, M., & Rao, H. (2004). Sources of durability and change in market
classifications: A study of the reconstitution of product categories in the
American mutual fund industry, 1944-1985. Social Forces, 82, 969-999.
Lounsbury, M., Ventresca, M. J., & Hirsch, P. M. (2003). Social movements, field frames
and industry emergence: A cultural-political perspective on U.S. recycling. Socio-
Economic Review, 1, 71-104.
Luhmann, N. (2008). Ideenevolution [The evolution of ideas]. Frankfurt am Main,
Germany: Suhrkamp.
Lumsden, C. J., & Singh, J. V. (1990). The dynamics of organizational speciation. In J.
V. Singh (Ed.), Organizational evolution: New directions (pp. 145-163). Newbury
Park, CA: Sage Publications.
Lyon, T. P., & Maxwell, J. W. (2004). Astroturf: Interest group lobbying and corporate
strategy. Journal of Economics and Management Strategy, 13, 561-597.
Maguire, S., & Hardy, C. (2009). Discourse and deinstitutionalization: The decline of
DDT. Academy of Management Journal, 52, 148-178.
266
Mallet, J., & Joron, M. (2007). Evolution of diversity in warning color and mimicry:
Polymorphisms, shifting balance, and speciation. Annual Review of Ecology and
Systematics, 30, 201-233.
Marquis, C. (2003). The pressure of the past: Network imprinting in intercorporate
communities. Administrative Science Quarterly, 48, 655-689.
Marrett, C. B. (1980). Influences on the rise of new organizations: The formation of
women's medical societies. Administrative Science Quarterly, 25, 185-199.
Mathewes-Green, F. (1996, July). Pro-life dilemma. Policy Review, 78, 40-43.
Maxwell, C. J. C. (2002). Pro-life activists in America: Meaning, motivation, and direct
action. Cambridge, NY: Cambridge University Press.
Mazza, C., & Alvarez, J. L. (2000). Haute couture and prêt-á-porter: The popular press
and the diffusion of management practices. Organization Studies, 21, 567-588.
McBride, D. E. (2008). Abortion in the United States: A reference handbook. Santa
Barbara, CA: ABC-CLIO.
McCaffrey, D., & Keys, J. (2000). Competitive framing processes in the abortion debate:
Polarization-vilification, frame saving, and frame debunking. Sociological
Quarterly, 41, 41-61.
McCarthy, I. P., & Gillies, J. M. (2003). Organisational diversity, configurations and
evolution. In E. Mitleton-Kelly (Ed.), Complex systems and evolutionary
perspectives on organisations: The application of complexity theory to
organisations (pp. 71-97). Boston, MA: Elsevier.
McCarthy, J. D., Wolfson, M., Baker, D. P., & Mosakowski, E. (1988). The founding of
social movement organizations: Local citizens’ groups opposing drunken driving.
In G. R. Carroll (Ed.), Ecological models of organizations (pp. 70-84).
Cambridge, MA: Ballinger.
McCarthy, J. D., & Zald, M. N. (1977). Resource mobilization and social movements: A
partial theory. American Journal of Sociology, 82, 1212-1241.
267
McKelvey, B. (1982). Organizational systematics: Taxonomy, evolution, classification.
Berkeley: University of California Press.
McKelvey, B. (1994). Evolution and organizational science. In J. A. C. Baum & J. V.
Singh (Eds.), Evolutionary dynamics of organizations (pp. 314-326). New York:
Oxford University Press.
McKendrick, D. G., & Carroll, G. R. (2001). On the genesis of organizational forms:
Evidence from the market for disk drive arrays. Organization Science, 12, 661-
683.
McKendrick, D. G., Jaffee, J., Carroll, G. R., & Khessina, O. M. (2003). In the bud? Disk
array producers as a (possibly) emergent organizational form. Administrative
Science Quarterly, 48, 60-93.
McLaughlin, P., & Haas, R. (2003, Fall). When is your due date? At the Center, 4(5).
Retrieved September 22, 2007, from http://www.atcmag.com/v4n3/article5.asp
McPherson, J. M. (1990). Evolution in communities of voluntary organizations. In J. V.
Singh (Ed.), Organizational evolution: New directions (pp. 224-245). Newbury
Park, CA: Sage Publications.
Mead, G. H. (1934). Mind, self and society. Chicago: University of Chicago Press.
Merton, R. K. (1957). The role-set: Problems in sociological theory. British Journal of
Sociology, 8, 106-120.
Mertus, J. A. (1990). Fake abortion clinics: The threat to reproductive self-determination.
Women and Health, 16, 95-113.
Meyer, D. S., & Staggenborg, S. (1996). Movements, countermovements, and the
structure of political opportunity. American Journal of Sociology, 101, 1628-
1660.
Meyer, J. W., & Scott, W. R. (1983). Organizational environments: Ritual and
rationality. Beverly Hills, CA: Sage.
268
Meyer, M. W., & Zucker, L. G. (1989). Permanently failing organizations. Newbury
Park, CA: Sage.
Mezias, J. M., & Lant, T. K. (1994). Mimetic learning and the evolution of organizational
populations. In J. A. C. Baum & J. V. Singh (Eds.), Evolutionary dynamics of
organizations (pp. 179-198). New York: Oxford University Press.
Mezias, J. M., & Mezias, S. J. (2000). Resource partitioning, the founding of specialist
firms, and innovation: The American feature film industry, 1912-1929.
Organization Science, 11, 306-322.
Miles, R. E., & Snow, C. C. (1978). Organizational strategy, structure, and process. New
York: McGraw-Hill.
Milstein, M. B., Hart, S. L., & York, A. S. (2002). Coercion breeds variation: The
differential impact of isomorphic pressures on environmental strategies. In A. J.
Hoffman & M. J. Ventresca (Eds.), Organizations, policy and the natural
environment: Institutional and strategic perspectives (pp. 152-172). Stanford,
CA: Stanford University Press.
Miner, A. S., Amburgey, T. L., & Stearns, L. B. (1990). Interorganizational linkages and
population dynamics: Buffering and transformational shields. Administrative
Science Quarterly, 35, 689-713.
Miner, A. S., & Raghavan, S. V. (1999). Interorganizational imitation: A hidden engine
of selection. In J. A. C. Baum & B. McKelvey (Eds.), Variations in organization
science: In honor of Donald T. Campbell (pp. 35-62). New York: Oxford
University Press.
Minkoff, D. C. (1999). Bending with the wind: Strategic change and adaptation by
women's and racial minority organizations. American Journal of Sociology, 104,
1666-1703.
Minkoff, D. C. (2002). Macro-organizational analysis. In B. Klandermans & S.
Staggenborg (Eds.), Methods of social movement research (pp. 260-285).
Minneapolis, MN: University of Minnesota Press.
269
Mische, A., & White, H. C. (1998). Between conversation and situation: Public switching
dynamics across network domains. Social Research, 65, 695-724.
Monahan, D. (2007, Spring/summer). What's in a name? At the Center, 8(2). Retrieved
September 22, 2007, from http://www.atcmag.com/v8n2/article11.asp
Monge, P. R., & Contractor, N. S. (2003). Theories of communication networks. New
York: Oxford University Press.
Monge, P. R., & Poole, M. S. (2008). The evolution of organizational communication.
Journal of Communication, 58, 679-692.
Moore, K., & Hala, N. (2002). Organizing identity: The creation of Science for the
People. Research in the Sociology of Organizations, 19, 209-335.
Moynihan, M. (1968). Social mimicry; Character convergence versus character
displacement. Evolution, 22, 315-331.
Mueller, G. P. (2008). Three-valued modal logic for reconstructing the semantic network
structure of a corpus of coded texts. In T. N. Friemel (Ed.), Why context matters:
Applications of social network analysis (pp. 37-54). Wiesbaden, Germany: VS
Verlag.
Munir, K., & Phillips, N. (2005). The birth of the 'Kodak moment': Institutional
entrepreneurship and the adoption of new technologies. Organization Studies, 26,
1665-1687.
Munson, Z. W. (2002). Becoming an activist: Believers, sympathizers, and mobilization
in the American pro-life movement (Doctoral dissertation, Harvard University,
2002). Dissertation Abstracts International, 63, 368.
Munson, Z. W. (2008). The making of pro-life activists: How social movement
mobilization works. Chicago, IL: University of Chicago Press.
270
Nelson, R. R. (2007). Comment on “Dismantling Lamarckism: Why descriptions of
socio-economic evolution as Lamarckian are misleading” by Hodgson and
Knudsen. Journal of Evolutionary Economics, 17, 349-352.
Nelson, R. R., & Winter, S. G. (1982). An evolutionary theory of economic change.
Cambridge, MA: Belknap Press of Harvard University Press.
Nguyen, P. (2005). Variety and the evolution of refinery processing. Industrial and
Corporate Change, 14, 469-500.
Nicholson-Crotty, J. (2007). Politics, policy, and the motivations for advocacy in
nonprofit reproductive health and family planning providers. Nonprofit and
Voluntary Sector Quarterly, 36, 5-21.
Nownes, A. J. (2004). The population ecology of interest group formation: Mobilizing for
gay and lesbian rights in the United States, 1950-98. British Journal of Political
Science, 34, 49-67.
Okamoto, D. G. (2006). Institutional panethnicity: Boundary formation in Asian-
American organizing. Social Forces, 85, 1-45.
Oliver, A. L. (2009). Networks for learning and knowledge creation in biotechnology.
New York: Cambridge University Press.
Papadakis, E. (1996). Environmental politics and institutional change. New York:
Cambridge University Press.
Pasteur, G. (1982). A classificatory review of mimicry systems. Annual Review of
Ecology and Systematics, 13, 169-199.
Peckham, G. W., & Peckham, E. G. (1892). Ant-like spiders of the family Attidae.
Occasional Papers of the Natural History Society of Wisconsin, 2, 1-84.
Péli, G. (1997). The niche hiker's guide to population ecology: A logical reconstruction
of organization ecology's niche theory. Sociological Methodology, 27, 1-46.
271
Perretti, F., Negro, G., & Lomi, A. (2008). E pluribus unum: Framing, matching, and
form emergence in U.S. television broadcasting, 1940-1960. Organization
Science, 19, 533-547.
Phillips, D. J. (2002). A genealogical approach to organizational life chances: The parent-
progeny transfer among Silicon Valley law firms, 1946-1996. Administrative
Science Quarterly, 47, 474-506.
Phillips, D. J., & Kim, Y.-K. (2008). Why pseudonyms? Deception as identity
preservation among jazz record companies, 1920-1929. Organization Science, 20,
481-499.
Phillips, M. E. (1994). Industry mindsets: Exploring the cultures of two macro-
organizational settings. Organization Science, 5, 384-402.
Phillips, N., Lawrence, T. B., & Hardy, C. (2004). Discourse and institutions. Academy of
Management Review, 29, 635-652.
Pietsch, T. W., & Grobecker, D. B. (1978, July 28). The compleat angler: Aggressive
mimicry in an Antennariid anglerfish. Science, 201, 369-370.
Podolny, J. M. (1993). A status-based model of market competition. American Journal of
Sociology, 98, 829-872.
Pólos, L., Hannan, M. T., & Carroll, G. R. (2002). Foundations of a theory of social
forms. Industrial and Corporate Change, 11, 85-115.
Popielarz, P. A., & McPherson, J. M. (1995). On the edge or in between: Niche position,
niche overlap, and the duration of voluntary association memberships. American
Journal of Sociology, 101, 698-720.
Popielarz, P. A., & Neal, Z. P. (2007). The niche as a theoretical tool. Annual Review of
Sociology, 33, 65-84.
Porac, J. F. (1994). On the concept of "organizational community". In J. A. C. Baum & J.
V. Singh (Eds.), Evolutionary dynamics of organizations (pp. 451-456). New
York: Oxford University Press.
272
Porac, J. F., & Thomas, H. (1990). Taxonomic mental models in competitor definition.
Academy of Management Review, 15, 224-240.
Porac, J. F., Thomas, H., & Baden-Fuller, C. (1989). Competitive groups as cognitive
communities: The case of Scottish knitwear manufacturers. Journal of
Management Studies, 26, 397-416.
Porac, J. F., Thomas, H., Wilson, F., Paton, D., & Kanfer, A. (1995). Rivalry and the
industry model of Scottish knitwear producers. Administrative Science Quarterly,
40, 203-227.
Porac, J. F., Ventresca, M. J., & Mishina, Y. (2002). Interorganizational cognition and
interpretation. In J. A. C. Baum (Ed.), The Blackwell companion to organizations
(pp. 579-598). Malden, MA: Blackwell.
Porac, J. F., Wade, J. B., & Pollock, T. G. (1999). Industry categories and the politics of
the comparable firm in CEO compensation. Administrative Science Quarterly, 44,
112-115.
Powell, W. W. (1990). Neither market nor hierarchy: Network forms of organization.
Research in Organizational Behavior, 12, 295-366.
Powell, W. W., & DiMaggio, P. J. (1991). The new institutionalism in organizational
analysis. Chicago, IL: University of Chicago Press.
Powell, W. W., White, D. R., Koput, K. W., & Owen-Smith, J. (2005). Network
dynamics and field evolution: The growth of interorganizational collaboration in
the life sciences. American Sociological Review, 110, 1132-1205.
Pressman, S. (2009). Charles Ponzi. In S. Matulich & D. M. Currie (Eds.), Handbook of
frauds, scams, and swindles: Failures of ethics in leadership (pp. 35-43). Boca
Raton, FL: Taylor & Francis.
Rao, H. (1998). Caveat emptor: The construction of nonprofit consumer watchdog
organizations. American Journal of Sociology, 103, 912-961.
273
Rao, H. (2002). Interorganizational ecology. In J. A. C. Baum (Ed.), The Blackwell
companion to organizations (pp. 541-556). Malden, MA: Blackwell.
Rao, H. (2009). Market rebels: How activists make or break radical innovations.
Princeton, NJ: Princeton University Press.
Rao, H., & Giorgi, S. (2006). Code breaking: How entrepreneurs exploit cultural logics to
generate institutional change. Research in Organizational Behavior, 27, 269-304.
Rao, H., & Kenney, M. (2008). New forms as settlements. In R. Greenwood, C. Oliver,
R. Suddaby & K. Sahlin (Eds.), The Sage handbook of organizational
institutionalism (pp. 352-370). Thousand Oaks, CA: Sage.
Rao, H., Monin, P., & Durand, R. (2003). Institutional change in Toque Ville: Nouvelle
cuisine as an identity movement in French gastronomy. American Journal of
Sociology, 108, 795-843.
Rao, H., Morrill, C., & Zald, M. N. (2000). Power plays: How social movements and
collective action create new organizational forms. Research in Organizational
Behavior, 22, 237-281.
Rao, H., & Singh, J. V. (1999). Types of variation in organizational populations: The
speciation of new organizational forms. In J. A. C. Baum & B. McKelvey (Eds.),
Variations in organization science: In honor of Donald T. Campbell (pp. 63-77).
New York: Oxford University Press.
Reeves, S. R. (2003). The development of post-abortion syndrome within the crisis
pregnancy center movement in America: A historical-theoretical study.
Unpublished master's thesis, Smith College, Northampton, Massachusetts.
Reger, R. K., & Huff, A. S. (1993). Strategic groups: A cognitive perspective. Strategic
Management Journal, 14, 103-123.
Rhee, M., Kim, Y.-C., & Han, J. (2006). Confidence in imitation: Niche-width strategy in
the UK automobile industry. Management Science, 52, 501-513.
274
Richards, R. A. (2009). Classification in Darwin’s Origin. In M. Ruse & R. J. Richards
(Eds.), The Cambridge companion to the "Origin of species" (pp. 173-193).
Cambridge, NY: Cambridge University Press.
Rindova, V. P., & Fombrun, C. J. (2001). Entrepreneurial action in the creation of the
specialty coffee niche. In C. B. Schoonhoven & E. Romanelli (Eds.), The
entrepreneurship dynamic: Origins of entrepreneurship and the evolution of
industries (pp. 236-261). Stanford, CA: Stanford University Press.
Romanelli, E. (1991). The evolution of new organizational forms. Annual Review of
Sociology, 17, 79-103.
Romanelli, E., & Khessina, O. M. (2005). Regional industrial identity: Cluster
configurations and economic development. Organization Science, 16, 344-358.
Rosa, J. A., & Porac, J. F. (2002). Categorization bases and their influence on product
category knowledge structures. Psychology and Marketing, 19, 503-532.
Ruef, M. (2000). The emergence of organizational forms: A community ecology
approach. American Journal of Sociology, 106, 658-714.
Ruef, M. (2004a). For whom the bell tolls: Ecological perspectives on industrial decline
and resurgence. Industrial and Corporate Change, 13, 61-89.
Ruef, M. (2004b). The demise of an organizational form: Emancipation and plantation
agriculture in the American South, 1860-1880. American Journal of Sociology,
109, 1365-1410.
Ruef, M., & Scott, W. R. (1998). A multidimensional model of organizational legitimacy:
Hospital survival in changing institutional environments. Administrative Science
Quarterly, 43, 877-904.
Ruxton, G. D., Sherratt, T. N., & Speed, M. P. (2004). Avoiding attack: The evolutionary
ecology of crypsis, warning signals, and mimicry. Oxford, NY: Oxford University
Press.
275
Sahlin, K., & Wedlin, L. (2008). Circulating ideas: Imitation, translation and editing. In
R. Greenwood, C. Oliver, R. Suddaby & K. Sahlin (Eds.), The Sage handbook of
organizational institutionalism (pp. 218-242). Thousand Oaks, CA: Sage.
Saini, A. (2008, February 4). Degrees of deception. New Statesman, 137, 19-20.
Sazima, I. (1977, December 8). Possible case of aggressive mimicry in a neotropical
scale-eating fish. Nature, 270, 510-512.
Schoenherr, R. A., & Young, L. A. (1990). Organizational demography and structural
change in the Roman Catholic church. In C. J. Calhoun, M. W. Meyer & W. R.
Scott (Eds.), Structures of power and constraint: Papers in honor of Peter M.
Blau (pp. 235-270). Cambridge, NY: Cambridge University Press.
Schoonhoven, C. B., & Romanelli, E. (2001). Emergent themes and the next wave of
entrepreneurship research. In C. B. Schoonhoven & E. Romanelli (Eds.), The
entrepreneurship dynamic: Origins of entrepreneurship and the evolution of
industries (pp. 383-408). Stanford, CA: Stanford University Press.
Scott, W. R. (1981). Organizations: Rational, natural, and open systems. Englewood
Cliffs, NJ: Prentice-Hall.
Scott, W. R. (1987). The adolescence of institutional theory. Administrative Science
Quarterly, 32, 493-511.
Scott, W. R. (2000). Institutions and organizations (2nd ed.). Thousand Oaks, CA: Sage.
Shumate, M., Fulk, J., & Monge, P. R. (2005). Predictors of the international HIV-AIDS
INGO network over time. Human Communication Research, 31, 482-510.
Sillince, J. A. A., & Suddaby, R. (2008). Organizational rhetoric: Bridging management
and communication scholarship. Management Communication Quarterly, 22,
5-12.
Silverman, J. (2007, May 9). States react to crisis pregnancy centers. The Washington
Post. Retrieved May 1, 2010, from http://www.washingtonpost.com/
wp-dyn/content/article/2007/05/09/AR2007050900291.html
276
Simons, T., & Ingram, P. (2003). Enemies of the state: The interdependence of
institutional forms and the ecology of the Kibbutz, 1910-1997. Administrative
Science Quarterly, 48, 592-621.
Simons, T., & Ingram, P. (2004). An ecology of ideology: Theory and evidence from four
populations. Industrial and Corporate Change, 13, 33-59.
Sine, W. D., & Lee, B. H. (2009). Tilting at windmills? The environmental movement
and the emergence of the U.S. wind energy sector. Administrative Science
Quarterly, 54, 123-155.
Singh, J. V., Tucker, D. J., & House, R. J. (1986). Organizational legitimacy and the
liability of newness. Administrative Science Quarterly, 31, 171-193.
Sommerfeld, D. H. (2008). Environmental influences on local nonprofit and for-profit
market structure: A national study of the nursing home industry (Doctoral
dissertation, University of Michigan, 2008). Dissertation Abstracts International,
68, 4484.
Sørensen, J. B. (1999). STPIECE: Stata module to estimate piecewise-constant hazard
rate models [Software wrapper and help file]. Retrieved May 13, 2010, from
http://econpapers.repec.org/RePEc:boc:bocode:s396801
Sørensen, J. B. (2004). Recruitment-based competition between industries: A community
ecology. Industrial and Corporate Change, 13, 149-170.
Sorenson, O., & Audia, P. G. (2000). The social structure of entrepreneurial activity:
Geographic concentration of footwear production in the United States, 1940-
1989. American Journal of Sociology, 106, 424-461.
Soule, S. A., & King, B. G. (2008). Competition and resource partitioning in three social
movement industries. American Journal of Sociology, 113, 1568-1610.
Spence, A. M. (1974). Market signaling: Informational transfer in hiring and related
processes. Cambridge, MA: Harvard University Press.
277
Staggenborg, S. (1991). The pro-choice movement: Organization and activism in the
abortion conflict. New York: Oxford University Press.
Starrett, A. (1993). Adaptive resemblance: A unifying concept for mimicry and crypsis.
Biological Journal of the Linnean Society, 48(4), 299-317.
Steinberg, R. (2006). Economic theories of nonprofit organizations. In W. W. Powell &
R. Steinberg (Eds.), The nonprofit sector: A research handbook (2nd ed., pp. 117-
135). New Haven, CT: Yale University Press.
Steiner, M. A. (2006). The rhetoric of Operation Rescue: Projecting the Christian pro-
life message. New York: T & T Clark.
Sterelny, K. (2008). Language and niche construction. In D. K. Oller & U. Griebel (Eds.),
Evolution of communicative flexibility: Complexity, creativity, and adaptability in
human and animal communication (pp. 215-231). Cambridge, MA: MIT Press.
Stevens, M. L. (2002). The organizational vitality of conservative Protestantism.
Research in the Sociology of Organizations, 19, 337-360.
Stinchcombe, A. L. (1965). Social structure and organizations. In J. G. March (Ed.),
Handbook of organizations (pp. 142-193). Chicago, IL: Rand McNally.
Strang, D., & Macy, M. W. (2001). In search of excellence: Fads, success stories, and
adaptive emulation. American Journal of Sociology, 17, 147-182.
Strang, D., & Sine, W. D. (2002). Interorganizational institutions. In J. A. C. Baum (Ed.),
The Blackwell companion to organizations (pp. 497-519). Malden, MA:
Blackwell.
Stuart, T. E., & Sorenson, O. (2003). The geography of opportunity: Spatial
heterogeneity in founding rates and the performance of biotechnology firms.
Research Policy, 32, 229-253.
Suchman, M. C. (1995). Managing legitimacy: Strategic and institutional approaches.
Academy of Management Review, 20, 571-610.
278
Suchman, M. C., Steward, D. J., & Westfall, C. A. (2001). The legal environment of
entrepreneurship: Observations on the legitimation of venture finance in Silicon
Valley. In C. B. Schoonhoven & E. Romanelli (Eds.), The entrepreneurship
dynamic: Origins of entrepreneurship and the evolution of industries (pp. 349-
382). Stanford, CA: Stanford University Press.
Summerhill, L. (1984). The story of Birthright: The alternative to abortion (2nd ed.).
Kenosha, WI: Prow Books.
Sundie, J. M., Cialdini, R. B., Griskevicius, V., & Kenrick, D. T. (2006). Evolutionary
social influence. In M. Schaller, J. A. Simpson & D. T. Kenrick (Eds.), Evolution
and social psychology (pp. 288-316). New York: Psychology Press.
Swaminathan, A. (1995). The proliferation of specialist organizations in the American
wine industry, 1941-1990. Administrative Science Quarterly, 40, 653-680.
Swaminathan, A. (1996). Environmental conditions at founding and organizational
mortality: A trial-by-fire model. Academy of Management Journal, 39, 1350-
1377.
Swaminathan, A. (1998). Entry into new market segments in mature industries:
Endogenous and exogenous segmentation in the U.S. brewing industry. Strategic
Management Journal, 19, 389-404.
Swaminathan, A. (2001). Resource partitioning and the evolution of specialist
organizations: The role of location and identity in the U.S. wine industry.
Academy of Management Journal, 44, 1169-1185.
Thijs, A. K. L. (2002). Perceptions of deceit and innovation in the Antwerp textile
industry (Sixteenth and Seventeenth Centuries). In T. van Houdt, J. L. de Jong, Z.
Kwak, M. Spies & M. van Vaeck (Eds.), On the edge of truth and honesty:
Principles and strategies of fraud and deceit in the early modern period (pp. 127-
147). Boston, MA: Brill.
Thornhill, S., & Amit, R. (2003). Learning about failure: Bankruptcy, firm age, and the
resource-based view. Organization Science, 14, 497-509.
279
Thornton, P. H. (2004). Markets from culture: Institutional logics and organizational
decisions in higher education publishing. Stanford, CA: Stanford Business Books.
Tilly, C. (2004). Social boundary mechanisms. Philosophy of the Social Sciences, 34,
211-236.
Tilly, C. (2006). Identities, boundaries, and social ties. Boulder, CO: Paradigm.
Tucker, D. J. (1994). Progress and problems in population ecoogy. In J. A. C. Baum & J.
V. Singh (Eds.), Evolutionary dynamics of organizations (pp. 327-333). New
York: Oxford University Press.
Tucker, D. J., Singh, J. V., Meinhard, A. G., & House, R. J. (1988). Ecological and
institutional sources of change in organizational populations. In G. R. Carroll
(Ed.), Ecological models of organizations (pp. 127-151). Cambridge, MA:
Ballinger.
Turner, J. R. G. (1988). The evolution of mimicry: A solution to the problem of
punctuated equilibrium. American Naturalist, 131(Suppl.), 42-66.
Twombly, E. C. (2003). What factors affect the entry and exit of nonprofit human service
organizations in metropolitan areas? Nonprofit and Voluntary Sector Quarterly,
32, 211-235.
Usher, J. M., & Evans, M. G. (1996). Life and death along gasoline alley: Darwinian and
Lamarckian process in a differentiating population. Academy of Management
Journal, 39, 1428-1466.
Van de Ven, A. H., & Grazman, D. N. (1999). Evolution in a nested hierarchy: A
genealogy of Twin Cities health care organizations, 1853-1995. In J. A. C. Baum
& B. McKelvey (Eds.), Variations in organization science: In honor of Donald T.
Campbell (pp. 185-212). New York: Oxford University Press.
Vanderwoerd, J. R. (2004). How faith-based social service organizations manage secular
pressures associated with government funding. Nonprofit Management and
Leadership, 14, 239-262.
280
Ventresca, M. J., & Mohr, J. W. (2002). Archival research methods. In J. A. C. Baum
(Ed.), The Blackwell companion to organizations (pp. 805-828). Malden, MA:
Blackwell.
Victor, B., & Stephens, C. (1994). The dark side of the new organizational forms: An
editorial essay. Organization Science, 5, 479-482.
Walker, E. T., & McCarthy, J. D. (2010). Legitimacy, strategy, and resources in the
survival of community-based organizations. Social Problems, 57, 315-340.
Walker, J. (1991). Mobilizing interest groups in America. Ann Arbor, MI: University of
Michigan Press.
Weber, K., Heinze, K. L., & DeSoucey, M. (2008). Forage for thought: Mobilizing codes
in the movement for grass-fed meat and dairy products. Administrative Science
Quarterly, 53, 529-567.
Weber, M. S. (2010). Media reinvented: The transformation of news in a networked
society. Unpublished dissertation, University of Southern California, Los Angeles.
Weinberg, S. L., & Abramowitz, S. K. (2002). Data analysis for the behavioral sciences
using SPSS. New York: Cambridge University Press.
Weitzman, M. L. (1992). On diversity. Quarterly Journal of Economics, 107, 363-405.
Weyer, J. (1989). "Reden über Technik" als Strategie sozialer Innovation: Zur Genese
und Dynamik von Technik am Beispiel der Raumfahrt in der Bundesrepublik
["Talking about technology" as a strategy for social innovation: On the origins
and dynamics of technology as illustrated by aerospace in Germany]. In M.
Glagow, H. Willke & H. Wiesenthal (Eds.), Gesellschaftliche
Steuerungsrationalität und partikulare Handlungsstrategien (pp. 81-114).
Pfaffenweiler, Germany: Centaurus-Verlagsgesellschaft.
Wickler, W. (1968). Mimicry in plants and animals (R. D. Martin, Trans.). New York:
McGraw-Hill.
281
Wilson, P. A. (2003). Faith-based organizations, charitable choice, and government.
Administration and Society, 35, 29-51.
Witcher, P. (2008, November 3). The bright, white bite. Savannah Daily News. Retrieved
May 5, 2010, from http://savdailynews.com/main.asp?SectionID=2014&
SubSectionID=2100&ArticleID=20363
Witt, U. (2000). Changing cognitive frames – changing organizational forms: An
entrepreneurial theory of organizational development. Industrial and Corporate
Change, 9, 733-755.
Wolf-Devine, C., & Devine, P. E. (2009). Abortion: A communitarian pro-life
perspective. In M. Tooley, C. Wolf-Devine, P. E. Devine & A. M. Jaggar (Eds.),
Abortion: Three perspectives (pp. 65-119). New York: Oxford University Press.
Yamaguchi, K. (1991). Event history analysis. Newbury Park, CA: Sage.
Yates, J. A., & Orlikowski, W. J. (2002). Genre systems: Structuring interaction through
communicative norms. Journal of Business Communication, 39, 13-35.
Yoshioka, T., Hermann, G., Yates, J. A., & Orlikowski, W. J. (2001). Genre taxonomy: A
knowledge repository of communicative actions. ACM Transactions on
Information Systems, 19, 431-456.
Young, R. C. (1988). Is population ecology a useful paradigm for the study of
organizations? American Journal of Sociology, 94, 1-24.
Zelizer, V. A. (1978). Human values and the market: the case of life insurance and death
in 19th-Century America. American Journal of Sociology, 84, 591-610.
Zelizer, V. A. (1981). The price and value of children: The case of children's insurance.
American Journal of Sociology, 86, 1036-1056.
Zuckerman, E. W. (1999). The categorical imperative: Securities analysts and the
illegitimacy discount. American Journal of Sociology, 104, 1398-1438.
282
Zuckerman, E. W., & Kim, T.-Y. (2003). The critical trade-off: Identity assignment and
box-office success in the feature film industry. Industrial and Corporate Change,
12, 27-67.
Zuckoff, M. (2005). Ponzi's scheme: The true story of a financial legend. New York:
Random House.
Zwerling, E. (2008, Fall). Dangerous masquerade. Ms. Magazine, 36, 50-54.
283
APPENDICES
Appendix A: CPCs Included in the Analysis Grouped by NTEE Major Group
NTEE major group label
NTEE major
group code(s)
Number of CPCs
listed
Percentage of
total number
Arts, culture, and humanities A 2 0.10
%
Education B 12 0.58 %
Environment and animals C 1 0.05 %
Health E, F, G 211 10.19 %
Human services I, J, L, M, O, P 1,778 85.89 %
International, foreign affairs Q 1 0.05 %
Public, societal benefit R, S, T, W 43 2.08 %
Religion related X 19 0.92 %
Unknown, unclassified Z 3 0.14 %
Total number of CPCs: 2,070 100.00 %
Note: The labels for NTEE major groups are provided at http://nccs.urban.org/classification/NTEE.cfm
284
Appendix B: Search Terms for Newspaper Article Research
Search description LexisNexis Boolean search terms
Search 1
Distinctly labeled CPCs
without clinic capabilities
pregnan! w/3 (program OR center) AND (alternatives to
abortion OR abortion alternatives OR Birthright OR
Heartbeat OR Sav-a-life OR ministr!) AND NOT
(ultrasound OR sonogram OR clinic)
Search 2
Crisis form CPCs without
clinic capabilities
crisis pregnancy center OR emergency pregnancy center
OR emergency pregnancy servic! OR pregnancy crisis
center OR pregnancy problem center AND NOT
(ultrasound OR sonogram OR clinic)
Search 3
Neutral form CPCs without
clinic capabilities
pregnan! AND (family resource center OR pregnancy aid
OR pregnancy care center OR pregnancy center OR
pregnancy counseling center OR pregnancy help center OR
pregnancy information center OR pregnancy resource
center OR pregnancy servic! OR pregnancy support center
OR pregnancy test! center OR women's care OR women's
center OR women's resource center OR women's servic!
OR CareNet or Care Net) AND NOT (crisis pregnancy
center OR emergency pregnancy center OR emergency
pregnancy servic! OR pregnancy crisis center OR
pregnancy problem center OR ultrasound OR sonogram OR
clinic)
Search 4
Neutral form CPCs with
clinic capabilities
pregnan! AND (pregnancy care clinic OR pregnancy clinic
OR pregnancy help clinic OR pregnancy help medical
clinic OR pregnancy crisis clinic OR pregnancy resource
medical center OR pregnancy support clinic OR women's
health center OR women's clinic OR women's medical
center)
Search 5
CPCs with clinic
capabilities
pregnan! AND (family resource center OR pregnancy aid
OR pregnancy care center OR pregnancy center OR
pregnancy counseling center OR pregnancy help center OR
pregnancy information center OR pregnancy resource
center OR pregnancy servic! OR pregnancy support center
OR pregnancy test! center OR women's care OR women's
center OR women's resource center OR women's servic!)
AND (ultrasound OR sonogram OR clinic)
285
Appendix C: Generalization Thesaurus Used for Textual Analysis
Original string Converted string Original string Converted string
isn t is not catholic church catholic_church
aren t are not baptist church baptist_church
wasn t was not episcopal church episcopal_church
weren t were not evangelical church evangelical_church
haven t have not christian church christian_church
hasn t has not lutheran church lutheran_church
hadn t had not presbyterian church presbyterian_church
won t will not congregation congregation
wouldn t would not congregational congregational
don t do not diocese diocese
doesn t does not parish parish
didn t did not ministries ministry
can t cannot ministry ministry
couldn t could not church church
shouldn t should not religious group religious_group
mightn t might not christian counseling christian_counseling
mustn t must not spiritual counseling spiritual_counseling
alternative abortion abortion_alternative pray pray
no abortion refer not_refer_abortion prayer prayer
no abortion referral not_refer_abortion praying praying
no refer abortion not_refer_abortion worship worship
no referral abortion not_refer_abortion catholic catholic
not abortion refer not_refer_abortion baptist baptist
not abortion referral not_refer_abortion episcopal episcopal
not refer abortion not_refer_abortion evangelical evangelical
not referrals abortion not_refer_abortion christian christian
care net care_net christianity christianity
carenet care_net lutheran lutheran
birth right birthright presbyterian presbyterian
birthline birthline minister minister
birthright birthright missionaries missionary
guttmacher guttmacher missionary missionary
guttmacher institute guttmacher pastor pastor
heartbeat international heartbeat_international chapel chapel
heartbeat heartbeat christ christ
operation rescue operation_rescue faith faith
naral naral god god
national right to life national_right_to_life grace grace
sav life sav_a_life religious religious
relay life relay_life bible bible
catholic social service catholic_social_service biblical biblical
286
Appendix C (Continued)
Original string Converted string Original string Converted string
abortion opponent abortion_opponent not abortion clinic not_abortion_clinic
abortion opposition oppose_abortion no abortion clinic no_abortion_clinic
opponent abortion oppose_abortion abortion clinic abortion_clinic
oppose abortion oppose_abortion adoption agency adoption_agency
opposed abortion oppose_abortion
community health
center
community_health_center
opposing abortion oppose_abortion
community pregnancy
center
community_pregnancy_center
opposition abortion oppose_abortion
community pregnancy
clinic
community_pregnancy_clinic
abortion foe abortion_foe crisis pregnancy center crisis_pregnancy_center
abortion protester abortion_protester fake clinic fake_clinic
anti abortion anti_abortion family care clinic family_care_clinic
abstinence
counseling
abstinence_counseling family resource center family_resource_center
abstinence
education
abstinence_education maternity home maternity_home
abstinence only
education
abstinence_only_education medical center medical_center
abstinence only
training
abstinence_only_training medical clinic medical_clinic
abstinence program abstinence_program medical facility medical_facility
abstinence training abstinence_training medical practice medical_practice
promote abstinence promote_abstinence pregnancy care center pregnancy_care_center
promoting
abstinence
promote_abstinence pregnancy care clinic pregnancy_care_clinic
abstinence abstinence pregnancy center pregnancy_center
chastity counseling chastity_counseling pregnancy clinic pregnancy_clinic
purity counseling purity_counseling
pregnancy counseling
center
pregnancy_counseling_center
not provide
contraceptive
not_provide_contraceptive pregnancy crisis center pregnancy_crisis_center
no contraceptive no_contraceptive pregnancy health clinic pregnancy_health_clinic
no birth control no_birth_control pregnancy health center pregnancy_health_center
birth control birth_control pregnancy help center pregnancy_help_center
condom condom pregnancy help clinic pregnancy_help_clinic
contraception contraception
pregnancy help medical
clinic
pregnancy_help_medical_clinic
contraceptive contraceptive
pregnancy information
center
pregnancy_information_center
fake abortion clinic fake_abortion_clinic
pregnancy resource
center
pregnancy_resource_center
287
Appendix C (Continued)
Original string Converted string Original string Converted string
pregnancy support
center
pregnancy_support_center gynecologist ob_gyn
pregnancy test center pregnancy_test_center physician medical_professional
pregnancy testing
center
pregnancy_test_center surgeon medical_professional
woman center woman_center healthcare provider healthcare_provider
women center women_center health care provider healthcare_provider
women clinic women_clinic options counselor options_counselor
women health center women_health_center
offer free pregnancy
test
offer_free_pregnancy_test
women health clinic women_health_clinic
offer free pregnancy
testing
offer_free_pregnancy_test
women medical center women_medical_center offer free pregnancy offer_free_pregnancy
women medical clinic women_medical_clinic free pregnancy test free_pregnancy_test
women resource
center
women_resource_center ultrasound ultrasound
health center health_center sonogram sonogram
health clinic health_clinic
pregnancy
counseling
pregnancy_counseling
wrc women_resource_center options counseling options_counseling
pcc pregnancy_crisis_center
post abortion
counseling
post_abortion_counseling
phc pregnancy_help_clinic cancer screening cancer_screening
phmc pregnancy_help_medical_clinic pap smear pap_smear
prc pregnancy_resource_center pap test pap_smear
cpc crisis_pregnancy_center pelvic exam pelvic_exam
no abortion provider no_abortion_provider physical exam physical_exam
not abortion provider no_abortion_provider
sexually transmitted
disease
std
abortion provider abortion_provider std test std_test
doctor medical_professional treat std treat_std
health professional health_professional std treatment treat_std
licensed nurse licensed_nurse hiv test hiv_test
licensed physician licensed_physician surgery surgery
medical practitioner medical_practitioner surgical surgical
medical professional medical_professional adoption referral adoption_referral
medical staff medical_staff refer adoption adoption_referral
nurse practitioner medical_professional referral adoption adoption_referral
registered nurse registered_nurse abortion activist abortion_activist
nurse medical_professional abortion advocate abortion_advocate
obgyn ob_gyn abortion counseling abortion_counseling
obstetrician medical_professional abortion minded abortion_minded
288
Appendix C (Continued)
Original string Converted string Original string Converted string
abortion movement abortion_movement patient patient
abortion right abortion_rights peer counsel peer_counsel
abortion right advocate abortion_rights_advocate phone hotline phone_hotline
abortion vulnerable abortion_vulnerable planned parenthood planned_parenthood
bogus clinic bogus_clinic pregnancy crisis pregnancy_crisis
charitable charitable pro abortion pro_abortion
charity charity pro choice pro_choice
choice women choice_women pro life pro_life
choose abortion choose_abortion promote abortion promote_abortion
choose life choose_life reproductive choice reproductive_choice
consultation consultation reproductive freedom reproductive_freedom
crisis pregnancies crisis_pregnancies reproductive health reproductive_health
crisis pregnancy crisis_pregnancies reproductive medicine reproductive_medicine
donate donate reproductive right reproductive_right
donation donation social service social_service
false information false_information specialty license plate specialty_license_plate
misleading information misleading_information specialty plate specialty_plate
not accurate information inaccurate_information unborn baby unborn_baby
inaccurate information inaccurate_information unborn child unborn_child
family planning family_planning unborn unborn
feminist feminist unplanned pregnancies unplanned_pregnancies
focus family focus_on_the_family unplanned unplanned
free confidential free_confidential unwed unwed
confidential confidential unmarried unwed
free maternity clothes free_maternity_clothes volunteer volunteer
fundraiser fundraiser volunteer counselor volunteer_counselor
fundraising fundraising walk life walk_for_life
human life human_life walk in walk_in
look like look_like walk in appointment walk_in_appointment
maternity care maternity_care abortion procedure abortion_procedure
maternity housing maternity_housing abortion service abortion_service
mislead mislead exam exam
motherhood motherhood examination examination
nonprofit nonprofit fertility fertility
non profit nonprofit free service free_service
offer abortion offer_abortion gyn gyn
parent class parent_class gynecological gynecological
parent education parent_education gynecology gynecology
parent skill parent_skill health care service health_care_service
289
Appendix C (Continued)
Original string Converted string Original string Converted string
health service health_service obstetrics and gynecology ob_gyn
health care health_care partial birth abortion partial_birth_abortion
healthcare healthcare not perform abortion not_perform_abortion
woman concern woman_concern perform abortion perform_abortion
women care women_care pregnancy aid pregnancy_aid
women concern women_concern pregnancy care pregnancy_care
women service women_service pregnancy service pregnancy_service
mammography mammography prenatal care referral prenatal_care_referral
medical care medical_care prenatal care prenatal_care
medical information medical_information provide abortion provide_abortion
medical procedure medical_procedure vaginal exam vaginal_exam
medical service medical_service womb womb
medical treatment medical_treatment adoption adoption
not provide abortion not_provide_abortion adopt adopt
obstetrical obstetrical choice choice
obstetric obstetric clinic clinic
290
Appendix D: Thesaurus Translating Concepts and Phrases Into Six Meta Concepts
Distinct elements Mimicry elements
Feature-based
abstinence_counseling, abstinence_education,
abstinence_only_education,
abstinence_only_training, abstinence_program,
abstinence_training, adopt, adoption,
adoption_referral, chastity_counseling,
christian_counseling, donate, donation,
false_information, free_maternity_clothes,
fundraiser, fundraising, inaccurate_information,
maternity_care, maternity_housing, mislead,
misleading_information, no_birth_control,
no_contraceptive, not_perform_abortion,
not_provide_abortion, not_provide_contraceptive,
not_refer_abortion, parent_class, parent_education,
parent_skill, post_abortion_counseling,
promote_abstinence, purity_counseling,
spiritual_counseling, womb
abortion_counseling, abortion_procedure,
abortion_service, birth_control, cancer_screening,
condom, confidential, consultation, contraception,
contraceptive, exam, examination, fertility,
free_confidential, free_pregnancy_test,
gynecological, gynecology, hiv_test, mammography,
medical_care, medical_information,
medical_procedure, medical_service,
medical_treatment, ob_gyn, obstetric, obstetrical,
offer_abortion, offer_free_pregnancy_test,
options_counseling, pap_smear, pelvic_exam,
perform_abortion, physical_exam,
pregnancy_counseling, prenatal_care,
prenatal_care_referral, provide_abortion, sonogram,
std_test, surgery, surgical, treat_std, ultrasound,
vaginal_exam, walk_in, walk_in_appointment
Identity-based
abortion_alternative, abortion_movement,
abortion_rights, anti_abortion, baptist, bible,
biblical, birthline, birthright, bogus_clinic,
care_net, catholic, charitable, charity, choose_life,
christ, christian, christianity, crisis_pregnancies,
crisis_pregnancy_center, episcopal,
emergency_pregnancy_center, evangelical, faith,
fake_abortion_clinic, fake_clinic, god, grace,
heartbeat, heartbeat_international, human_life,
look_like, lutheran, maternity_home, motherhood,
no_abortion_clinic, no_abortion_provider,
nonprofit, not_abortion_clinic, oppose_abortion,
pray, prayer, pregnancy_crisis,
pregnancy_crisis_center, presbyterian,
problem_pregnancy_center, pro_life, religious,
sav_a_life, worship
abortion_clinic, abortion_provider, choice,
choose_abortion, clinic, community_health_center,
community_pregnancy_center,
community_pregnancy_clinic, family_care_clinic,
family_planning, family_resource_center, feminist,
health_care, health_care_service, health_center,
health_clinic, health_service, healthcare_provider,
medical_center, medical_clinic, medical_facility,
medical_practice, pregnancy_aid, pregnancy_care,
pregnancy_care_center, pregnancy_care_clinic,
pregnancy_center, pregnancy_clinic,
pregnancy_counseling_center,
pregnancy_health_center, pregnancy_health_clinic,
pregnancy_help_center, pregnancy_help_clinic,
pregnancy_help_medical_clinic,
pregnancy_information_center,
pregnancy_resource_center, pregnancy_service,
pregnancy_support_center, pregnancy_test_center,
pro_abortion, pro_choice, promote_abortion,
reproductive_choice, reproductive_freedom,
reproductive_health, reproductive_medicine,
reproductive_right, social_service, woman_center,
woman_concern, women_care, women_clinic,
women_concern, women_health_center,
women_health_clinic, women_medical_center,
women_medical_clinic, women_resource_center
291
Appendix D (Continued)
Distinct elements Mimicry elements
Structure-based
abortion_foe, abortion_minded, abortion_opponent,
abortion_protester, abortion_rights_advocate,
abortion_vulnerable, adoption_agency,
baptist_church, catholic_church,
catholic_social_service, chapel, christian_church,
church, congregation, congregational, diocese,
episcopal_church, evangelical_church,
focus_on_the_family, lutheran_church, minister,
ministry, missionary, national_right_to_life,
operation_rescue, parish, partial_birth_abortion,
pastor, peer_counsel, phone_hotline,
presbyterian_church, religious_group,
specialty_license_plate, specialty_plate, unborn,
unborn_baby, unborn_child, unplanned,
unplanned_pregnancies, unwed, volunteer,
volunteer_counselor, walk_for_life
abortion_activist, abortion_advocate, guttmacher,
guttmacher, health_professional, licensed_nurse,
licensed_physician, medical_practitioner,
medical_professional, medical_staff, naral, ob_gyn,
ob_gyn, options_counselor, patient,
planned_parenthood, registered_nurse
Abstract (if available)
Abstract
Communication scholarship has significantly improved organizational ecology models by emphasizing that the members of organizational communities are participating actively in the enactment of their environment. Organizational forms are viewed as contested and multi-faceted action frames which emerge from community interactions and simultaneously shape them. Form negotiations mediate the evolutionary fate of population members adhering to a form. Existing institutional ecology approaches focus on discursive strategies foregrounding the distinctive and unique organizational identity underlying a certain form. The assumption prevails that organizational populations can only thrive if their form is recognized clearly by diverse audiences as this entails that it has achieved cognitive legitimacy. But not all organizational populations benefit from clearly communicating who they are, particularly if they are operating in an environment that attaches low cognitive legitimacy to their form. The biological concept of aggressive mimicry, which refers to a competitive evolutionary strategy based on form ambiguity, illuminates the circumstances when members of a population exploit the lack of clarity about their form to their benefits. Among other tactics, aggressive mimics in organizational populations engage in purposive labeling of their names to increase their resemblance with a population of models.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Communicating organizational knowledge in a sociomaterial network: the influences of communication load, legitimacy, and credibility on health care best-practice communication
PDF
Effects of economic incentives on creative project-based networks: communication, collaboration and change in the American film industry, 1998-2010
PDF
Contagious: social norms about health in work group networks
PDF
The missing link: the role of organizational culture in technology change management communication
PDF
The role of social media for community-based organizations focused on environmental justice in southeast Los Angeles
PDF
A study of diversification In the outdoor recreation industry and its connection to the Black, Indigenous, and People of Color (BIPOC) urban health equity gap
Asset Metadata
Creator
Heiss, Bettina Maria Richards
(author)
Core Title
Organizational mimicry in American social movement communities: an analysis of form communication effects on the evolution of crisis pregnancy centers, 1989-2009
School
Annenberg School for Communication
Degree
Doctor of Philosophy
Degree Program
Communication
Publication Date
11/10/2010
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cognitive legitimacy,community ecology,Cox regression,event history analysis,evolution,institutional ecology,labeling,meta network analysis,mimicry,negative binomial regression,nonprofit organizations,OAI-PMH Harvest,organizational forms,organizational identities,piecewise constant exponential models,population ecology,pro-life,social movement organizations
Place Name
USA
(countries)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Monge, Peter R. (
committee chair
), Fulk, Janet (
committee member
), Kennedy, Mark T. (
committee member
)
Creator Email
bmrheiss@gmail.com,heiss@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m3519
Unique identifier
UC1142758
Identifier
etd-Heiss-4187 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-420936 (legacy record id),usctheses-m3519 (legacy record id)
Legacy Identifier
etd-Heiss-4187.pdf
Dmrecord
420936
Document Type
Dissertation
Rights
Heiss, Bettina Maria Richards
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
cognitive legitimacy
community ecology
Cox regression
event history analysis
evolution
institutional ecology
labeling
meta network analysis
mimicry
negative binomial regression
nonprofit organizations
organizational forms
organizational identities
piecewise constant exponential models
population ecology
pro-life
social movement organizations