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
/
The functions of the middleman: how intermediary nonprofit organizations support the sector and society
(USC Thesis Other)
The functions of the middleman: how intermediary nonprofit organizations support the sector and society
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THE FUNCTIONS OF THE MIDDLEMAN: HOW INTERMEDIARY NONPROFIT
ORGANIZATIONS SUPPORT THE SECTOR AND SOCIETY
by
Yusun Cho
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
(PUBLIC POLICY AND MANAGEMENT)
August 2018
Copyright 2018 Yusun Cho
i
ACKNOWLEDGEMENTS
First and foremost, I am extremely grateful to my advisor Professor Peter J. Robertson
for his valuable guidance and consistent support for me. I appreciate for all his contributions of
time and efforts for encouraging my research. Completion of this dissertation was possible only
because of his unconditional support. I would also like to thank my committee members,
Professor Shui Yan Tang and Professor Gary D. Painter for their professional advisement and
valuable comments which augmented my dissertation with various perspectives. I could never
have accomplished this work without their support. My sincere thanks also go to Professor
Jae-jin Yang who made it possible for me to pursue my academic career.
I thank my colleagues at USC and all my friends for providing constant support and
friendship for me. My special thanks go to Chunhong Chon and Kyungran Kim, who looked
after me with tremendous love and support.
I am deeply thankful to my entire family for their love and support. I would not have
made this work without them. I would like to thank to my father in law Seyoo Shim and mother
in law Kyungnan Yoo for their love, immense support, and encouragement. Special thanks to my
beloved husband, Han Suk Shim who helped me for overcoming difficulties with his abundant
love and encouragement.
I dedicate this dissertation to my mother Eunhee Park who has devoted her life to me.
Any words cannot express how grateful I am to her for all of the sacrifices that she has made on
my behalf. This dissertation is also dedicated to the memory of my beloved father Mu-yeon Cho
and my grandmother Kija Lee, who gave me incredible support and love.
Finally, I would like to thank everyone who helped and supported me. This journey
would not have been possible without the support of my family, professors, mentors, and friends.
ii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ............................................................................................................. i
LIST OF TABLES ......................................................................................................................... vi
LIST OF FIGURES ...................................................................................................................... vii
ABSTRACT ................................................................................................................................. viii
CHAPTER 1. INTRODUCTION ................................................................................................... 1
Background ................................................................................................................................. 2
Definition of Nonprofit Intermediaries .................................................................................... 2
Importance of Nonprofit Intermediaries .................................................................................. 4
Functions of Nonprofit Intermediaries .................................................................................... 5
Purpose of the Study ................................................................................................................... 7
CHAPTER 2. INTERMEDIARIES AS CAPACITY BUILDERS: SPECIALIZATION IN
NONPROFIT MANAGEMENT SUPPORT ORGANIZATIONS ................................ 9
Background ............................................................................................................................... 10
Nonprofit Capacity Building and Management Support Organizations ............................... 10
Specialization and Generalization in Nonprofit Management Support Organizations ......... 12
Organizational Ecology: Resource Partitioning, Density Dependence, and Market Capacity
............................................................................................................................................... 15
Research Hypotheses................................................................................................................. 18
Resource Partitioning between Generalist and Specialist MSOs .......................................... 18
Density Dependence: Competition among Specialist MSOs ................................................ 19
Market Capacity and Organizational Niche: Clients and Demand for Services ................... 19
iii
Data and Method ....................................................................................................................... 20
Data ........................................................................................................................................ 20
Empirical Models and Variables ........................................................................................... 23
Results ....................................................................................................................................... 25
Overview of Nonprofit MSOs ............................................................................................... 25
Logistic Regression Analysis of Founding of Specialist MSOs............................................ 26
Discussion and Conclusion ....................................................................................................... 30
CHAPTER 3. INTERMEDIARIES AS COORDINATORS OF POLICY IMPLEMENTATION:
EQUITY AND EFFECTIVENESS OF A JOB ASSISTANCE AND TRAINING
PROGRAM ................................................................................................................... 34
Background ............................................................................................................................... 35
Description of the Workforce Investment Act Program ........................................................ 35
Local Workforce Investment Boards: Intermediaries as Policy Implementing Coordinators38
Performance Evaluations of the Workforce Investment Act ................................................. 40
Performance Issues in the WIA Job Training Program............................................................. 42
Equity Issues .......................................................................................................................... 42
Effectiveness Issues ............................................................................................................... 44
Equity and Effectiveness in a Quasi-Governmental Form of LWIB ..................................... 46
Research Questions................................................................................................................ 48
Data and Method ....................................................................................................................... 49
Data ........................................................................................................................................ 49
Sample ................................................................................................................................... 51
Analysis and Variables .......................................................................................................... 51
iv
Results ....................................................................................................................................... 53
Descriptive Statistics ............................................................................................................. 53
Equity of the WIA Job Training Program ............................................................................. 56
Effectiveness of the WIA Job Training Programs ................................................................. 61
Discussion and Conclusion ....................................................................................................... 68
CHAPTER 4. INTERMEDIARIES AS POLITICAL NETWORK MOBILIZERS: THE ROLE
OF SOCIAL MEDIA IN PROMOTING CIVIC ENGAGEMENT WITH
NONPROFIT ADVOCACY ORGANIZATIONS ....................................................... 75
Background ............................................................................................................................... 76
Advocacy Nonprofit Organizations and the Use of Social Media ........................................ 76
Literature Review on Nonprofit Organizations and Social Media ........................................ 77
Research Questions ................................................................................................................... 79
Civic Engagement with Advocacy Organizations vial Social Media .................................... 79
Network Effects of Social Media .......................................................................................... 80
Twitter as an Effective Tool for Civic Engagement .............................................................. 83
Research Hypotheses................................................................................................................. 84
Resource Dependence Theory and the Use of Social Media ................................................. 84
Factors Affecting Civic Engagement on Social Media ......................................................... 85
Method ...................................................................................................................................... 87
Sample and Data Collection .................................................................................................. 87
Variables ................................................................................................................................ 89
Results ....................................................................................................................................... 92
Descriptive Statistics on Twitter Use of Nonprofit Advocacy Organizations ....................... 92
v
Factors Affecting Civic Engagement with Nonprofit Advocacy Organizations on Twitter . 93
Discussion and Conclusion ....................................................................................................... 99
CHAPTER 5. CONCLUSION.................................................................................................... 103
Summary of Findings .............................................................................................................. 103
Contributions and Implications for the Field .......................................................................... 106
Limitations of the Study and Future Research Directions ...................................................... 107
REFERENCES ........................................................................................................................... 110
vi
LIST OF TABLES
Table 2.1 The Average Number and Density of Nonprofit MSOs in 2013 .................................. 25
Table 2.2 Logistic Regression on the Factors Affecting Founding of New Specialist MSOs ...... 28
Table 3.1 Descriptive Statistics of the Sampled WIA Adult Program Participants ...................... 55
Table 3.2 Multivariate Logistic Regression of Receipt of Job-training Services on Client
Demographic Characteristics ........................................................................................ 57
Table 3.3 Differences between training services and core & intensive services recipients by the
organizational form of LWIBs ...................................................................................... 59
Table 3.4 Multivariate Logistic Regression of Receipt of Job-training Services on Client
Demographic Characteristics by the organizational form of LWIBs ........................... 60
Table 3.5 Estimated Effects of the Job-training Program Compared with the Core or Intensive
Services ......................................................................................................................... 62
Table 3.6 USDOL’s Performance Evaluation of LWIBs in 2015 ................................................ 63
Table 3.7 Logistic regression of the effects of training program and LWIB organizational status
on employment rate and retention rate .......................................................................... 66
Table 3.8 Multivariate regression of the effects of training program and LWIB organizational
status LWIBs on earnings and earnings changes .......................................................... 67
Table 4.1 Twitter Presence of the 73 Human and Civil Rights Organizations ............................. 88
Table 4.2 Descriptive Statistics of the Sample ............................................................................. 92
Table 4.3 Correlations of the Explanatory Variables .................................................................... 94
Table 4.4 Twitter Activity and Engagement across Advocacy Topic Areas ................................ 95
Table 4.5 Factors Affecting Network-based Engagement on Twitter .......................................... 97
Table 4.6 Contrast Analyses of the Patterns of Engagement on Twitter across Advocacy Issue
Areas ............................................................................................................................. 99
vii
LIST OF FIGURES
Figure 2.1 The Number of Generalist and Specialist Nonprofit MSOs from 1990 to 2013 ......... 13
Figure 2.2 The Number of Founding and Disbanding of Generalist and Specialist MSOs from
1990 to 2013 ................................................................................................................. 14
Figure 3.1 Distribution of Governmental and Nonprofit LWIBs by States .................................. 54
Figure 4.1 Types of Information Diffusion Structure ................................................................... 81
viii
ABSTRACT
The purpose of this dissertation is to develop a better understanding of intermediary
nonprofit organizations and their roles in supporting the sector and society. Intermediary
organizations are middlemen that connect various social actors and help them collaborate in an
effective and efficient manner in order to solve intense social, political, and economic problems
in our society. Intermediaries are especially significant in the nonprofit sector where many
organizations have limited organizational capacity and resources. Despite their growing
importance, nonprofit intermediaries remain invisible in the literature. Filling that gap, this
dissertation consists of three essays that cover nonprofit intermediaries’ three main functions:
capacity building, coordinating policy implementing networks, and mobilizing political
networks.
The first study aims to investigate the growth and evolution of nonprofit management
support organizations and analyze the factors affecting their diversification. Nonprofit
management support organizations provide capacity-building services for other nonprofits and
play a contributing role in improving the sector’s performance. By analyzing panel data on the
population of these organizations from 1990 to 2013, this research found the number of nonprofit
management support organizations has grown dramatically during the last 20 years, as the
demands for capacity building services have increased. This research determined that the
organizational population of nonprofit management support organizations has evolved from
sector-wide management organizations assisting a variety of nonprofits in multiple domains to
management support organizations serving nonprofits operating in a specific service area. The
empirical analysis suggests that high competition between nonprofit management organizations
hinders their further growth.
ix
The second study explores the function of nonprofit intermediaries as coordinators of
collaborative policy implementing networks. In particular, this research investigates local
workforce investment boards, which coordinate US job assistance programs in local areas, and
the efficacy and equity of their varied organizational structures. Specifically, these boards exist
as two types: government-controlled nonprofit organizations and governmental agencies. By
analyzing the client data of the US Workforce Investment Act Adult program in 17 states, this
research found that the organizational form of local workforce investment boards is significantly
associated with program outcomes. The empirical analysis indicates that nonprofit workforce
investment boards were more effective in increasing their clients’ employment but less effective
in growing their earnings. This implies that nonprofit boards may focus on the quantity, rather
than the quality, of job placements, demonstrating the adverse effects of using non-governmental
agencies to deliver public services.
The third study investigates nonprofit advocacy organizations, which serve as political
network mobilizers, and the effectiveness of their social media use for mobilizing civic
engagement. Looking at how 70 advocacy intermediaries use Twitter, this research found that
advocacy intermediaries benefit from using social media; their organizational information and
activities are diffused to a large audience through users’ social networks, and it is read and shared
not only by their direct stakeholders but also friends of the stakeholders. In addition, this
research found that the patterns of civic engagement with organizations on Twitter differ by
advocacy topic areas. The findings indicate that civil rights nonprofit advocacy organizations
have higher levels of civic engagement than other organizations. This implies that the public is
more likely to engage in advocacy issues covering more general topics, and those issues are more
likely to be diffused, while social media activities for special interest groups are more likely to be
x
shared only by members of those specific interest groups.
By investigating three main functions of intermediary nonprofit organizations, this
dissertation provides evidence that nonprofit intermediaries perform a significant role in
strengthening the sector and society by performing a wide range of mediating roles, including
brokers, networkers, bridges, communicators, and resource mobilizers. This dissertation is
expected to contribute to the field of nonprofit and organizational studies by offering a
comprehensive understanding of nonprofit intermediaries and highlighting their importance,
which has received little attention in the literature.
1
CHAPTER 1. INTRODUCTION
An intermediary is defined as “any person or organization whose function places it
between two other persons or organizations” (Szanton, 2004, p.10). Similarly, intermediary
organizations are in-between organizations that act as agents or brokers in many aspects of
organizational processes between two or more parties. They add distinct value beyond what the
parties would be able to generate by themselves (Honig, 2004). These organizations are
especially important in fields with high environmental uncertainty and limited operational
capacity and resources. Within these domains, intermediaries are essential in helping
organizations work efficiently to achieve their mission and goals (de Souza Briggs, 2003).
Likewise, intermediary organizations serve a significant role in the nonprofit sector, where many
organizations have limited organizational capacity and resources (Brown & Kalegonkar, 2002).
Furthermore, increasing demands for collaborative relationships between organizations highlight
the importance of intermediaries as brokering organizations. Intermediary nonprofit
organizations not only strengthen the nonprofit sector and but also society at large by performing
their crucial functions, such as providing technical and management assistance, governing
resources, and managing relationships among various social actors (de Souza Briggs, 2003;
Shea, 2011).
Despite their growing importance, nonprofit intermediaries still remain invisible to
schlars, and we have little understanding of them. Therefore, this dissertation aims to investigate
the functions of nonprofit organizations and to analyze how these functions enhance not only the
nonprofit sector but also society. Three dissertation chapters examine intermediary organizations
as: 1) capacity builders, 2) coordinators of collaborative policy implementation networks, and 3)
political network mobilizers.
2
Background
Definition of Nonprofit Intermediaries
“Intermediaries” is a term widely used in the literature to describe organizations
operating between other actors, but there is no consensus as to what intermediaries are (Moss,
2009; Shea, 2011). Generally, “intermediary organizations are defined by their structural
position, namely ‘intermediary’ is any organization that mediates the relationship(s) between two
or more social actors” (Van der Meulen, et al., 2005, p. 3). However, the most distinctive feature
of intermediary organizations is that “intermediation is their raison d'etre”
1
(Van der Meulen,
Nedeva, & Braun, 2005, p.4). Based on this main feature, Honig (2004, p. 67) defines
intermediary organizations as “organizations that occupy the space in between at least two
parties. Intermediary organizations primarily function to mediate or to manage change in both
those parties.”
Researchers have used the term ‘intermediary organizations’ to describe a wide range of
organizations across different domains (Moss, 2009), including ‘funding intermediaries’
gathering funds from donors and regranting them to a defined set of grantees (Benjamin, 2010);
‘community-based intermediary organizations’ connecting various types of community members
for enhancing local community (Shea, 2011); ‘capacity-building intermediaries’ providing
technical assistance for improving organizational capacity (Szanton, 2004); ‘civic intermediaries’
linking citizens to the larger political structure (LeRoux, 2007); ‘public-private intermediary
organizations,’ coordinating social service provision governance (Mosely, 2014); ‘policy-
implementing intermediaries’ operating between policymakers and implementers (Honig, 2004);
1
“Raison d'etre” means “The claimed reason for the existence of something or someone; the sole or ultimate
purpose of something or someone.” (https://en.wiktionary.org/wiki/raison_d%27%C3%AAtre)
3
and ‘innovation intermediaries’ filling gaps in information, knowledge, and innovation (Howell,
2006; Winch, 2007). These studies were mostly descriptive, and theoretical discussions about
intermediary organizations are limited (Van der Meulen, et al., 2005).
In the nonprofit literature, de Souza Briggs (2003) broadly defines intermediary
nonprofit organizations as go-between organizations helping other nonprofits to be more
effective. He notes that nonprofit intermediaries primarily address various social problems in the
nonprofit sector, such as conflicts among stakeholders and organizations and limited capacity in
organizational performance and accountability. The role of the nonprofit sector has become
larger and stronger in the United States, as nonprofit organizations complement the role of for-
profit organizations and governments (Grønbjerg & Paarlberg, 2001; Hansmann, 1987;
Weisbrod, 1998). Thus, the roles of nonprofit intermediaries in strengthening the sector and
society also have become significant.
Nonprofit intermediaries aim to enhance the capacities of organizations and
communities of which they are a part (Shea, 2011). Communities are “networks of individuals,
groups, and/or organizations connected by similar beliefs, identities, geographies, cultures, or a
host of other variables” (Shea, 2011, p.59). Thus, strengthening community entails enhancing
networks of social members and bridging gaps, such as fragmentation of segments of the
community, resource mismatches, and cultural differences. By bridging various types of social
actors, intermediary organizations aim to increase information, resource-sharing, and interaction.
Nonprofit intermediaries perform a wide range of mediating roles, including broker, networker,
bridge, communicator, and resource mobilizer, and they work as a body of knowledge, skills, and
resources (Benjamin, 2010; Lopez, Kreider, & Coffman, 2005; Shea, 2011). However, the roles
of intermediaries are not widely known due to a dearth of research on that topic (Shea, 2011).
4
Importance of Nonprofit Intermediaries
Organizations are embedded in the environment in which they operate. As a result,
dealing with environmental complexity—such as governing resources, managing uncertainty,
and building relationships with multiple stakeholders—is the issue that organizations must
address in order to enhance organizational viability, efficiency, and effectiveness (Baum &
Rowley, 2002; Scott, 2004; Sternlieb, Bixler, & Huber-Stearns, 2013). The problem is that no
single organization has complete control over all the necessary components for addressing
environmental complexity and uncertainty. Therefore, organizations should interact with other
elements in their environments in order to secure necessary resources for their survival (Davis &
Cobb, 2010; Pfeffer & Salancik, 2003).
However, when interacting and working with others, organizations must consider, and
minimize, transaction costs, which are the costs associated with negotiating, coordinating, and
monitoring their partner organizations (Brown and Potoski, 2003; Williamson, 1981). Additional
costs, including time and money for processing information and protecting contracts, are
inevitably involved in collaborating with others, hindering efficiency. Intermediaries help other
organizations reduce these transaction costs by 1) searching for and evaluating information to
identify suitable matching organizations, 2) preparing organizations for contracts and exchanges
by enhancing their capacity, 3) building and mending relationships among organizations, and 4)
providing a credible environment for building trust among social members (Benjamin, 2010).
Ultimately, nonprofit intermediaries can facilitate efficient and effective collaboration between
social members. Other organizations may overcome environmental complexity and uncertainty.
5
Functions of Nonprofit Intermediaries
The definitions of intermediaries vary greatly across areas where they are used, so there
is no clear consensus about the way to identify intermediary nonprofit organizations. However,
intermediary organizations can be distinguished by their functions (Moss, 2009). Intermediaries
facilitate connections and interactions between actors with diverse backgrounds, and they play a
significant role in producing knowledge, resources, and relationship-building (Sternlieb, et al.,
2013). Many researchers have discussed the various functions of intermediary organizations in
the nonprofit sector (Benjamin, 2010; de Souza Briggs, 2003; Lopez, et al., 2005; Shea, 2011;
Smith, 2007; Szanton, 2004; Wynn, 2000); based on a review of this literature, the functions of
nonprofit intermediaries can be broadly classified into three roles: 1) capacity builders, 2)
coordinators of policy implementing networks, and 3) political network mobilizers.
First, nonprofit intermediaries function as capacity builders of the nonprofit sector.
Nonprofit organizations are often challenged by a lack of managerial expertise for achieving
their organizational missions and goals (Wimberley & Rubens, 2002). In order to help these
organizations, nonprofit intermediaries provide management and technical assistance services,
including staff development, leadership skills, board governance, strategic planning, and
financial management (Brown & Kalegonkar, 2002; Da Vita & Fleming, 2001; Renz, 2008); they
ultimately aim to build strong nonprofit sectors by enhancing the managerial capacity of other
nonprofit organizations (Brwon & Kalegonkar, 2002). Through offering capacity-building
services, intermediary organizations bridge funding entities (e.g. foundations and government
agencies) and community organizations. Szanton (2004) found that intermediaries link
foundations with their grantees by helping collect information and providing technical assistance.
Nonprofit capacity builders also mediate between government agencies and community-based
6
organizations by cultivating the managerial capacity of the community organizations that
encountered difficulty in partnering directly with government agencies (Shea, 2011; Smith, 2007;
Wynn, 2000).
Secondly, nonprofit intermediaries function as coordinators of collaborative networks for
policy implementation. Single organizations have faced challenges when dealing with
complicated social problems alone, therefore some organizations choose to partner with others to
address those issues through joint resources (Arya & Lin, 2007; Guo & Acar, 2005). In
particular, public and nonprofit organizations have widely collaborated in delivering various
social services (Johansen & LeRoux, 2012). However, organizational collaboration does not
always guarantee positive outcomes; organizations require additional effort for negotiating,
coordinating, and monitoring collaborative networks (Milward & Provan, 2000). Nonprofit
intermediaries manage and coordinate policy implementation and service networks, maximizing
the effectiveness of networks by connecting, brokering, and coaching participating organizations;
they facilitate communication among members of the network and encourage programmatic and
strategic collaborations (Shea, 2011).
Lastly, nonprofit intermediaries function as political network mobilizers. Advocacy
nonprofit intermediaries play a significant role in bridging individuals and organizations to shape
agendas. Specifically, they identify stakeholder groups who share similar interests and help them
to organize and facilitate coalitions and social movements by consolidating various social actors.
Furthermore, nonprofit advocacy intermediaries link citizens to larger political structures and
promote the participation of underrepresented groups (LeRoux, 2007). They provide a platform
for member organizations to network and promote common benefits, so they ultimately achieve
success in shaping policies that affect their members (Balassiano & Chandler, 2009).
7
Purpose of the Study
Intermediary nonprofit organizations are middlemen that link various social actors,
facilitating their collaborative in order to solve intense social, political, and economic problems
in our society. They aim to enhance the nonprofit sector and society at large by functioning as
capacity builders, coordinators of collaborate networks, and political network mobilizers.
However, despite their growing importance, intermediary nonprofit organizations remain
invisible, and we have little understanding of them. In order to illuminate their unknown aspects
and contribute to the organizational theory and nonprofit management research field, the chapters
of this dissertation investigate their various roles as capacity builders, coordinators of policy-
implementing networks, and political network mobilizers.
Chapter Two, “Intermediaries as Capacity Builders: Specialization in Nonprofit
Management Support Organizations,” focuses on intermediaries as capacity builders. This
chapter investigates the growth and evolution of nonprofit management support organizations,
and analyzes the factors affecting the diversification of their forms based on organizational
ecology and resource partitioning frameworks. Chapter Three, “Intermediaries as Coordinators
of Collaborative Policy-Implementing Networks: Equity and Effectiveness of a Job Assistance
and Training Program,” studies intermediaries as coordinators of collaborative policy-
implementing networks. By focusing on local workforce investment boards, which are the
agencies that implement and govern job assistance programs, this chapter investigates the role of
policy-implementing intermediaries in service outcomes (i.e. equity and effectiveness of job
training programs). Chapter Four, “Intermediaries as Political Network Mobilizers: The Role of
Social Media in Promoting Civic Engagement with Nonprofit Advocacy Organizations,”
examines intermediaries as political network mobilizers and investigates nonprofit advocacy
8
organizations, the ways they use social media for mobilizing civic engagement in sociopolitical
movements and the effects of these efforts on patterns of engagement.
Ultimately, the overarching question that this dissertation aims to answer is how do
nonprofit intermediaries support the sector and society. Chapter Five, “Conclusion,” closes the
dissertation by summarizing the major findings of this study, in addition to discussing study
limitations and contributions to theory and practice.
9
CHAPTER 2. INTERMEDIARIES AS CAPACITY BUILDERS: SPECIALIZATION IN
NONPROFIT MANAGEMENT SUPPORT ORGANIZATIONS
This chapter focuses on the role that intermediaries serve as capacity builders and
investigates the rise of specialist nonprofit management support organizations (MSOs). These
particular organizations are nonprofits that serve other nonprofit entities through training and
consulting on various organizational issues, such as leadership, human resource development,
staff development, building alliances, and knowledge and resource management (Abramson &
McCarthy, 2012; Backer, Bleeg, & Groves, 2004; Brown & Kalegaonkar, 2002; Connor, Kadel‐
Taras, & Vinokur‐Kaplan, 1999). Nonprofits face many challenges including the lack of basic
business infrastructure and managerial expertise (Glimer & Hughes, 2013; Wimberely, 2002). In
order to improve their limited organizational capacity, an increasing number of nonprofits are
turning to MSOs for these services. Although much attention has been placed on strengthening
the capacity of the nonprofit sector, only a few studies have explored nonprofit MSOs to date.
In order to shed light on the evolution of nonprofit MSOs, this chapter investigates the
growth and evolution of MSOs from 1990 to 2013 and the factors influencing their
diversification, including whether they adopt a generalist or specialist approach. Specifically, a
generalist MSO assists other nonprofits regardless of their service areas, whereas a specialist
MSO only targets a very limited number of clients within a specific service area (Abramson &
McCarthy, 2012; Backer et al., 2004). Over the past decade, the MSO field has evolved from
sector-wide management organizations assisting a variety of nonprofits in multiple domains, to
MSOs serving a dedicated service area. MSOs have been key in the sector in improving the
capacity of not only individual organizations but also the nonprofit sector as a whole, yet little is
known about the diversification of nonprofit MSOs. In addition, only a few empirical studies
10
have examined the organizational population of nonprofit MSOs. The population of specialist
MSOs grew rapidly over the past 20 years, but the growth slowed down and the total number of
specialist MSOs even decreased after 2005. However, we have limited knowledge about the
evolution of the organizational population of nonprofit MSOs. Thus, this research aims to
investigate not only the factors fueling its growth and diversification but also the factors
preventing its future sustainability.
This study applies organizational ecology and resource partitioning frameworks to
nonprofit MSOs to explain their evolution and diversification. By analyzing National Center for
Charitable Statistics data from 1990 to 2013, this research finds that three distinctive
mechanisms affect the evolution of the specialist MSO population: resource partitioning between
generalist and specialist MSOs, density dependence of specialist MSOs and market capacity. By
drawing from organizational ecology theories, this study enhances our understanding of
nonprofit MSOs and the factors affecting their growth and diversification of their forms. This
research contributes to the nonprofit literature as this is the first study to examine the
diversification of nonprofit MSOs. Furthermore, this is the first empirical study to investigate the
factors affecting the growth and decline of the population of nonprofit MSOs over time. Despite
the growing importance of nonprofit MSOs and increased demands for their services in the
sector, nonprofit MSOs remain small in number. The empirical analysis of this research provides
the answer to the question why nonprofit MSOs are specialized and their further growth is
limited, which is an uninvestigated research topic in the sector.
Background
Nonprofit Capacity Building and Management Support Organizations
Amateurism is one of the challenges hindering the development of the nonprofit sector;
11
that is, many nonprofit organizations lack the managerial and organizational skills required to
survive and achieve their missions successfully (Brown & Kalegonkar, 2002; Cairns, Harris, &
Young, 2005; Minzner, et al., 2014). Much effort has been made to deal with the capacity
limitations of nonprofits. For example, funders have invested in technical assistance, training,
and capacity building programs (Minzner et al., 2014; Millesen, Carman, & Bies, 2010).
Representing one solution to address capacity needs, nonprofit MSOs are nonprofit organizations
that provide management assistance, organizational development, and other consulting and
support services, with the goal of improving organizational effectiveness (Abramson &
McCarthy, 2012; Connolly et al., 2003; Renz, 2008; TCC Group, 2010). Furthermore, they
provide a wide range of management and technical assistance services for staff development,
leadership skills, board governance, strategic planning, fundraising, and financial management
(Backer, et al., 2004; Brown & Kalegonkar, 2002; Connolly et al., 2003; Da Vita & Fleming,
2001). The primary task of nonprofit MSOs is to help community organizations accomplish their
missions effectively, and they ultimately contribute to the establishment of a strong nonprofit
sector and civil society as capacity builders (Brown & Kalegonkar, 2002).
The evolution of nonprofit management assistance services by specialized providers (i.e.
nonprofit MSOs) began in the late 1960s in response to nonprofit sector’s need for
rationalization and professionalization (Abramson & McCarthy, 2012). Rationalization refers to
“the integration of formalized roles and rules” in an organization (Hwang & Powell, 2009, p
272), and professionalization denotes “a shift away from amateur or personalized responses to
needs or problems and toward technical and often standardized approaches to providing services
that reflect expert knowledge gained through specialized training” (Abramson & McCarthy,
2012, p 438). The growth of the nonprofit sector and the expansion of government contracting
12
with nonprofits required these organizations to work in an effective and efficient manner by
improving their internal management process (Abramson & McCarthy, 2012). This led to more
widespread adoption of professionalism and rationalization among nonprofits, especially as the
sector has observed the success of organizations that embrace a rationalized and professionalized
organizational structure (Abramson & McCarthy, 2012; Hwang & Powell, 2009). Nonprofit
MSOs emerged to satisfy these specific needs, enhancing the professional development and
managerial capacity of nonprofit organizations. Since the 1970s, the population of MSOs has
grown significantly, and these entities played important roles in strengthening the sector as
capacity builders (Abramson & McCarthy, 2012). The number of nonprofit MSOs of the U.S. in
1990 totaled 1,670; this number grew to 3,150 in 2013.
2
Specialization and Generalization in Nonprofit Management Support Organizations
In recent decades, the organizational population of nonprofit MSOs has diversified from
sector-wide management organizations assisting a variety of nonprofits in multiple domains, to
MSOs serving a field occupied by nonprofits operating in a specific service area (Backer et al.,
2004). Based on theoretical concepts from the resource-partitioning literature, this research
defines the first type as generalist MSOs and the second type as specialist MSOs. Generalist
MSOs assist other nonprofits regardless of their service areas, whereas specialist MSOs only
target a very limited number of clients within a specific service area. As an example of a
generalist MSO, the Alliance for Nonprofit Excellence works to strengthen nonprofits toward
excellence by providing various types of management services, such as the Program for
Nonprofit Excellence, an intensive three-year capacity building program for nonprofit
organizations. While the Alliance for Nonprofit Excellence does not specify their client
2
The total number of nonprofit MSOs was highest in 2004, and it was 3857.
13
nonprofits’ service areas, National Arts Strategies, Inc., which is an example of a specialist MSO,
defines its mission as the stabilization of communities of arts organizations in the United States.
Another example of a specialist MSO in an education service area is Excellent Education
Development, which provides substantial technical assistance services for charter schools in
California. Despite the distinct difference between the two approaches of nonprofit MSOs, only a
few studies have focused on this topic.
Figure 2.1
The Number of Generalist and Specialist Nonprofit MSOs from 1990 to 2013
Source: Calculation based on National Center for Charitable Statistics Core Financial Files
Figure 2.1 shows the growth patterns of generalist and specialist MSOs from 1990 to
2013 in 170 metropolitan areas. The noteworthy pattern found in Figure 1 is that the population
of specialist MSOs sharply increased, while the number of generalist MSOs rose only slightly.
Furthermore, although the number of specialists doubled over the last 20 years, this number
started to decrease after 2005.
0
500
1000
1500
2000
2500
3000
3500
Specialist MSOs Generalist MSOs
14
Figure 2.2
The Number of Generalist and Specialist MSOs Founded and Disbanded from 1990 to 2013
Source: Calculation based on National Center for Charitable Statistics Core Financial Files
Figure 2.2 shows the founding and disbanding patterns of generalist and specialist
MSOs. The number of newly founded specialist MSOs has decreased after 2002, while the
number of disbanded specialist MSOs has sharply increased after 2002. The number of
disbanded specialist MSOs has exceeded the number of founded specialist MSOs after 2005, and
this is the reason why the total number of specialist MSOs has decreased after 2005 as shown in
Figure 2.1.
The research questions addressed in this study began with these observations: What
factors contribute to the increase then decrease of the population of specialist MSOs? In
addressing these questions, this study specifically investigates the factors affecting the founding
of new specialist MSOs based on the theoretical frameworks of organizational ecology and
resource partitioning.
0
50
100
150
200
250
Founding of specialists Disbanding of specialists
Founding of generalists Disbanding of generalists
15
Organizational Ecology: Resource Partitioning, Density Dependence, and Market Capacity
Organizational ecology has investigated environments in which organizations compete,
attempting to explain the process of the rise and demise of organizations in a population by
emphasizing competition. A population of organizations refers to an aggregate of organizations
that share similar missions and common resources in the same environment, such that they
contend for limited resources in a market (Hannan & Freeman, 1977). If organizations do not fit
into the environment, they will fail or be eliminated from the market (Baum & Singh, 1996).
Along with the environmental selection process, ecology theory believes that the population of
organizations reaches an optimal point through this competition mechanism (Hannan &
Freeman, 1977).
The main research questions in organizational ecology examine how environments
influence intra-population conditions, such as 1) the founding of new organizational forms and
organizations, 2) the disbanding of organizational forms and organizations, and 3) density
dependence and population dynamics (Singh & Lumsden, 1990; Aldrich & Ruef, 2006).
Population dynamics, which are endogenous processes including foundings, disbandings, and
density changes in a population, influence the future status of a population because they affect
resource availability and competition in the population (Aldrich & Ruef, 2006; Dobrev & Kim,
2006; Wholey & Brittain, 1986).
Density dependence refers to “the relationship between population growth processes and
the size of the population itself” (Aldrich & Ruef, 2006, p.214). According to Hannan and
Freeman (1989), the size of a population reflects two underlying processes: legitimation and
competition. An increase in density intensifies the legitimacy of a population at first, but this
increase will lead to high competition for resources and thereby hinder the influx of new
16
organizations into a market. In other words, the theoretical argument about organizational
founding is that prior foundings signal sufficient resources (e.g., knowledge, learning
opportunities, and networks in a market), encouraging the influx of new organizations; however,
as more organizations enter a market, increased competition discourages further foundings,
which means that there is a nonmonotonic pattern and an inverse curvilinear effect of density
dependence on foundings (Singh & Lumsden, 1990; Hannan & Freeman, 1977; Aldrich & Ruef,
2006; Tucker, et al., 1990).
In addition, many researchers have investigated issues of organizational mortality,
especially the factors affecting organizational disbanding. The effects of competition and density
dependence are the most prominent factors used to explain organizational mortality too; high
density in a market will increase the mortality rate of organizations due to the increase of
competition (Singh & Lumsden, 1990). Ecological perspectives have effectively addressed the
density dynamic issues of an organizational population and researchers in the nonprofit field
have also applied these ecological theoretical frameworks to investigate factors in the founding
or failure of nonprofit organizations empirically (Bielefeld, 1994; Twombly, 2003; Hager,
Galaskiewicz, & Larson, 2004; Tucker, et al., 1990; Galaskiewicz & Bielefeld, 1998).
Organizational births and deaths are influenced by market capacity. Organizational
competition in a market is occurred because of limited market capacity. Hannan and Freeman
(1990, p.940) noted that “as long as the resources which sustain organizations are finite and
populations have unlimited capacity to expand, competition must ensue.” Due to the finite
resources, a market has their own carrying capacity, which refers to the maximum density that an
environment can support (Howard & Ruef, 2006). The inverse curvilinear effect of density
dependence on foundings exist because the total population overshoot the environment’s
17
carrying capacity; when population density reaching the carrying capacity, the number of new
foundings declines at higher levels of prior foundings as a market encounters higher competition
for a finite set of resources (Galaskiewicz & Bielefeld, 1998; Tucker, 1990). Consequently,
market competition and population dynamics are determined by carrying capacity of a market.
One of the criticisms about the density dependence model is that the effect of density is
not evenly distributed across a population. In response to this observation, resource partitioning
theory assumes that competition within a population has different effects depending on the
niches where organizations are focused (Aldrich & Ruef, 2006). The idea of resource partitioning
theory derives from an organization’s strategy to maximize its viability. When a market is
dominated by strong organizations, smaller and weaker organizations fail unless they find a
niche where they can occupy a competitive position. Therefore, in order to survive in crowded
market conditions, weaker and smaller organizations must seek a viable position by targeting
their products to various resources segments, leading to the resource partitioning of a market and
generating two different forms of organizations, generalists and specialists (Carroll, 1985;
Carroll & Swaminathan, 2000). The resource partitioning perspective suggests two types of
organizations based on the nature of their niche: 1) generalist organizations target wide industry
areas, and 2) specialist organizations choose narrow homogenous targets and avoid competition
with generalists in a crowded market. Generalists and specialists are fundamentally interrelated
(Carroll, 1985). According to Carroll (1985), generalists and specialists compete for the same
resources in unconcentrated markets. However, as concentration increases, generalists begin to
focus on the center of a market, while specialists move to peripheral niches to avoid direct
competition with generalists. As a result, this leads to the resource partitioning of a market
(Carroll, Dobrev, & Swaminathan, 2002; Aldrich & Ruef, 2006).
18
Density dependence and resource partitioning models in organizational ecology suggest
that competition and market capacity (i.e. resources and potential clients) are key factors
determining positioning in a market; this implies that nonprofit MSOs that serve specialized
subfields employ differentiation strategies for maximizing their viability by avoiding competition
with generalist MSOs. Although specialist MSOs would not be able to occupy larger markets,
they can strengthen their position in a market niche by providing differentiated services. In order
to explore the factors affecting the founding of specialist MSOs, this study offers research
hypotheses based on these theoretical frameworks, and empirically tests them by using nonprofit
panel data from 1990 to 2013.
Research Hypotheses
Resource Partitioning between Generalist and Specialist MSOs
The main argument of resource partitioning theory is that organizational forms are
decided by competition and resource availability in a market. Carroll (1985) noted that
organizations tend to specialize by locating in peripheral niches of a market when market
concentration is high because specialists try to avoid direct competition with generalists (Aldrich
& Ruef, 2006). Thus, new organizations tend to be specialized when a market is occupied by
generalists. In this research, the density of MSOs refers to the MSOs-to-nonprofit organizations
ratio. A higher density of MSOs signifies that a higher number of MSOs are serving nonprofits in
the sector. Based on Carroll’s model of resource-partitioning theory, this study proposes the first
hypothesis regarding the relationship between generalist and specialist MSOs:
Hypothesis 1: The density of generalist MSOs is positively associated with the founding
of new specialist MSOs.
19
Density Dependence: Competition among Specialist MSOs
The density dependence model predicts an inverse U-shaped relationship between
organizational density and the rate of foundings of new organizations in that market (Hannan &
Freeman, 1989). The density dependence model assumes that increasing organizational density
in a market signals market success, so new organizations attempt to enter into a market.
However, after reaching the market capacity, high density in a market discourages the
establishment of new specialists as existing organizations compete with limited resources and
clients. In order to test the effect of competition among the specialists, this research propose the
following hypothesis:
Hypothesis 2: The density of specialist MSOs has an inverse U-shaped relationship with
the founding of new specialist MSOs.
In addition, prior disbanding of organizations is also expected to influence the founding
of new specialist MSOs because it can be a signal that a population has exceeded its carrying
capacity (Carroll & Delacroix, 1982). Therefore, this research assumes the relationship between
the disbanding of specialist MSOs and their new founding as:
Hypothesis 3: The disbanding of specialist MSOs in a certain service area is negatively
associated with the founding of new specialist MSOs serving that area.
Market Capacity and Organizational Niche: Clients and Demand for Services
Market capacity and organizational niche, which are represented by potential clients and
demand for services, are important factors for making a decision to establish an organization.
When organizational niche is expanding, organizations tend to participate in a market. As a
result, it can be assumed that more specialist MSOs would be established with an increase in the
number of potential client nonprofit organizations in a certain service area in which they provide
20
services. In addition, there can be a greater demand for management support services if a
nonprofit service area is competitive; nonprofit organizations desire management support
services for maximizing viability. More specialist MSOs would enter a market if there are high
demands for services in that service area. Therefore, this research hypothesizes the effects of
market capacity and organizational niche on the founding of new specialist MSOs as:
Hypothesis 4: The number of potential clients (i.e. nonprofits) in a certain service area is
positively associated with the founding of new specialist MSOs serving that area.
Hypothesis 5: The market competition of a nonprofit service area is positively associated
with the founding of new specialist MSOs serving that area.
Data and Method
Data
In order to assess the factors affecting the growth of specialist MSOs, this research built
a dataset using panel data for 170 metropolitan statistics areas (MSAs)
3
and the status of their
nonprofit sector (e.g., number of nonprofit organizations by service area, number of nonprofit
MSOs by service area, etc.) from 1990 – 2013
4
. MSOs serve organizations that are located in
geographically close places, and they compete with other MSOs near them rather than compete
with MSOs in distant places. Therefore, this research sets the geographical scope at the MSA
level and investigates the founding patterns of specialist MSOs within a metropolitan area.
National Center for Charitable Statistics (NCCS) Core Financial Files are used to build the
dataset in this research. The Core Files contain information about nonprofit organizations that
3
“A geographic entity delineated by the Office of Management and Budget for use by federal statistical agencies.
Metropolitan statistical areas consist of counties (or equivalent entities) associated with at least one urbanized area
of at least 50,000 population, plus adjacent counties having a high degree of social and economic integration with
the core as measured through commuting ties” (http://www.census.gov/population/metro/data/glossary.html)
4
The NCCS Core Files are available from 1989 to 2015. In order to address incomplete data issues, data from 1990
to 2013 are used in this research.
21
report gross receipts of at least $50,000; the information includes the unique employer
identification number (EIN), annual financial information (e.g. expense, income, asset, and
revenue), geographic location, founding year, and a mission area of nonprofit organizations. By
using the annual Core Files data that started from 1989, this study calculated the number of
nonprofit organizations and nonprofit MSOs, as well as the founding of new MSOs by their
service area in each MSA every year, in order to build the panel dataset.
The National Taxonomy of Exempt Entities (NTEE)
5
classification system is used to
distinguish the service areas of nonprofit organizations. In the Core Files, every nonprofit
organization provides an NTEE Core Code (NTEE-CC) for indicating its mission area. NTEE-
CC offers a definitive classification system for charitable entities according to their
organizational purposes and targeted service areas (NCCS, 2007). The classification system
consists of a letter identifying the service areas (i.e. arts, education, etc.) and two digits
indicating the specific mission of the nonprofit (i.e. advocacy, technical and management
support, etc.). In this research, nonprofit service areas are classified into 10 categories according
to 10 major NTEE categories
6
: arts, culture, and humanities; education; environment; health,
human services; international; public and societal benefit; religion; mutual benefit; and unknown.
Moreover, the NTEE-CC classification system is also used to identify nonprofit MSOs.
Generalist MSOs are nonprofit organizations that possess an ‘S50 – Nonprofit Management’
code, indicating nonprofits that “provide technical assistance for other nonprofit organizations
that need management support in areas like board development, facility administration, fiscal
administration etc.” (NCCS, 2007). Specialist MSOs are nonprofits that hold ‘02- Management
5
http://nccs.urban.org/classification/ntee.cfm
6
http://exploresusquehanna.org/knowledgebase/detail.php?linkID=4323&category=40023&xrefID=7275
22
& Technical Assistance’ as their NTEE code, indicating nonprofits that provide consultation,
training, and other forms of management assistance services to nonprofit groups in a specific
service area. For example, organizations having an ‘A02’ NTEE code are nonprofits that provide
management assistance services only to other nonprofits in a dedicated arts, culture, and
humanities service area. Meanwhile, nonprofit organizations having an ‘E02’ code are nonprofit
entities helping management and technical aspects for other nonprofits in the health area.
Using the Core Files to build the dataset for this research comes with challenges and
limitations. For one, many nonprofit organizations appear irregularly in the Core Files, i.e. only
in certain years, so their information is missing for other years in the dataset. Nonprofit
organizations are required to submit tax forms annually, but many of them are granted extensions
by the Internal Revenue Service. It is important to note that missing information in the Core Files
does not necessarily mean the disbanding of the organizations. While organizations may be
absent in some years, they may appear again later in the data (NCCS, 2006). As a result,
additional data management is required to address the missing data problem. Without treating
this issue, the calculation the number of nonprofits and MSOs can be distorted. Therefore, if an
organization is absent from the Core Files for several years but appears in later years, the
research considers that it was alive during the years of absence in the Core Files.
Secondly, the NTEE classification system may not reflect correct information about the
mission of the nonprofits (DiMaggio et al., 2002). There is a concern that, while the
classification system accurately represents the broad service areas of nonprofit organizations, it is
less accurate in denoting their specialized functions and activities. In order to avoid potential
mistakes in identifying nonprofit MSOs when using the NTEE-CC classification system, this
study utilizes the NTEE confidence rating in the NCCS Core Files. The Core Files provide the
23
NTEE classification confidence rating of a nonprofit organization; this provides information
whether identification of nonprofit service areas represented by NTEE-CC classification codes is
accurate or not. By counting a nonprofit MSOs only when it has an A or B confidence rating, this
study attempts to minimize potential bias and mistakes in sampling nonprofit MSOs from the
NCCS dataset.
This research addresses these two major issues with the dataset as explained above, but
there are still other limitations in the data. Information about many organizations is missing in
many years because nonprofits can be exempt from filing a tax form. Organizational assets are
required to calculate the level of competition in a nonprofit service area for Hypothesis 5. This
research imputes financial information for the missing years by using linear interpolation, which
can limit the accuracy of the analysis. In addition, the NCCS Core Files contain information only
on those nonprofits whose gross receipts are more than $50,000, so small organizations are not
included in the dataset. However, in spite of these limitations, the NCCS Core Files still provide
exclusive descriptions of the nonprofit sector and the only available panel data about the
nonprofit organizations. In minimizing the dataset’s bias, it is expected that using the Core Files
will allow meaningful analysis of the organizational dynamics of nonprofit MSOs.
Empirical Models and Variables
This research empirically analyzes the factors affecting the founding of new specialist
MSOs by using the MSA level panel data from 1990 to 2013. The dependent variable in the
empirical model is a binary variable reflecting the founding of new specialist nonprofit MSOs,
whether any new specialist MSOs are created in each MSA in a year; the empirical analyses is
conducted separately by the service areas.
Six independent variables are included in the empirical model for testing the research
24
hypotheses. Hypothesis 1 about resource partitioning theory is tested by a variable reflecting the
density of generalist MSOs in a MSA in a year, measured by the percentage of generalist MSOs
in the nonprofit sector. Hypothesis 2 about the density dependence is examined by a variable
reflecting the density of specialist MSOs in a MSA in a year, measured by the percentage of
specialist MSOs in a certain nonprofit service area, as well as the square of this variable.
Hypothesis 3 about the effects of organizational disbanding on the founding of new organizations
is tested by a variable reflecting organizational deaths in a MSA in a year, measured by the
number of disbanded specialist MSOs in a certain service area. Hypotheses 4 and 5 about market
capacity and organizational niches are tested by a variable reflecting the number of potential
clients measured by the number of nonprofits in a service area and a variable reflecting the
demand for services measured by the Herfindahl-Hirschman Index (HHI); HHI is an indicator of
the amount of competition and market concentration between firms based on their size. By using
the assets of nonprofit organizations, the HHI of each nonprofit service area is calculated for
each MSA in each year. The total population and median household income of a MSA in a year
are included in the empirical model as control variables. The data are obtained from American
Fact Finder,
7
but the data about population and median household income at the MSA level are
only available from 2005, so the 2005 MSA information is used to fill in the missing data before
2005.
This research uses the random effects logistic regression model
8
for predicting the
founding of new specialist MSOs with panel data on 170 MSAs from 1990 to 2013. Since the
new founding of a specialist in the current year is influenced by factors during the previous year,
7
https://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml
8
Fixed effects logistic regression model is not used in this research because it drops many MSAs from the analysis
that do not have any new founding of specialist MSOs from 1990 to 2013. So, this research uses random effects
models, and LR tests of rho for all models were statistically significant.
25
the dependent variable is lagged for one year. The empirical analysis is conducted separately by
nonprofit service areas because specialist MSOs compete with others in their service areas.
Among 10 major NTEE categories of nonprofit service areas, seven areas -arts, culture, and
humanities, education, environment, health, human services, public and societal benefit, and
religion- are analyzed in this research. International and mutual benefit service areas are
excluded due to the small number of specialist MSOs founded in 1990 to 2013. Unknown service
areas are also excluded from the analysis.
Results
Overview of Nonprofit MSOs
Table 2.1 presents descriptive statistics for the nonprofit MSOs in 170 metropolitan
areas in 2013, which is the most recent year in the dataset used in this research.
Table 2.1
The Average Number and Density of Nonprofit MSOs in 2013
Number of MSOs
Density (ratio) of MSOs
in a nonprofit service area
M (SD) M (SD)
Generalist MSOs 3.33 (7.29) 0.19% (0.21)
Specialist MSOs (Total) 15.64 (34.92) 0.85% (0.49)
Specialists in Art, culture, and
humanities
0.69 (2.10) 0.25% (0.60)
Specialists in Education 3.29 (7.41) 1.11% (1.37%)
Specialists in Environments 0.73 (1.65) 0.95% (2.36)
Specialists in Health 2.48 (5.36) 1.37 (1.87)
Specialists in Human services 4.16 (9.22) 0.66 (0.73)
Specialists in Public and societal benefit 2.87 (7.25) 1.20 (1.44)
Specialists in Religion 0.98 (2.28) 0.70 (1.49)
According to Table 1, the average number of generalist and specialist MSOs are 3.33
and 15.64, which are relatively small numbers. Furthermore, their densities in the nonprofit
26
sector, measured by the ratio of MSOs to the number of nonprofit organizations, are 0.19% and
0.85% of the nonprofit sector. This means that there are about 1.9 generalist MSOs and about 8.5
specialist MSOs for every 1,000 nonprofit organizations. Among seven service areas, health is
the service area having the highest density of specialist MSOs.
Logistic Regression Analysis of Founding of Specialist MSOs
Table 2.2 provides the results of the logistic regression analyses for predicting the
founding of new specialist MSOs by nonprofit service areas. Hypothesis 1, which assumes a
positive association between the density of generalists and the founding of specialist MSOs, is
confirmed in the areas of human services and public and societal benefit, although the
association was not statistically significant in the other areas. This result supports the idea of
resource partitioning theory, which posits that organizations tend to be specialized for avoiding
competition with generalists in a crowded market.
Hypothesis 2 and 3, focusing on the density dependence model surrounding the
competition effect among specialist MSOs, are statistically confirmed as presented in Table 2.2.
Across all nonprofit service areas, the square of the density of specialist MSOs shows negative
odds ratios and the density of the specialist shows positive odd ratios. The results accord with
density dependence theory, which explains that the high density of organizations negatively
influences the founding of new organizations due to the increase in market competition (Singh &
Lumsden, 1990). The theoretical model argues that increased density of organizations signals
sufficient resources and the business potential of a market, so it encourages an influx of new
organizations; nevertheless, growing the number of organizations in a market consequently
extends competition, therefore it ultimately discourages further inflow of organizations (Singh &
Lumsden, 1990; Hannan & Freeman, 1977; Tucker, et al., 1990). As presented in Table 2.2, the
27
empirical analysis of this research confirmed the inverse U-shaped relationship between the
density and the founding of organizations. Furthermore, according to the results in Table 2.2, the
negative association between the organizational death of specialist MSOs and the founding of
new specialist MSOs is confirmed in the areas of health, human services, and religion. This
result confirms Hypothesis 3, which implies that organizational disbandings can be a signal of
high competition in a market, so it ultimately discourages the founding of new specialist MSOs.
Hypotheses 4 and 5 about the effects of market capacity and organizational niches are
also confirmed by the results in Table 2.2. The analytic result points out that the number of
potential clients (i.e. number of nonprofit organizations) in a service area has a positive impact
on the founding of new specialist MSOs. However, the effect is very small, which indicates that
every increase of 1,000 nonprofit organizations in a service area heightens the odds of a new
specialist MSO forming by roughly 1 to 3%. The negative associations between HHI of a
nonprofit service area and the founding of new specialist MSOs presented in Table 2.2 confirm
Hypothesis 5 regarding the demand for services; high competition among organizations in a
nonprofit service area is positively associated with the founding of new specialist MSOs, as
many nonprofits may desire management support services in a competitive environment.
Ultimately, the logistic regression analysis supported all hypotheses in this study; as
resource partitioning and density dependence theories explain, the competition with generalists,
the competition among specialists, and market capacity are key factors affecting the founding of
new specialist MSOs. According to the empirical analysis, MSOs benefit from specialization by
avoiding direct competition with generalists. Potential clients and demand for services are
important factors in predicting new specialists, but the study results suggest that competition
among specialists is the most influential factor.
28
Table 2.2
Logistic Regression on the Factors Affecting Founding of New Specialist MSOs
Arts, culture, and humanities Education Environment Health
Odds ratio [95% CI]
Odds ratio [95% CI]
Odds ratio [95% CI]
Odds ratio [95% CI]
Resource partitioning
Competition with generalists
Density of generalists
3.148
[0.597, 16.599]
1.607
[0.614, 4.204]
2.838
[0.781, 10.307]
1.758
[0.602, 5.134]
Density dependence
Competition among specialists
Density of specialists
49.163
[14.917, 162.033]
***
5.832
[3.780, 8.996]
***
1.899
[1.597, 2.259]
***
3.303
[2.426, 4.495]
***
Density of specialists
2
0.416
[0.279, 0.622]
***
0.865
[0.819, 0.914]
***
0.984
[0.977, 0.990]
***
0.922
[2.426, 4.495]
***
Organizational death
Disbanding of
specialists
0.907
[0.494, 1.666]
0.877
[0.700, 1.098]
0.533
[0.167, 1.704]
0.587
[0.396, 0.871]
**
Market capacity & organizational niche
Potential clients
Number of NPs in a
service area
1.001
[1.000, 1.001]
***
1.001
[1.000, 1.001]
***
1.003
[1.001, 1.005]
***
1.001
[1.001, 1.002]
***
Demand for services
HHI of a NP service
area
0.058
[0.002, 1.444]
*
0.423
[0.128, 1.394]
0.099
[0.016, 0.611]
**
0.056
[1.813, 3.584]
***
Control variables
Population
2.378
[1.557, 3.631]
***
3.809
[2.573, 5.641]
***
2.555
[1.672, 3.905]
***
2.549
[1.813, 3.584]
***
Median household
income
0.212
[0.017, 2.653]
0.075
[0.011, 0.486]
**
0.223
[0.026, 1.928]
0.085
[0.016, 0.436]
**
Number of observations 3896
3898
3986
3881
Number of MSA 170
170
170
170
Wald Chi
2
102.96 *** 167.97 *** 115.36 *** 157.04 ***
Log Likelihood -324.401
-887.526
-383.243
-789.761
*p < .05, **p <.01, ***p < .001
29
Table 2.2 (continued)
Logistic Regression on the Factors Affecting Founding of New Specialist MSOs
Human Services Public and societal benefit Religion
Odds ratio [95% CI] Odds ratio [95% CI] Odds ratio [95% CI]
Resource partitioning
Competition with generalists
Density of generalists
2.568
[1.197, 5.511]
**
2.288
[0.963, 5.440]
*
2.153
[0.550, 8.433]
Density dependence
Competition among specialists
Density of specialists
8.549
[4.431, 13.213]
***
3.720
[2.742, 5.049]
***
3.928
[2.703, 5.709]
***
Density of specialists
2
0.754
[0.683, 0.945]
***
0.902
[0.865, 0.941]
***
0.918
[0.886, 0.952]
***
Organizational death
Disbanding of specialists
0.760
[0.610, 0.945]
**
0.825
[0.613, 1.110]
1.166
[0.693, 1.960]
**
Market capacity & organizational niche
Potential clients
Number of NPs in a service
area
1.001
[1.000, 1.001]
***
1.001
[1.001, 1.002]
***
1.001
[1.000, 1.002]
Demand for services
HHI of a NP service area
0.099
[0.013, 0.740]
**
0.185
[0.038, 0.891]
**
0.601
[0.175, 2.065]
***
Control variables
Population
2.506
[1.941, 3.234]
2.660
[1.952, 3.624]
***
2.936
[1.972, 4.604]
Median household income
0.155
[0.042, 0.574]
**
0.372
[1.952, 3.624]
0.923
[0.093, 9.168]
Number of observations 3989
3986
3875
Number of MSA 170
170
169
Wald Chi
2
351.01 *** 241.96 *** 113.96 ***
Log Likelihood -1036.199 -789.427 -430.556
*p < .05, **p <.01, ***p < .001
30
Discussion and Conclusion
MSOs have been termed the capacity builders of the nonprofit sector, but only a few
studies have focused on this topic. In order to fill this gap, this research aimed to investigate the
evolution of nonprofit MSOs and the factors influencing their specialization by adopting
organizational ecology and resource partitioning frameworks. This research focused on five
hypotheses regarding competition between generalists and specialists, competition among
specialists (i.e. density dependence and the effects of organizational deaths), potential clients,
and demands for services and tested them through the analysis of the NCCS data from 1990 to
2013 in 170 metropolitan areas.
First, from the perspective of an ecological framing of resource partitioning, market
conditions dominated by generalists facilitate the growth of specialists by forcing organizations
to specialize their service area to maximize viability. Thus, this research expected that the density
of generalists has a positive effect on the founding of new specialist MSOs, and it was confirmed
by the empirical analysis in the two service areas of human services and public and societal
benefit. Secondly, the high density of organizations having similar functions is expected to be
negatively associated with the growth of similar types of organizations because it aggravates
market competition. Based on this theoretical implication, this research assumed that the density
of specialist MSOs has an inverse U-shaped relationship with the founding of new specialist
MSOs, and this is confirmed in all nonprofit service areas. Third, organizational death is
expected to have a negative association with the founding of new specialist MSOs, and this is
confirmed by the empirical model in the areas of health, human services, and religion. Lastly,
market capacity and organizational niches, which are represented by the potential clients that
MSOs can serve and demands for services, are expected to be positively associated with the
31
growth of specialist MSOs and they are also confirmed by the empirical analysis.
From the analysis, this research finds that MSOs can benefit from specialization by
protecting themselves from direct competition with generalists. The results also point out that
specialist MSOs are subject to intense competition from the other specialist MSOs serving the
same service area. Potential clients and demands for services are positively associated with the
founding of new specialist MSOs as the hypotheses assumed. Furthermore, this research finds
that the current market size and capacity of nonprofit MSOs in the nonprofit sector are very
small, implying high competition already exists in a market as they have not grown bigger.
Nonprofit MSOs’ small market capacity can impede their further development. The number of
specialist MSOs sharply increased over the last 20 years, but in-depth analysis pointed out that
both generalist and specialist nonprofit MSOs are encountering high competition in the sector.
The small size of the nonprofit MSO population and high competition in the sector can imply
low demand for capacity building services by nonprofit MSOs, although this needs further
examination. Alternatively, this could mean that nonprofit MSOs are not actively utilized as
capacity builders in the sector despite considerable discussion of the importance of their role in
the nonprofit literature. Boris (2001) mentioned that researchers should have a responsibility to
connect the literature on nonprofit capacity building to actual practice. The results of this study
suggest that nonprofit MSOs are in a nascent stage; efforts to develop them are required because
MSOs are an effective management tool for strengthening the nonprofit sector.
There are several limitations of this study. First, there were challenges on using the
NCCS Core Files for this research. Although the Core Files are a unique panel dataset about
understanding the population of the nonprofit sector, it has missing information issues. The Core
Files are constructed by annual tax forms of nonprofit organizations that they submit to the
32
Internal Revenue Services. Since many nonprofits are exempted to submit their annual tax forms,
there are missing information in those years. Thus, this research had to fill missing data by using
linear interpolation. Furthermore, the Core Files do not provide information about disbanding of
organizations, while they provide information about founding years of organizations. Therefore,
this research had to presume organizational disbanding by tracing organizations’ last tax filing
year. Consequently, these data issues can limit the accuracy of the analyses of this research.
Second, the NTEE-CC classification system that this research used for figuring out generalist
and specialist MSOs may not perfectly identify them; the NTEE-CC classification system may
be less accurate in denoting the specialized functions and activities of nonprofits. This research
addressed this issue by using NTEE classification confidential rate, but future in-depth research
needs to be conducted with better methods for identifying nonprofit MSOs. Third, this research
assumes that nonprofit MSOs compete at the MSA level to serve nonprofit organizations.
However, some of MSOs are national level MSOs serving nonprofits nation-wide. For example,
Pathways Community Network, Inc. supports human services providers in seven states by
providing technical assistance and training services this research assumes that nonprofit MSOs
compete at the MSA level to serve nonprofit organizations. However, some of MSOs are national
level MSOs serving nonprofits nation-wide. For example, Pathways Community Network, Inc.
supports human services providers in seven states by providing technical assistance and training.
Because there are some MSOs whose services are available nation-wide, it is possible that
research findings and results are inaccurate if the analysis includes national level nonprofit
MSOs as this research built the dataset at the metropolitan level.
Despite the limitations of this study, this research is expected to contribute to the
literature and practice of nonprofit management by empirically analyzing the populations of
33
nonprofit management support organizations and investigating factors affecting their
diversification and growth. Although the diversification of nonprofit management support
organizations is distinct feature, only few studies have focused on it. This research is the first
empirical study to explore two different types of nonprofit management support organizations
and to empirically investigate factors affecting their population growth and decline. The findings
of this research provide the answer why the number of nonprofit MSOs remains small despite the
constant discussion of their importance in practice.
This research empirically explored the diversification and growth of the nonprofit
MSOs, yet we still have little in-depth knowledge about them. Thus, qualitative studies about the
service and outcome differences between generalist and specialist MSOs needs to be conducted
in the future. The role of nonprofit management organizations and their effectiveness in
strengthening the capacity of the nonprofit sector also would be next research questions that
needs to be investigated.
34
CHAPTER 3. INTERMEDIARIES AS COORDINATORS OF POLICY
IMPLEMENTATION: EQUITY AND EFFECTIVENESS OF A JOB ASSISTANCE AND
TRAINING PROGRAM
This chapter studies intermediaries’ role as coordinators of policy implementation by
focusing on local workforce investment boards (LWIBs), which are agencies implementing and
managing job assistance programs put forth via the Workforce Investment Act (WIA). LWIBs are
intermediaries that are positioned between federal and state government (i.e. funding entity) and
One-Stop operators (i.e. street-level bureaucracy delivering services). Furthermore, LWIBs are
coordinators, who broker partner organizations; they cooperate with government agencies, local
businesses, nonprofits, education institutions, and community-based organizations for planning
and implementing WIA programs in order to vitalize the local economy and workforce system.
LWIB is a central actor that influences performance and effectiveness of the WIA programs.
Despite the importance of LWIBs as policy implementation coordinators, they have received
little attention by scholars.
One of the notable characteristics of LWIBs is their organizational legal status. About
one third of 571 LWIBs of the United States are taking an organizational form of a quasi-
government, which is a government-controlled nonprofit organization, while the rest of LWIBs
are holding a form of traditional government entity. The quasi-governmental LWIB is a mixed
organizational form of public and private organizations for optimizing the process of policy
implementation. The founding of quasi-governmental organizations aims to enjoy benefits of
both public and private organizations. The quasi-governmental LWIBs seek to improve their
organizational effectiveness from the advantages of their organizational form, yet there is little
evidence or discussion on this topic. Therefore, this research aims to investigate the performance
35
differences of LWIBs according to their organizational form by studying those that implement
the WIA programs.
The main research question of this study is empirically investigating the role of the
policy implementing intermediaries and their organizational form in service equity and
effectiveness. With a sample of 160,486 recipients of the WIA Adult program served by 253
LWIBs in 17 states, this research finds that LWIBs selectively provide the WIA job training by
serving those more able to find employment, out of all those eligible, and the job training
services produce better labor market outcomes than WIA core or intensive services. Although the
empirical analyses indicate that organizational form is not significantly associated with the
equity of the program, it is significantly associated with program effectiveness. This study is
expected to contribute to the literature and practice of the WIA as well as the public management
literature more generally by empirically investigating the association between organizational
forms of social service agencies and the equity and effectiveness of government services, which
have rarely been addressed in previous studies.
Background
Description of the Workforce Investment Act Program
History of US workforce policies and the Workforce Investment Act. U.S. workforce
policy began with Wager-Peyser Act in 1933, which established a nationwide system of public
employment offices and job placement services. In the 1960s, the Manpower Development and
Training Act (MDTA) was legislated as the first employment and training program in the U.S.
MDTA was replaced by the Comprehensive Employment and Training Act (CETA) in 1973.
Subsequently, the Job Training Partnership Act (JTPA) of 1982 replaced CETA in order to
continue provision of job training services. In 1998, WIA replaced JTPA and it has been the
36
primary federal employment and job training programs.
9
WIA has worked to help low income
people to enter unsubsidized employment, improve the quality of the workforce, and reduce
welfare dependency (Besharov & Cottingham, 2011; Cebrien, 2012; O'Shea & King, 2001).
One of distinct features of WIA compared with previous programs is its One-Stop
Career Center service delivery system, which means “a conceptual approach to service delivery
intended to provide a single point of access for receiving a wide range of workforce development
and employment services, either on-site or electronically, through a single system.”
10
WIA
mandates the creation of One-Stop Career Centers in local areas, which are physical sites where
WIA job assistance services, training services, and other related social services
11
are provided.
Various federally funded social programs and providers coordinate, co-locate, and integrate
information and service provision at One-Stop Career Centers, so individuals who want to access
the WIA program and other social services can receive services conveniently.
WIA has six guiding principles. First, WIA integrates multiple employment and training
programs at the “street level” through the creation of One-Stop Centers. Second, WIA empowers
clients to select services which meet their needs. Third, WIA ensures universal access of the
services for all individuals without exclusion. Fourth, WIA highlights accountabilities of
implementing the programs by monitoring the annual performance of state and local Workforce
9
In September 2014, the Workforce Innovation and Opportunity Act (WIOA) was passed as the reauthorization and
modernization of WIA of 1998. As a result, WIOA replaced WIA on July 1, 2015. WIOA maintains much of the
structure and funding streams (Barnow, & Smith, 2015). The data used in this dissertation contains information on
individuals served by the WIA adult program between January 1, 2014 and June 30, 2016. Therefore, this
dissertation decides to use the name of WIA as it is, because it was the original name of the program.
10
https://law.lis.virginia.gov/vacode/title2.2/chapter24/section2.2-2470/
11
WIA legislation requires One-stop Career Centers to provide 17 federal programs, for instance Employment
Services (Wagner-Peyser), Trade Adjustment Assistance, Unemployment Insurance, Job Corps, V ocational
Rehabilitation Program, and Vocational Education (Perkins Act). In addition to these mandatory partners, some
federally funded programs, such as Temporary Assistance to Needy Families (TANF) and the Supplemental
Nutrition Assistance Program (SNAP), are provided at One-Stop Career Centers optionally (Decker & Berk, 2011).
37
Investment Boards (WIBs). Fifth, WIA emphasizes the role of local WIBs and private sector
actors to design and implement the programs which meet the local labor market’s demands and
supplies. Sixth, state and local WIBs have strong autonomy and flexibility in implementing for
each regional economy (Barnow & King, 2005; Beaulieu, 1999).
Structure of the WIA Adult program. WIA provides employment and training services
for three targeted groups: adults, dislocated workers, and youth. An individual who is at least 18
years of age is eligible for the WIA Adult services, and individuals who are laid off without
cause are eligible for the WIA Dislocated Worker services. Low-income youth aged 14 to 21 are
eligible for WIA Youth services if they have barriers to employment, such as a deficiency of
basic skills, being dropped out of school, etc. (Decker & Berk, 2011; Wandner, 2002; Workforce
Investment Act, 2000). This research focuses on analyses of the WIA adult program, which is the
largest part of the WIA activity.
The WIA Adult program consists of three levels of services: core services, intensive
services, and training services. Core services are designed to inform and educate individuals
about labor markets of local areas; the services include job listings, computer access, and
workshops on resume writing and interview skills. These are self-assisted services: clients can
access these services at One-Stop Career Centers and they can obtain job-related information,
post their resume, and submit job applications on online websites.
12
If individuals fail to find
employment, they are eligible for intensive services. Intensive services are staff-assisted job
search and occupational development services through direct interaction with staff in the One-
Stop Centers. These are more in-depth job assistance services, such as comprehensive
assessments of clients, group counseling for job seekers, individual counseling and career
12
https://www.careeronestop.org/ or state and local WIB websites.
38
planning, and internships or short-term prevocational experience placements. If core and
intensive services ae not sufficient for clients to find employment, they become eligible for
training services. State WIBs select eligible training providers, which can be educational
institution, private firms, and nonprofits. Training services are not universal, so clients must
apply for the programs and case managers select participants. Participants can obtain
occupational skill training, on-the-job training, private sector training, and so on. Trainings are
WIA-funded services and are financed through vouchers or Individualized Training Accounts
(ITAs). Clients can choose what training program they want to take but training providers have
to get certified from state WIBs and the skills need to be in demand in the local economy.
Priority for resource-intensive services (i.e. training services) is given to low income clients
whenever funds are scarce (D’Amico & Salzman, 2004; Decker & Berk, 2011; Javar & Wadner,
2004; Marco, Almandsmith, & Hague, 2003; O'Shea & King, 2001).
Local Workforce Investment Boards: Intermediaries as Policy Implementing Coordinators
Roles and functions of LWIBs. LWIBs are intermediary organizations that manage and
implement the WIA in local areas by coordinating the WIA system with federal, state, and local
agencies and One-Stop Centers. WIA legislation mandates establishment of LWIBs in local
areas; there are 571 designated LWIBs in the United States.
13
The WIA requires LWIBs to work
on strategic planning, policy development, and oversight of the local workforce investment
systems (Workforce Investment Act, 2000). Functions of LWIBs can be summarized as below:
14
13
According to the US Department of Labor (https://doleta.gov/performance/reporting/pdf/wibdefinitionspy2015.xlsx),
571 LWIBs exist in all 50 states plus Washington D.C. in 2015. 571 LWIBs manage the workforce investment
system of their designated local areas; a local workforce area may include several counties or it may be a
metropolitan area. The services of LWIBs are delivered by One-Stop Centers. Clients need to visit One-Stop Centers
in their local areas to obtain the services.
14
http://www.state.nj.us/njsetc/policy/certification/resources/documents/WIB%20Roles%20and%20Responsibilities
%202-2012.pdf
39
Creating the budget and monitoring fiscal expenditures of the WIA programs;
Working directly with business and industry of local areas for understanding the needs of
the local economy;
Overseeing, evaluating, and monitoring performance of the programs and services;
Designating and certificating a One-Stop Center operator, which is a single entity or
consortium of entities that operate a one-stop center;
Connecting, brokering, and coaching participation of public sector employers in the
workforce investment systems;
Establishing standards for procurements of services;
Developing strategic visions (e.g. identification of goals) for local workforce investment
systems;
Negotiating and reporting local WIA performance to the state WIB.
The listed functions indicate that the role of LWIBs as intermediary organizations of the
workforce investment system is highly significant because success of the WIA programs depends
on collaboration among various actors of the workforce system, such as local businesses, public
agencies, One-Stop operators, and state WIBs.
The number of board members of the LWIBs varies from 20 to 40 across local areas.
LWIB membership must include representatives from local business, local education institutions,
labor organizations, community-based organizations, One-Stop partners, and economic
development agencies; the majority (i.e. more than 50%) should be local business
representatives. The members of LWIBs are appointed by an elected chief executive officer of a
unit of a general local government in a local area, such as the mayor or county commission chair.
40
Flexibility and autonomy of LWIBs. WIA is the federally funded program, but the
regulation ensures strong flexibility and autonomy of state and local WIBs when implementing
the programs (Barnow & King, 2005; Beaulieu, 1999; Workforce Investment Act, 2000). LWIBs
can choose how to allocate their funds between core, intensive, and training services.
Furthermore, they can decide how quickly clients transit from core services to more resource-
intensive services (Decker & Berk, 2011). LWIBs can also decide a price cap on training
services, so the amount of dollars and duration of job training vouchers are significantly different
across localities (D’Amico & Salzman, 2004; Decker & Berk, 2011).
15
In order to produce the
greatest impact of the program, LWIBs make strategic decisions on how best to use WIA funds
given their local economic conditions (D’Amico & Salzman, 2004). As a result of this strong
autonomy and flexibility, implementation and outcomes of the WIA programs inevitably vary
across local areas.
Performance Evaluations of the Workforce Investment Act
Program outcomes of LWIBs are managed by state WIBs, and the U.S. Department of
Labor (USDOL) manages state level performance; USDOL negotiates annual expected levels of
performance with each state, and a state WIB negotiates these expectations with each LWIB in
its state (Decker & Berk, 2011; Workforce Investment Act, 2000). Performance data of LWIBs is
collected quarterly, and three common measures are used for evaluating the performance of WIA
Adult services, as follows:
16
15
Dollar limits range from under $2,000 per ITA holder to a high of $7,500 or more with a modal value of $5,000
(D’Amico & Salzman, 2004, p.120).
16
https://www.doleta.gov/performance/guidance/tools_commonmeasures.cfm
41
Entered employment: the number of adult participants who are not employed at the
starting date of participation and are employed in the first quarter after the exit quarter,
divided by the number of total adult participants who exit during the quarter;
Employment retention: the number of adult participants who are employed in the first
quarter after the exist quarter and are still employed in both the second and their quarters
after the exit quarters, divided by the number of total adult participants who exist during
the quarter;
Average earnings: for those adult participants who are employed in the first, second, and
third quarters after the exit quarter, total earnings in the second plus third quarters after
the exit quarter, divided by the number of adult participants who exit during the quarter.
State WIBs manage the performance of LWIBs through an incentive system. LWIBs get
graded as ‘meeting’ performance when they achieve 80% to 100% of their negotiated
performance level. LWIBs get graded as ‘failing’ when they achieve less than 80% of their
negotiated performance level, and LWIBs get graded as ‘exceeding’ when they achieve higher
than 100% of their negotiated performance level. State WIBs provide monetary incentives
17
to
LWIBs if they ‘exceed’ negotiated performance levels. LWIBs get imposed sanctions (e.g.
reduction in amount of grant, receiving technical assistance and corrective actions
18
) if they fail
to meet performance levels (Barnow & King, 2005; D’Amico et al., 2001; Heinrich et al., 2009;
Workforce Investment Act, 2000; Workforce Innovation and Opportunity Act, 2014).
17
The WIOA signed into law that eliminates incentive awards for state performance. Program year 2013 is the last
year that incentive grants were awarded to states under WIA.
https://www.doleta.gov/performance/results/incentives_sanctions_archived.cfm
18
If state WIBs fail to meet the State adjusted levels of performance, state WIBs get a reduction in the amount of
grants (WIOA, 2014). The Governor can take corrective action, such as requiring the appointment and certification
of a new local board, if LWIBs fail to meet the negotiated levels of performance (WIA, 2000; WIOA, 2014).
42
Performance Issues in the WIA Job Training Program
Equity Issues
The emphasis on the flexibility of LWIBs, performance evaluation, and incentive system
in the WIA legislation engenders unintended behavioral responses of LWIBs, called “cream
skimming” behavior. With great autonomy given by WIA legislation, LWIBs make strategic
decisions on the way they operate and manage the programs, such as how adults move through
the service levels, how priority for target groups is given, whether training is emphasized, price
caps placed on job-training, and so forth (D’Amico, et al., 2004). LWIBs can take their flexible
approach to dealing with clients. They can move clients more quickly from core and intensive
services to training services in the case that they believe clients demonstrate quick need of
intervention (D’Amico & Salzman, 2004). More importantly, LWIBs can choose who gets
training services, which are the most resource-intensive services. Consequently, the potential for
“cream skimming”, i.e., serving only those with greatest potential for success, has been a long-
standing controversial issue on providing job training services (Anderson, Burkhauser, &
Raymond, 1993; Bell & Orr, 2002; Heckman & Smith, 2004; Nilsen, 2002).
Some individual characteristics, such as age, sex, education, language limitation, veteran
status, offender status, and race has been studied as factors affecting substantial differences in
labor market outcomes. For example, there is discrimination against ageism in employment
because of a narrowed recruitment pool and employer’s perception on poor returns on
investment in human capital (Taylor & Walker, 1994). Women has lower employment rate than
men and their earnings is only about 70 percent of men’s (Blau & Kahn, 1996; Darity & Mason,
1998). Education is expected to have a positive association with job placement and earnings as
there are beliefs that more educated individuals have higher levels of human capital, skills, and
43
productivity (Bills, 2003; Weiss, 1995). Adults who have limited English proficiency often lack
education credential and have lower levels of literacy; consequently, they are concentrated in
low-wage work (Wrigley et al., 2003). Employment rate and earnings of ex-offenders are lower
than the others because of barriers, such as poor skills and work experiences, health problems,
and negative perceptions of employers on ex-offenders (Holzer, Raphael, & Stoll, 2003).
Individuals with disabilities have lower levels of employment rate and incomes than the general
population (Blackorby & Wagner, 1996). Blacks are less likely to receive job offers than whites
(Bertrand & Mullainathan, 2003; Kirschenman & Neckerman, 1991; Turner, Fix, & Struyk,
1991). However, veterans are frequently found to earn more than non-veterans because of the
expectations that they are more skilled (Tray, 1982). In sum, being older, women, less educated,
limited English proficient, offenders, and black would work as barriers to employment and
earnings, while being veteran may work as a favorable condition in labor market.
WIA encourages serving “at-risk” populations, such as low income or individuals having
barriers to employment, but state and local WIBs have discretion to select the population that
they will serve (Bradley, 2013). Funding constraints in combination with performance standards
and WIA’s evaluation system of WIA can make local areas narrow eligibility for job training
services. Particularly with the performance incentive system, LWIBs express concerns that the
evaluation system can promote “cream skimming” where in state and local WIBs spend a
significant effort on producing ‘good-looking’ performance rather than providing appropriate
services to clients (Barnow & Smith, 2015; Heinrich, 2007). For example, Holcomb and Barnow
(2004) found that current performance measures (i.e. calculating employment rate, retention rate,
and average earnings of participants) of the WIA programs produce a powerful disincentive for
44
LWIBs to serve people with disabilities because they need a longer time period and higher costs
for training services.
The potential of creaming behavior by LWIBs brings a strong concern about equity for
the WIA job training services, with regards to who gets what services. Furthermore, high levels
of performance among LWIBs can be a result of client discrimination, in which they provide
high-cost job training services only to program enrollees who already have employment
advantages. Previous studies found empirical evidence of creaming behavior among social
service agencies when providing job training services (Anderson, et al., 1993; Bell & Orr, 2002;
Buck, 2002; Courty & Marschke, 2007; Heckman & Smith, 2004; Holcomb & Barnow, 2004).
Thus, this research will investigate this issue further in the context of LWIBs.
Effectiveness Issues
As discussed, LWIBs have motives to perform well on the outcome measures of the
WIA programs (i.e. employment rate, retention rate, and average earnings after the programs)
because of the performance incentive system. LWIBs may focus their resources towards clients
who are more likely to help improve their outcome measures rather than those most in need
(Moore & Gorman, 2009). Therefore, creaming behavior of LWIBs can result in the performance
of LWIBs being overestimated. Consequently, the current evaluation using LWIB-level data may
provide imperfect information about their performance and the effectiveness of the programs.
Many researchers investigated impacts of the WIA programs (Chrisinger, 2013; Decker
& Berk, 2011; Fleissig, 2014; Heinrich et al, 2009; Hollenbeck et al., 2005; Moore & Gorman,
2009). Instead of core services, which are universal services, researchers focused on the analyses
of intensive and job training services. Fleissig (2014) calculated return on investment of a job
training program and intensive services from California One-Stop Centers, and the results
45
showed the positive impact of the WIA programs on earnings of recipients. However, he found
that job training programs had a lower return on investment than intensive services. Unlike
Fleissig (2014), who conducted an evaluation study on both intensive and training program, most
WIA evaluation studies focused on measuring the effectiveness of job-training program, which is
the most resource intensive and high cost services. Researchers examined the effects of agencies’
creaming behavior, which hinders the measurement of the net impacts of the programs. Thus, in
order to avoid the effect of creaming behavior on measuring the net impacts of the program,
researchers used a non-experimental method through propensity-score matching technique. This
method allows researchers to select matching pairs with identical information about job-related
abilities that affect labor market outcomes, so it helps reduce overestimation on the net impact of
job training program by controlling for the effect of creaming behavior (Chrisinger, 2013;
Heinrich et al., 2009; Murnane & Willett, 2010). Many researchers found a positive impact of the
WIA adult training programs on program participants’ earnings (Decker & Berk, 2011; Heinrich
et al., 2009; Hollenbeck, et al., 2005). However, Moore and Gorman (2009) argued that WIA
training intervention did not yield positive impacts after controlling for participant demographic
characteristics. Moreover, after minimizing the differences among WIA participants who
received training and less-intensive services by using propensity-score matching, Chrisinger
(2013) found that the two groups experienced approximately equivalent earning growth after the
program.
This review of the impacts of the WIA job-training program provides two significant
implications. Firstly, the current evaluation of the performance of LWIBs provides imperfect
information; although performance of a LWIB meets or exceeds the negotiated level of
performance, it may not indicate that the programs provided by that LWIB are effective because
46
of the potential of creaming behavior. Thus, performance of LWIBs needs to be measured with
controlling for the creaming effects. Secondly, the net impacts of the WIA job-training program
are still inconclusive. Further studies are needed for measuring the net impact of the training
program.
Equity and Effectiveness in a Quasi-Governmental Form of LWIB
One of distinctions among LWIBs that has received little attention is their organizational
forms. LWIBs are intermediary administrative agencies that manage and implement local
workforce investment systems by linking federal, state, and line staff at One-Stop Centers along
with various private and public partners in the local area. There were 571 LWIBs in 50 states in
2015, and 32% of the have taken an organizational status as a quasi-governmental form, which is
a government-controlled nonprofit organization.
19
The government-controlled nonprofit organization, which “mixes a private legal status
(i.e. incorporated as a nonprofit corporation) with public governance (i.e. having at least one
director who is appointed by a unit of government)”, is a particular subset of quasi-governmental
organizations
20
(QGOs) that has been prevalent in local areas (Mead & Warren, 2016, p. 292).
Mead and Warren (2016) discussed reasons for choosing a QGO form instead of a public agency,
which is a traditional organizational form. They noted that QGOs can be seen as a way of
overcoming government and market failures. A QGO provides a balance of government
influence and independence that is desirable for political and legal reasons. For example, a QGO
has a higher level of autonomy than public entities because it can be free from complicated laws
19
Guidestar, which is a search engine for nonprofit organizations, and each LWIB’s website were used for
determining the organizational form of LWIBs. Detailed procedure on this is described in the following section.
20
Researchers have expressed difficulties to define QGOs precisely (Cole, 1998, Moe, 2001). In this research, we
have focused on a particular type of QGO, “an organization incorporated as a private, nonprofit organization, but run
by a board of directors that is composed of government officials or directors appointed by a unit of traditional
government” (Mead & Warren, 2016, p. 289).
47
and regulations imposed on public agencies. Moreover, a QGO can obtain the trust of individuals
who are skeptical of the government. In addition to this, a QGO can facilitate a public-private
partnership or intergovernmental agreement because it is an embodiment of a partnership
between private and public institutions (Mead & Warren, 2016; van Thiel, 2004).
Quasi-governmental LWIBs (hereafter nonprofit LWIBs) are registered as nonprofit
organizations, but the board members are appointed by an elected chief executive officer of a
unit of general local government and they receive funding from federal and state government. In
practice, experts of WIA have reported some common advantages of forming LWIBs as a
nonprofit corporation.
21
Firstly, it is easier for nonprofit LWIBs than governmental LWIBs to
obtain alternative sources of revenue, such as giving from foundations (Becker, 2010). Secondly,
nonprofit LWIBs are nimbler on treating human resource and procurement issues than
governmental LWIBs; it is difficult to fire or discipline a public sector employee, and a high
level of rigidity in human resource management is counterproductive to organizational change.
WIA legislation also requires collaboration with local business firms, education institutions, and
community-based organizations for implementing the WIA programs. Considering the fact that
many in the private sector have a negative perception of government (Van Slyke & Roch, 2004),
formation of nonprofit LWIBs can be a way to facilitate collaboration with various partners in
private sector (Mendel & Brudney, 2012). This literature review about quasi-governmental
organizations supports the expectation that nonprofit LWIBs perform better than governmental
LWIBs in delivering the WIA programs due to the potential merits discussed above. However,
21
Detailed information about making a LWIB a nonprofit corporation is described here
http://www.onetacademy.mahermaher.org/view/1987/info . This presentation describes the comparative benefits of
nonprofit LWIBs as “nonprofit entity can often be more flexible with contracts/personnel, more able to remain
above political infighting and can more easily access foundation funds”.
48
the issue of performance differences between two different organizational forms of LWIBs has
received little attention in the research and even in practice.
Furthermore, the organizational status of LWIB can influence not only the effectiveness
(i.e. performance) of the programs but also the equity of the programs. Previous studies raised
concerns about the low quality of services and client discrimination when social services are
provided by private organizations (Heinrich & Lynn, 2000; Van Slyke, 2003). For example,
Heinrich and Lynn (2000) found a strong association between types of administrative structure
and emphasis of performance measurement. They found that service delivery areas with a larger
administrative role for local public officials were less likely to highlight performance
measurement, while the areas with more private sector representatives as administrative entities
(e.g. private industry council) were highly focused on performance measurement. They also
found that administrative entities with more public officials were more inclined to focus on
“hard-to-serve” groups. Therefore, it can be expected that quasi-governmental LWIBs, which are
administrative entities with more private sector representatives, show a higher level of creaming
behavior. Although creaming behavior of these agencies has been discussed in the literature,
research has not yet examined variations in LWIBs creaming behavior as a function of
organizational form.
Research Questions
Previous sections discussed the equity and effectiveness issues of the WIA job-training
program and the link between these issues and the organizational form of LWIBs. From the
literature review, the equity and effectiveness of the WIA job-training program as well as the
organizational form of LWIBs are significant topics that needs to be investigated. Therefore, this
research aims to investigate the following research questions. First, this research examines the
49
equity issues of the WIA job-training program. Are there any differences between the
demographic characteristics of clients who received training services and clients who received
only core or intensive services? What demographic characteristics of individuals increase the
likelihood of getting job training services? Consequently, do LWIBs discriminate against clients
with multiple employment barriers whose outcomes are likely to be low by providing fewer job
training services? Secondly, this research investigates the relevance of organizational form to
equity by comparing quasi-governmental LWIBs with governmental LWIBs. Are quasi-
governmental LWIBs more likely to show creaming behavior in their provision of job-training
services? Thirdly, this research investigates the effectiveness of the WIA job-training programs.
Are outcomes of job-training programs better than core and intensive services? Fourthly, this
research analyzes the effectiveness of the WIA job-training program by LWIB’s organizational
form. Are there differences in WIA job-training outcomes between nonprofit LWIBs and
governmental LWIBs? Does organizational form matter to the effectiveness of the program? By
investigating these research questions, this study is expected to contribute to the literature and
practice of WIA as well as public management literature by providing empirical evidence on the
association between organizational form and the equity and effectiveness of social service
delivery.
Data and Method
Data
The dataset used in this research is the Workforce Investment Act Standardized Record
Data (WIASRD) that is published by the U.S. Department of Labor (USDOL).
22
This dataset
provides information about the individuals served, services provided, and outcomes attained
22
https://www.doleta.gov/performance/results/
50
under WIA. WIASRD data is available from PY 2005 to PY2015 Quarter 4.
23
This research
especially uses the PY 2015Q4 Public Use WIASRD file, which contains data on individuals
who were served by WIA programs between January 1, 2014 and June 30, 2016; it contains all
clients who began participation before July 1, 2016 and either exited on or after January 1, 2014
or had not exited as of March 31, 2016.
24
The focus of this research is the WIA Adult program,
so this research limits the scope of data to individuals who participated in the Adult program. In
addition to the individual level data on those who participated in the WIA Adult program, LWIB
level data were collected through the State WIA annual reports.
25
These reports contain
negotiated and achieved performance levels of LWIBs in their states.
The organizational status of LWIBs, whether they are governmental or nonprofit, are
determined by using Guidestar and each LWIB’s website. As indicated above, there has been
little discussion about a quasi-governmental form of LWIBs in research and practice, so there is
no concrete data source about the organizational form of LWIBs. Thus, this research used the
Guidestar website, which provides a search engine for information about US nonprofit
organizations. Guidestar
26
covers more than 1.8 million Internal Revenue Service (IRS)-
registered nonprofit entities. After getting the list of LWIBs from USDOL
27
and Careeronestop
websites,
28
the Guidestar search engine and the website of each LWIB was utilized to figure out
its organizational form, i.e. whether a LWIB is registered as a nonprofit entity.
23
Annual data is available for PY 2005 – PY 2008. Starting in PY 2009, USDOL published quarterly WIASRD.
24
https://www.doleta.gov/performance/results/WIASRD/PY2015/PY_2015Q4_Quarterly_WIASRD_Report_13July2017.pdf
25
https://www.doleta.gov/performance/results/AnnualReports/annual_report.cfm
26
http://www.guidestar.org
27
https://doleta.gov/performance/reporting/pdf/wibdefinitionspy2015.xlsx
28
https://www.careeronestop.org/LocalHelp/WorkforceDevelopment/find-workforce-development-boards.aspx
51
Sample
The sample of this research is 160,481 individuals served by 253 LWIBs in 17 states.
Not all states provided information about the negotiated and achieved performance level of their
LWIBs, and those that did not provide this information in 2015 were excluded from the sample.
29
States with only one type of LWIB were also excluded from the research; having only one type
can be the result of state policy direction, so they were also excluded.
30
Finally, states that have
only one WIB are not included in the analysis.
31
As a result, 17 states were included in the
sample, namely, Alabama, Arkansas, Florida, Georgia, Illinois, Indiana, Maryland,
Massachusetts, Michigan, Minnesota, Jew Jersey, New Mexico, North Carolina, Pennsylvania,
Virginia, Washington, and Wisconsin. Among individuals who participated in the WIA Adult
program in these LWIBs, 160,481 individuals were finally included in the sample.
Analysis and Variables
The empirical analysis of this research has four research goals: 1) investigating creaming
behavior by LWIBs, 2) analyzing the equity of WIA Adult job training programs as a function of
the LWIB’s organizational form, 3) measuring the effectiveness of the WIA Adult job training
program, and 4) analyzing the effectiveness of WIA Adult job training program as a function of
LWIB’s organizational form. In order to investigate their creaming behavior, the demographic
information of clients who received job-training services and clients who did not (i.e. those who
received core or intensive services) are compared by using t-test as a preliminary step. For an in-
depth analysis of the creaming behavior, logistic regression is conducted for predicting the
29
California, Kentucky, Missouri, New York, Ohio, Oklahoma, Oregon, Tennessee, Texas, and West Virginia were
excluded from the research because they did not provide local level performance information in their annual report
in 2015.
30
Arizona, Connecticut, Hawaii, Iowa, Kansas, Louisiana, Maine, Mississippi, Montana, Nebraska, New
Hampshire, and Rhode Island were excluded.
31
Alaska, Delaware, Washington D.C., North Dakota, South Dakota, Utah, Vermont, and Wyoming were excluded.
52
probability of receiving training services. By using independent variables that are related to the
employment of clients, such as their education level, income status, language limitation, race,
gender, disability status, veteran status, offender status, employment status, single parenthood,
and public assistance recipient, this research predicts the possibility of receiving job-training
services and checks whether these individual characteristics are significantly associated with the
possibilities of receiving training services. Unobserved characteristics of local areas are
controlled by using a dummy variable for each local area.
The second step of the analysis investigates the effectiveness of the job-training services.
This research uses t-test on four outcome variables of the WIA programs (i.e. employment rate,
retention rate, average earnings after the program, and earnings changes after the program)
32
between clients who received job-training services and clients who did not. The original
USDOL’s WIA Adult program evaluation included only three measures: employment rate,
retention rate, and earnings. However, this research adds earnings change as one of outcome
measures; it is possible that clients who already have high-wage jobs get high-wage jobs again,
so including earnings changes as an outcome variable can be expected to address overestimation
of the net impact of the job training services.
After checking the preliminary results of the t-test, this research uses propensity-score
matching technique for measuring the impacts of the program. The propensity-score matching
entails identifying matching pairs of job-training recipients and non-training service recipients
with similar information about age, education level, and other job-related abilities (Chrisinger,
2013; Heinrich et al., 2009; Murnane & Willett, 2010). Regression analyses for assessing the net
32
The definition of three outcome measures, employment rate, retention rate, and average earnings after the
program, are described in the background section.
53
impact of the program are conducted with the matched sample, so the clients’ demographic
factors affecting outcomes of the program can be controlled. As a result, the propensity-score
matching method allows measuring the net impacts of the training program. After matching the
pairs of individuals, logistic regression analyses are conducted for two dependent variables (i.e.
employment rate and retention rate) because they are binary variables. The other two dependent
variables, earnings and earnings changes, were analyzed by using ordinary least squares
regression. The main research variable, whether clients received the job-training services or not,
is included as an independent variable along with the listed individual client-level demographic
variables. Previous studies found that the effects of training programs vary with local labor
market conditions (Heinrich & Mueser, 2014; Lechner & Wunsch, 2009), so unobserved
characteristics of local areas are controlled by using a dummy variable for each local area.
The last step of the research is comparing nonprofit LWIBs and governmental LWIBs in
terms of their creaming behavior and program effectiveness. First, logit and ordinary least
squares (OLS) regression analyses are conducted with the sample classified into two groups
according to organizational form. Second, in order to measure the effects of organizational status
on clients’ outcomes, a dummy variable for nonprofit LWIB is added to the regressions. An
interaction term reflecting the product of the organizational form dummy variable and a training
service dummy variable is also added in order to investigate the performance differences of the
job training services in the two organizational forms.
Results
Descriptive Statistics
The sample of this study is 160,481 individuals who participated in the WIA adult
program and were served by 253 LWIBs in 17 states, which are Alabama, Arkansas, Florida,
54
Georgia, Illinois, Indiana, Maryland, Massachusetts, Michigan, Minnesota, New Jersey, New
Mexico, North Carolina, Pennsylvania, Virginia, Washington, and Wisconsin.
Figure 3.1 presents the distribution of LWIBs by organizational form in the 17 sampled
states. Among 253 LWIBs in the sample, 125 were governmental LWIBs and 128 were nonprofit
LWIBs. As shown in Figure 3.1, the mix of form varies across states; nonprofit LWIBs were
dominant in Florida, Indiana, and Wisconsin, while governmental LWIBs were utilized more in
Georgia, Illinois, and New Jersey.
Figure 3.1
Distribution of Governmental and Nonprofit LWIBs by States
Table 3.1 provides the descriptive statistics of the sampled WIA Adult program
participants. Among 160,481 clients who participated in the WIA adult program, 34.99%
received job-training services. Clients who were served by governmental LWIBs were 44.14%,
and those who were served by nonprofit LWIBs were 55.68%. The average age of clients was
39.38 and 55.43% of them were female. Average education level is higher than high school and
0
2
4
6
8
10
12
14
16
18
20
Governmental LWIBs Nonprofit LWIBs
55
less than beyond bachelor. Less than 1% of the clients had limited English proficiency, 7.16% of
them were veterans, 6.40% of them were offenders, and 12.11% had a disability. Only 15.16 %
of the clients were employed, and 25.77 % were the recipients of public assistance. Clients who
had a low income were 49.30%, and 11.52% of clients were single parents. Blacks comprised
36.74% of the sample, which was the second largest racial group of the sample.
33
Table 3.1
Descriptive Statistics of the Sampled WIA Adult Program Participants (N=160,481)
Total
(N=160,481)
Core or
Intensive
recipients
(N=105,619)
Training
Recipients
(N=54,862)
Difference
Age 39.38 41.66 34.99 -6.67 ***
Sex (1=female) 55.43% 54.28% 57.66% 3.39% ***
Education (0-3 scale)
34
1.36 1.37 1.35 -0.03 ***
Limited English
language proficiency
0.98% 0.86% 1.22% 0.36% ***
Veteran 7.16% 7.88% 5.78% -2.10% ***
Offender 6.40% 5.26% 8.61% 3.35% ***
Employment status
35
15.16% 6.55% 31.73% 25.18% ***
Disability 12.11% 8.12% 19.81% 11.69% ***
Received public
assistance
25.77% 17.88% 40.95% 23.08% ***
Low income 49.30% 38.22% 70.63% 32.41% ***
Single parent 11.52% 5.91% 22.31% -11.52% ***
Race: black 36.74% 38.65% 33.07% -5.58% ***
Governmental LWIBs
70841
(44.14%)
46066
(65.03%)
24775
(34.97%)
Nonprofit LWIBs
89640
(55.86%)
30087
(33.56%)
59553
(66.44%)
*p < .05, **p <.01, ***p < .001
33
The largest race group is white (not Hispanic), which was 49.79% of the sample, while Hispanics were 8.33 % of
the sample.
34
Education level is coded on a 0 to 3 scale; 0 – less than high school, 1 – high school and equivalent 2- bachelor’s
degree and equivalent, 3 – education beyond bachelor’s degree
35
Employment status is coded as 0 when a client is unemployed, and it is coded as 1 when a client is employed or
employed but received notice of termination.
56
Table 3.1 also presents a comparison of the demographic characteristics of recipients of
training services and recipients of core or intensive services; all differences were statistically
significant. According to the results presented in Table 3.1, the training service recipient group
had a higher percentage of women, limited English language proficient clients, offenders, those
already employed, clients with disabilities, clients that received public assistance, low-income
individuals, and single parents than the core or intensive service group, but fewer veterans and
blacks.
Equity of the WIA Job Training Program
Equity between clients of job-training services and clients of core or intensive
services. In order to examine creaming behavior of LWIBs, results of a multivariate logistic
regression of the dummy variable indicating receipt of training services on the clients’
demographic variables are presented in Table 3.2. According to Table 3.2, age, sex, education
level, limited English language proficiency, offender status, employment status, receipt of public
assistance, low income, single parenthood, and race are significantly associated with the receipt
of the WIA job training services. Those who are older, women, more educated, with limited
English language proficiency, offenders, unemployed, and black are less likely to receive training
services. Except for education level, the results confirm the creaming behavior of LWIBs, which
means individuals having barriers to employment are less likely to receive the training services.
However, public assistance recipients, low-income individuals, and single parents are more
likely to receive job-training services. These results indicate that LWIBs tend to serve more low-
income clients as the WIA legislation requires, yet LWIBs discriminate against other important
client characteristics (e.g. sex, language ability, etc.) that significantly affect job market
outcomes.
57
Table 3.2
Multivariate Logistic Regression of Receipt of Job-training Services on Client Demographic
Characteristics (N=158,227)
36
Variables Odds ratio (95% CI)
Age 0.975 [0.973, 0.975] ***
Sex 0.815 [0.783, 0.841] ***
Education (base group = less than high school)
High school or equivalent 2.661 [2.474, 2.862] ***
Bachelor's degree or equivalent 1.894 p1.735, 2.068] ***
Beyond Bachelor's degree or equivalent 1.957 [1.799, 2.129] ***
Limited English language proficiency 0.627 [0.549, 0.715] ***
Veteran 0.960 [0.900, 1.023]
Offender 0.901 [0.849, 0.956] ***
Employment status(1=employed) 3.161 [3.018, 3.310] ***
Disability 1.003 [0.988, 1.018]
Received public assistance 1.197 [1.144, 1.252] ***
Low income 1.971 [1.881, 2.064] ***
Single parent 1.563 [1.490, 1.640] ***
Race: black 0.822 [0.791, 0.854] ***
*p < .05, **p <.01, ***p < .001
Equity of the WIA job-training services by LWIB organizational status. Table 3.3
provides a comparison of the demographic characteristics of clients in the two service groups for
each of the two LWIB organizational forms. Similar patterns can be seen in both governmental
and nonprofit LWIBs. One interesting finding is that nonprofit LWIBs seem to provide job
training services to clients who have more favorable conditions for job placement. For example,
both nonprofit and governmental LWIBs tend to provide training services to younger clients but
the age gap between the training service recipients and core or intensive services recipients is
36
The total number of observations is less than the total sample size, which is 160,481, because some local areas
are dropped from the analysis due to the absence of training service recipients.
58
greater in nonprofit LWIBs than in governmental LWIBs. Furthermore, blacks are less likely to
receive training services in nonprofit LWIBs than governmental LWIBs. Unemployed clients are
the majority of the recipients of both training services and core or intensive services. However,
the percentage gap between the two groups is smaller in governmental LWIBs (18.11%) than
nonprofit LWIBs (31.01%), which means nonprofit LWIBs are more likely to discriminate
against the unemployed clients to provide training services. Veterans are more likely to get
training services in nonprofit LWIBs than in governmental LWIBs; veteran status is expected to
have positive association with employment and earnings. This pattern towards age, employment
status, veteran status, and race is statistically confirmed by the differences in difference
technique.
In contrast, nonprofit LWIBs tend to have smaller differences between the training
services recipient group and the core or intensive services recipient group in the number of
clients with disabilities. For example, the percentage of the disadvantaged clients is higher in the
training services recipient group than the core or intensive recipient group, but nonprofit LWIBs
have smaller percentage gaps between the two groups. This implies that nonprofit LWIBs
provide job training to disadvantaged clients, but they are less willing to do so compared to
governmental LWIBs.
Another interesting finding is that both governmental and nonprofit LWIBs tend to
provide training services to public assistance recipients, low-income individuals, and single
parents, but governmental LWIBs are more likely to provide training services to those
individuals. This indicates that governmental LWIBs are more likely to service serve “at-risk”
population than nonprofit LWIBs.
59
Table 3.3
Differences between training services and core & intensive services recipients by the organizational form of LWIBs
Governmental LWIBs Nonprofit LWIBs
Differences in
differences
Total
No
training
Received
training
Difference
Total
No
training
Received
training
Difference
Age 39.35 41.62 35.12 -6.50 *** 39.40 41.69 34.88 -6.81 *** 0.31 *
Sex (1=female) 55.62% 54.36% 57.97% 3.60% *** 55.28% 54.21% 57.41% 3.20% *** -0.40%
Education 1.32 1.33 1.30 -0.03 *** 1.40 1.41 1.39 -0.02 *** -0.01
Limited English
language proficiency
1.16% 1.01% 1.45% 0.44% *** 0.84% 0.74% 1.03% 0.29% *** 0.15%
Veteran 7.52% 8.51% 5.67% -2.84% *** 6.88% 7.39% 5.88% -1.52% *** -1.32% ***
Offender 5.87% 4.61% 8.21% 3.60% *** 6.83% 5.76% 8.94% 3.18% *** 0.42%
Employment status 12.61% 6.27% 24.38% -18.11% *** 17.18% 6.77% 37.78% -31.01% *** 12.90% ***
Disability 11.50% 6.33% 21.11% 14.78% *** 12.59% 9.50% 18.73% 9.23% *** 5.55% ***
Received public
assistance
26.55% 17.67% 43.06% 25.39% *** 25.15% 18.04% 39.22% 21.18% *** 4.21% ***
Low income 51.33% 37.38% 77.28% 39.89% *** 47.69% 38.87% 65.15% 26.28% *** 13.61% ***
Single parent 4.26% 22.90% 10.78% 18.64% *** 12.10% 7.19% 21.82% 14.63% *** 4.01% ***
Race: black 38.32% 38.25% 38.44% 0.18% 35.49% 38.95% 28.65% -10.31% *** 10.49% ***
*p < .05, **p <.01, ***p < .001
60
Table 3.4
Multivariate Logistic Regression of Receipt of Job-training Services on Client Demographic
Characteristics by the organizational form of LWIBs
Governmental LWIBs
(N=68,801)
Nonprofit LWIBs
(N=89,426)
Odds ratio (95% CI)
Odds ratio (95% CI)
Age 0.977 [0.975, 0.989] *** 0.973 [0.971, 0.974] ***
Sex 0.797 [0.754, 0.841] *** 0.829 [0.790, 0.869] ***
Education (base group = less than high school)
High school or equivalent 2.613 [2.335, 2.925] *** 2.680 [2.434, 2.950] ***
Bachelor's degree or equivalent 2.161 [1.890, 2.470] *** 1.691 [1.504, 1.900] ***
Beyond Bachelor's degree or
equivalent
2.138 [1.875, 2.437]
1.837 [1.645, 2.051] ***
Limited English language
proficiency
0.808 [0.678, 0.964 ** 0.467 [0.382, 0.572] ***
Veteran 0.969 [0.988, 1.066]
0.951 [0.873, 1.037]
Offender 1.048 [0.953, 1.153]
0.819 [0.759, 0.884] ***
Employment status (1=employed) 3.460 [3.216, 3.723] *** 2.931 [2.761, 3.112] ***
Disability 1.003 [0.977, 1.029]
1.002 [0.983, 1.020]
Received public assistance 1.264 [1.181, 1.352] *** 1.156 [1.089, 1.227] ***
Low income 2.411 [2.246, 2.587] *** 1.680 [1.579, 1.787] ***
Single parent 1.958 [1.812, 2.115] *** 1.365 [1.284, 1.451] ***
Race: black 0.787 [0.743, 0.833] *** 0.860 [0.816, 0.907] ***
*p < .05, **p <.01, ***p < .001
Similar to Table 3.2, Table 3.4 provides the results of the logistic regression models
examining the factors that predict the receipt of job training services, comparing the results for
the two LWIB organizational forms. As confirmed in Table 3.4, the results of the logistic
regression provide evidence of creaming behavior of LWIBs. Age, sex, education level, limited
English language proficiency, employment status, receipt of public assistance, having low
income status, being a single parent, and being black are significantly associated with getting job
training services in both governmental and nonprofit LWIBs. Clients who are older, women, with
lower education level than high school, with limited English language proficiency, and black are
61
less likely to get job-training services, while these who receive public assistance, have low
incomes, are employed and single parents are more likely to receive them. This pattern of results
is similar to the results presented in Table 3.3, which shows creaming behavior of LWIBs.
However, unlike Table 3.3, which shows relatively clear evidence that nonprofit LWIBs
are more likely to prefer clients who have more favorable conditions for job placement, the
results presented in Table 3.4 are inconclusive regarding such distinction by organizational form.
For example, both governmental and nonprofit LWIBs are more likely to provide job training
services to clients who are younger, men, and without language limitations; however, nonprofit
LWIBs are more selective about age and language proficiency, while governmental LWIBs are
more selective about sex. Thus, it is difficult to conclude that nonprofit LWIBs tend to higher
levels of creaming behavior according to these results.
Effectiveness of the WIA Job Training Programs
Effectiveness of job-training services. Table 3.5 shows the estimated effects of the job-
training program compared to the core and intensive services with unmatched and matched
samples. Previous analyses provided evidence of creaming behavior, that LWIBs tend to provide
training services to clients who are most able to achieve high levels of labor market outcomes.
So, this research estimates the effects of the job-training program by using a propensity-score
matching method to avoid the overestimated performance caused by creaming behavior. In Table
3.6, “unmatched” results show the treatment effects (i.e. the job-training program effects) before
matching the sample and “matched” results show the treatment effects after matching the sample.
Before matching the sample, the training recipients have higher levels of outcomes on all four
measures and the differences between them and clients receiving core or intensive services are
statistically significant.
62
Table 3.5
Estimated Effects of the Job-training Program Compared with the Core or Intensive Services
Outcome
variables
Sample
Job
Training
Core/
Intensive
Difference S.E. T-stat
Employment
rate
Unmatched 0.81 0.67 0.14 0.003 49.43 ***
Matched 0.79 0.75 0.04 0.004 9.36
***
Retention
rate
Unmatched 0.91 0.86 0.04 0.002 19.89 ***
Matched 0.88 0.86 0.02 0.004 5.54 ***
Earnings
Unmatched 16650.65 13415.93 3234.72 77.40 41.79 ***
Matched 14585.53 12277.48 2308.05 119.29 19.35
***
Earnings
changes
Unmatched 7387.95 374.51 7013.44 74.79 93.77 ***
Matched 7151.67 4318.46 2833.21 122.72 23.09 ***
*p < .05, **p <.01, ***p < .001
The employment rate of the training recipients was 13.8% higher than the others, and the
employment retention rate was 4.4% higher. The average earnings of the training program
recipients after the program were $3234.72 higher than core or intensive service recipients. The
training recipients earned $7387.95 more after participating in the program, while core or
intensive service recipients made only $374.51 more. However, as shown in Table 3.5,
differences between the job-training program recipients and core or intensive program recipients
become smaller after matching the sample. The differences in employment rates and retention
rates decrease by as much as 9.7% and 2.1% after matching the sample, and the differences in
earnings and earnings changes between two groups decrease by as much as $926.67 and
$4180.23 after matching the sample. Particularly, the difference in earnings changes decreases
dramatically from $7013.44 to $2833.21 after matching the sample. This indicates that the net
impact of the training program is overestimated before using the matching method. In
conclusion, job-training program participants show a 4.1% higher employment rate and a 2.3%
63
higher job retention rate. Moreover, the job training recipients have $2038.05 higher earnings,
with $2833.21 greater earnings increase, than core or intensive service recipients in the sample.
Effectiveness of the WIA job-training program by LWIB organizational form. Table
3.6 presents the performance evaluation results of 253 sampled LWIBs in terms of three common
outcome measures of the WIA Adult program in 2015. USDOL evaluates the performance of the
WIA adult program on employment rate, retention rate, and earnings of the clients.
37
According
to Table 3.6, most LWIBs meet or exceeded their negotiated performance levels; they failed to
achieve only 1.19% of their targets. The interesting finding is that nonprofit LWIBs performed
significantly better than governmental LWIBs. Nonprofit LWIBs exceeded nearly 70% of their
performance measures, yet governmental LWIBs exceed only 58.06% of the performance
measures.
Table 3.6
USDOL’s Performance Evaluation of LWIBs in 2015
Governmental LWIBs Nonprofit LWIBs Total
p value
a
N % N % N %
Not Met 5 1.34 4 1.03 9 1.19 0.004**
Met 151 40.59 113 29.20 264 34.78
Exceed 216 58.06 270 69.77 486 64.03
Total 372 100 387 100 759 100
a
p-Value by Pearson χ
2
test
*p < .05, **p <.01, ***p < .001
37
USDOL’s evaluation system on LWIB’s performance is described in the previous section of this chapter. After
annual negotiation on the expected performance level, LWIBs get graded as ‘meeting’ performance when they
achieve 80% to 100% of their negotiated performance level. LWIBs get graded as ‘failing’ when they achieve less
than 80% of their negotiated performance level, and LWIBs get graded as ‘exceeding’ when they achieve higher
than 100% of their negotiated performance level.
64
At first glance, the USDOL’s evaluation results imply that nonprofit LWIBs
outperformed governmental LWIBs. However, due to LWIB’s creaming behavior, these results
are imperfect evidence to conclude that nonprofit LWIBs perform better; the higher performance
of nonprofit LWIBs could be the result of their behavior serving clients who already have
employment advantages. Therefore, this research conducts multivariate regression and logistic
regression on outcome measures using the matched sample to investigate more rigorously
whether nonprofit LWIBs performed better than governmental LWIBs.
Table 3.7 presents the multivariate logistic regression results of the effects of job-
training program and LWIB organizational status on clients’ employment and retention rates after
the program, controlling for other client demographic characteristics. The results of both the
unmatched and matched samples indicate that training program is significantly associated with
both employment rate and retention rate. It is a noteworthy that the odds ratios of the training
services decrease after matching the sample; the odds ratio of the training program on
employment rate decreases from 1.237 to 1.157 and the odds ratio of the training program on
retention rate decreases from 1.289 to 1.268. This confirms the fact that the impact of the training
program may be overestimated due to the creaming behavior if the evaluation of the program
uses the unmatched sample. Although the magnitudes of the odds ratios decrease, the job training
program is positively associated with increasing employment and retention rates after controlling
for client characteristics. However, nonprofit LWIBs do not have significantly different with
employment and retention rates especially with the matched samples. After matching the sample,
the interaction effect between nonprofit LWIBs and training services is significantly associated
with employment rate (OR=1.152, 95% CI=1.040, 1.275), which means that nonprofit LWIBs
65
performed better on the training programs than governmental LWIBs in terms of improving
client employment rates.
Table 3.8 provides the multivariate OLS regression results of the effects of training
program and LWIB organizational status on clients’ earnings and earnings changes after the
program, controlling for other client demographic characteristics. The job training services
significantly increase clients’ earnings and earnings change of the clients. Similar to the effects
of the training program on employment and retention rates as shown in Table 3.7, the magnitude
of the coefficient of the training program on earning changes decreases after matching the
sample, which implies that the effects of the training program are overestimated when the sample
is not matched. However, the magnitude of the coefficient of the training program on earnings
increases after matching the sample, which implies that the effects of the training programs are
underestimated when the sample is not matched. After matching the sample, the training service
recipients have $2792.35 more in earnings and $3627.168 more earnings change than those who
received core or intensive services. Although the variable indicating nonprofit status is not
statistically significant, the interaction between nonprofit form and the training program is
statistically significant. Unlike the results shown in Table 3.7, the interaction effect is negatively
associated with client earnings and earnings changes. This interaction effect indicates the
differences in the training service effect between nonprofit and governmental LWIBs, so the
result implies that the training program in nonprofit LWIBs is less effective than in governmental
LWIBs. In sum, it can be concluded that the job training program of nonprofit LWIBs was more
effective in increasing employment but less effective in increasing earnings.
66
Table 3.7
Logistic regression of the effects of training program and LWIB organizational status on employment rate and retention rate
Employment Rate Retention Rate
Before matching (N=136,455) After matching (N=35,890) Before matching (N=87,644) After matching (N=26,161)
Odds ratio (95% CI)
Odds ratio (95% CI)
Odds ratio (95% CI)
Odds ratio (95% CI)
Training 1.237 [1.163, 1.316] *** 1.157 [1.073, 1.248] *** 1.289 [1.175, 1.415] *** 1.268 [1.132, 1.420] ***
Nonprofit agencies 2.128 [1.157, 3.903] ** 2.264 [0.348, 14.720]
1.203 [0.490, 2.957]
1.087 [0.117, 10.053]
NP X Training 1.111 [1.022, 1.208] ** 1.152 [1.040, 1.275] *** 0.878 [0.777, 0.992] ** 0.882 [0.760, 1.024]
Age 0.983 [0.983, 0.984] *** 0.982 [0.980, 0.984] *** 1.000 [0.998, 1.002]
1.000 [0.997, 1.004]
Sex 0.927 [0.903, 0.951] *** 0.923 [0.872, 0.977] *** 1.094 [1.046, 1.145] *** 1.145 [1.053, 1.245] ***
Education (base group = less than high school)
High school or
equivalent
1.281 [1.225, 1.339] *** 1.478 [1.332, 1.640] *** 1.416 [1.312, 1.529] *** 1.328 [1.143, 1.544] ***
Bachelor's degree or
equivalent
1.385 [1.311, 1.463] *** 1.792 [1.569, 2.048] *** 1.532 [1.390, 1.690] *** 1.453 [1.199, 1.760] ***
Beyond bachelor's
degree or equivalent
1.427 [1.354, 1.504] *** 1.755 [1.548, 1.989] *** 1.680 [1.530, 1.845] *** 1.617 [1.346, 1.942] ***
Limited English language
proficiency
1.252 [1.080, 1.451] *** 1.098 [0.899, 1.342]
1.039 [0.850, 1.272]
0.853 [0.650, 1.119]
Veteran 0.862 [0.813, 0.903] *** 0.817 [0.743, 0.898] *** 0.887 [0.823, 0.955] *** 0.890 [0.770, 1.030]
Offender 0.857 [0.813, 0.903] *** 0.855 [0.789, 0.927] *** 0.719 [0.664, 0.778] *** 0.698 [0.622, 0.783] ***
Employment status
(1=employed)
1.884 [1.438, 2.468] *** 1.175 [0.736, 1.874]
1.743 [1.632, 1.849] *** 1.667 [1.510, 1.839] ***
Disability 0.949 [0.936, 0.962] *** 0.952 [0.932, 0.972] *** 1.002 [0.978, 1.027]
1.008 [0.973, 1.044]
Received public assistance 0.848 [0.814, 0.883] *** 0.885 [0.829, 0.945] *** 0.798 [0.749, 0.850] *** 0.865 [0.786, 0.952] ***
Low income 0.938 [0.906, 0.971] *** 0.873 [0.806, 0.946] *** 0.825 [0.776, 0.877] *** 0.758 [0.676, 0.850] ***
Single parent 1.093 [1.039, 1.150] *** 1.121 [1.045, 1.202] *** 1.149 [1.070, 1.235] *** 1.081 [0.978, 1.194]
Race: black 1.229 [1.195, 1.264] *** 1.082 [1.020, 1.148] *** 0.981 [0.934, 1.030] 0.959 [0.878, 1.047]
*p < .05, **p <.01, ***p < .001
67
Table 3.8
Multivariate regression of the effects of training program and LWIB organizational status LWIBs on earnings and earnings changes
Earnings Earning Change
Before matching (N=75,546) After matching (N=21,936) Before matching (N=85,592) After matching (N=25,310)
Coefficient (Std. Err.)
Coefficient (Std. Err.)
Coefficient (Std. Err.)
Coefficient (Std. Err.)
Training 2710.885 (2710.885) *** 2792.350 (172.750) *** 3734.520 (164.564) *** 3627.168 (188.637) ***
Nonprofit agencies -1695.424 (1379.051)
-2818.222 (2795.288)
2272.518 (1436.974)
-4578.985 (4075.202)
NP X Training -795.719 (207.674) *** -532.065 (224.288) *** -822.989 (213.356) *** -1005.635 (246.102) ***
Age 61.692 (2.974) *** 44.956 (5.016) *** -83.932 (3.050) *** -67.511 (5.485) ***
Sex -3341.906 (74.440) *** -2761.209 (125.331) *** -285.124 (76.688) *** -941.534 (137.328) ***
Education (base group = less than high school)
***
High school or
equivalent
2014.302 (152.750) *** 1950.054 (266.227) *** 22.358 (152.099)
1154.570 (284.410) ***
Bachelor's degree or
equivalent
7178.526 (178.919) *** 5713.256 (314.576) *** 544.012 (180.542) *** 2775.609 (339.909) ***
Beyond Bachelor's
degree or equivalent
5642.343 (171.890) *** 4736.999 (302.050) *** 789.047 (173.104) *** 2604.584 (325.227) ***
Limited English language
proficiency
-2325.028 (333.719) *** -1888.809 (414.714) *** 337.490 (342.678)
-1.065 (455.130)
Veteran 69.516 (135.445)
380.198 (231.488)
90.519 (138.881)
-283.695 (251.003)
Offender -1808.672 (152.917) *** -1176.892 (200.542) *** 712.990 (153.637) *** 677.205 (214.772) ***
Employment status
(1=employed)
890.200 (89.822) *** 1086.476 (129.938) *** -174.391 (93.580) * 163.771 (143.728)
Disability -93.429 (40.785) * -103.995 (53.121) * -21.863 (42.165)
-71.642 (58.173)
Received public assistance -1304.606 (108.423) *** -1422.619 (149.701) *** -304.805 (111.119) ** -494.321 (161.773) ***
Low income -1871.097 (98.318) *** -2197.843 (159.970) *** 2193.954 (101.638) *** 2003.582 (176.277) ***
Single parent 807.350 (112.715) *** 736.816 (150.178) *** -427.899 (116.755) *** -142.779 (163.376)
Race: black -2603.668 (82.339) *** -1528.218 (134.537) *** 153.110 (84.519) * -628.806 (146.703) ***
R-squared 0.2011
0.1834
0.1689
0.0908
Adjusted R-squared 0.1983 0.1744 0.1663 0.0821
*p < .05, **p <.01, ***p < .001
68
Discussion and Conclusion
This research aimed to investigate the equity and effectiveness of the WIA Adult job-
training program with consideration of the organizational status of the LWIBs. This study
investigated 1) whether LWIBs selectively serve some clients when providing job-training
program by serving those more able to find employment from among those eligible, 2) whether
nonprofit LWIBs are more likely to selectively serve clients than governmental LWIBs, 3)
whether the job-training program produces higher labor market outcomes than the WIA core or
intensive services, and 4) whether training programs in nonprofit LWIBs result in better
outcomes than programs provided by governmental LWIBs. These research questions were
empirically examined with the sample of 160,481 clients served by 253 LWIBs in 17 states.
Consistent with previous research (Anderson, et al., 1993; Bell & Orr, 2002; Buck,
2002; Courty & Marschke, 2007; Heckman & Smith, 2004; Holcomb & Barnow, 2004), the
possibility of creaming behavior by LWIBs was empirically detected. The findings of this
research revealed that LWIBs selectively provided the job-training services to clients based on
individual characteristics, such as age, sex, language proficiency, offender status, and race, which
significantly affected job placement. WIA emphasizes immediate job placement for the
unemployed, so clients can access training only after they fail to find job placement through core
or intensive services.
However, there were also some inconsistent findings about the equity issues associated
with the WIA job-training program. Hegewisch and Luyri (2010) found that women were more
likely to receive training, while the multivariate logistic regression analysis of this research
found that women are less likely to receive training after controlling for other factors. In the
sample of this study, 57.66% of the training service recipients were women, so it can be
69
considered that women were more likely to receive training, as Hegewisch and Luyri (2010)
found. However, women were also the majority of core or intensive service recipients, and the
descriptive statistics of the sample did not provide the total number of applicants by sex.
Therefore, considering the results presented in Table 3.2 and Table 3.4, it can be concluded that
women have more limited opportunities to obtain training than men. Another inconsistent finding
is about prioritizing training services for low income individuals. D’Amico and Salzman (2004)
found that low income or welfare recipients were less likely to receive training, even though
WIA regulations give priority for training services to low income individuals. However, the
findings presented in Table 3.2 and Table 3.4 indicated that low income clients and public
assistance recipients were more likely to receive training services after controlling for other
factors, which is inconsistent with the finding of D’Amico and Salzman (2004). Furthermore,
Hopkins, Hansman, and Monaghan (2009) noted that an incumbent worker would not be able to
access training unless the first two services failed to provide a better position. They indicated that
this can hinder the development of job skills and the competitiveness of an incumbent worker.
However, the logistic regression results shown in Table 3.2 and Table 3.4 confirmed this
argument by demonstrating that the employed are more likely to receive training services than
clients who are currently unemployed.
Although the findings of this research empirically confirmed the behavior of LWIBs that
prefer clients who have more favorable conditions for job placement when providing job training
services, results regarding the behavioral differences as a function of their organizational form of
LWIBs were inconclusive as shown in Table 3.4. Nonprofit LWIBs selectively provided training
based on age and language proficiency, but governmental LWIBs selectively provided training
based on sex, employment status, and race. Although Heinrich and Lynn (2000) argued that
70
administrative entities with more public officials are more inclined to focus on “hard-to-serve”
groups, but their claim was not confirmed in this study. A notable finding is that both types of
LWIBs gave priority for training services to low income clients and public assistance recipients,
but governmental LWIBs were more likely to do so. LWIBs provide the WIA services at One-
Stop Career Centers. Within these One-Stop Centers, collaboration among local agencies
providing social welfare services are important for integrated services (Cohen, Timmons, &
Fesko, 2005). It can be assumed that governmental LWIBs collaborate more easily with other
public social service agencies, so that low income clients and public assistance recipients
immediately get referred to training services.
The creaming behavior of LWIBs can exaggerate the impact of the WIA adult job
training services. Therefore, this study measured the impact of training services compared with
core or intensive services by using propensity-score matching. Although Chrisinger (2013) found
that training service recipients and core or intensive service recipients experienced
approximately equivalent earnings growth after the program, the results of the analyses presented
in Table 3.5, Table 3.7, and Table 3.8 indicate that training recipients achieved better outcomes
than the others. They had higher percentage of employment and retention rate as well as higher
earnings and earnings change. Although the magnitude of the impact decreased when analyzing
the matched sample, the outcomes for training recipients were still higher than for the others.
These findings are consistent with previous studies that found training services increase earnings
(Decker & Berk, 2011; Hollenbeck, 2009).
LWIBs have great levels of flexibility and autonomy in their governance and
administrative structures, the way they operate and manage the program, how adults move
through the service levels, how priority for target groups is given, whether training is
71
emphasized, and so forth (D’Amico et al., 2004). Because of a shortage of funds, LWIBs are
cautious in authorizing training services. To optimize the performance of their programs, LWIBs
make strategic decisions about how best to use WIA funds. Some areas demonstrate a strong
commitment to training, while others focus more on core or intensive services (D’Amico &
Salzman, 2004). Therefore, it is obvious that the impacts of the programs vary according to how
LWIBs design and implement them. Among various factors that may lead to performance
differences across LWIBs, this research focused on their organizational status as the key factor
affecting program effectiveness. Nonprofit LWIBs, which are quasi-governmental organizations,
have been established to improve organizational efficacy, so this research expected that nonprofit
LWIBs would produce better performance than government LWIBs.
A descriptive comparison of performance between nonprofit and governmental LWIBs,
which is presented in Table 3.6, seems to indicate that nonprofits performed better in
implementing the WIA Adult program. However, the results presented in Table 3.7 and Table 3.8
reveal the fact that training services of nonprofit LWIBs were more effective in increasing
employment rate and retention rate, but they were less effective in increasing earnings and
earnings changes. The findings presented in Table 3.8 are inconsistent with the findings of
Heinrich’s (2002) research, which found that if the private sector engages in implementing job-
training program as administrative entities of federal funds, program participants in those local
areas have higher earnings. Further investigation is needed to address this inconsistent result.
However, the findings presented here imply that nonprofit LWIBs may focus on quantity of job
placements instead of quality of job placements, similar to the argument by Heinrich and Lynn
(2000) that private sector representatives as administrative entities focus more on performance
measurement.
72
This research investigated the issues of equity and effectiveness in WIA job-training
programs. Although the findings revealed that LWIBs prefer clients who have more favorable
conditions for job placement, LWIBs may not hold all the blame for this behavior, as they face
complicated issues that may lead to this discriminative behavior. There have been some concerns
about the local the flexibility allowed by WIA legislation because it can cause ambiguity and a
lack of clarity in implementing the program. Furthermore, limited funding has been indicated as
a barrier of WIA implementation. Performance standards and expectations increased, but
resources are limited (Cohen, et al., 2005). Equity and effectiveness issues arise due to these
conditions. For example, the underlying philosophy and priorities of WIA do not favor people
with disabilities. WIA is based on short-term services with a goal of immediate employment, but
individuals with disabilities need long-term support more than other kinds of job seekers.
Furthermore, limited funding can make LWIBs hesitate to use resources for intensive and
training services, so they may spend more funding on core services. However, core services,
which emphasize self-guided services, are difficult to access for clients with disabilities
(Holcomb & Barnow, 2004). Buck (2002) noted that LWIBs have raised concerns about how to
assess the large number of clients to determine who gets intensive and training services because
of universal access of the WIA programs. Furthermore, some local areas reported difficulties in
providing training services to as many disadvantaged individuals as they would like (Buck,
2002).
Use of LWIBs underlines market-based strategies, which rely on market incentives (e.g.
performance standards) to encourage efficiency, however this contributes to the equity and
effectiveness issues that must be addressed. Firstly, performance standards need to reflect the
objectives of the programs, whether it is high impact of programs or equity of services (Barnow
73
& Gubits, 2003). Secondly, LWIBs need to consider ways to improve their performance other
than through creaming behavior. For example, some LWIBs contract with for-profit firms for
consulting services such as information system development, research, monitoring services, etc.
(Marco, et al., 2003). These efforts are significant in improving the quality of services and
management in the One-Stop system (Javar & Wadner, 2004).
There are several limitations of this study. This research investigated equity and
effectiveness issues in LWIBs according to their organizational form of LWIBs, whether it is a
governmental or quasi-governmental form. Although this research found meaningful differences
between them and supported the findings through the theoretical review, the detailed
mechanisms causing these differences were not empirically investigated in this study. In
addition, this study addresses an important research topic, namely a quasi-governmental form of
LWIBs, but deeper research on this topic should be pursed in the future. Although this research
found that training program participants had better labor market outcomes than recipients of
other services, this research does not provide any information about the return on investment of
the programs. For example, Fleissig (2014) found that training programs have lower return on
investment than intensive services. Thus, it could be controversial as to whether job-training
services, which are relatively costly services, are worth investing in. These questions should also
be addressed in future research. Lastly, this research used a certain period of data, which is 2014-
2016 program year information, so the long-term generalizability of the findings cannot be
determined. Future longitudinal research using time-series data on WIA programs could
contribute to a better understanding of the causal dynamics underlying the issues explained in
this research.
74
Despite the limitations of this study, this research is expected to contribute to the
literature and practice of the job assistance and training programs. By addressing program
effectiveness and service recipient inequity together, this study supports a recommendation that
performance incentives and evaluation systems should clearly reflect the objectives or program,
whether it is the maximizing impact of programs or service equity. Furthermore, the findings
warn about the riskiness of high ambiguity in policy implementation, which brings confusion to
public service agencies. Most importantly, this research highlights the importance of policy
implementing intermediary organizations and their influence on the equity and effectiveness of
governmental services.
75
CHAPTER 4. INTERMEDIARIES AS POLITICAL NETWORK MOBILIZERS: THE
ROLE OF SOCIAL MEDIA IN PROMOTING CIVIC ENGAGEMENT WITH
NONPROFIT ADVOCACY ORGANIZATIONS
This chapter focuses on the role of intermediaries as political network mobilizers by
investigating nonprofit advocacy organizations and their use of social media. The development
of information and communication technology, such as online social media, has significantly
changed the ways in which people interact with each other (Hara & Estrada, 2005). For example,
besides having direct personal interactions, people now build virtual social relations in cyber
space and communicate through social network services, such as Facebook and Twitter.
Arguably, these new online-based social network platforms have enriched civil society and
enhanced socio-political grassroots activities. Social media allow information and ideas to
diffuse rapidly, thereby increasing public awareness about social issues. Moreover, the internet
has helped democratize the process of collective action and facilitate civic engagement by
lowering the barrier to participate in social movements.
The purpose of this chapter is to understand the functions of advocacy nonprofits and the
ways they use social media for mobilizing civic engagement in socio-political movements.
Despite the rapid growth of web-based advocacy activities and online civic engagement, many
questions on this topic remain uninvestigated. Advocacy organizations use social media and
other online tools in order to facilitate the flow of information through online social networks
and encourage civic participation by removing constraints of time, space, and costs (Guo &
Saxton, 2014; Klein, 1999). However, there is no clear empirical evidence supporting the idea
that social media use promotes civic engagement. In order to fill this gap, this study aims to
explore empirically factors that facilitate network-based engagement on nonprofit advocacy
76
organizations’ social media.
By analyzing the civic engagement patterns of 70 advocacy nonprofit organizations on
Twitter, this research finds that advocacy nonprofits enjoy multiple benefits from using social
media. The findings of this research indicate network-based information diffusion effectively
elicits civic engagement with organizational messages on Twitter. Moreover, this research
provides evidence that the effects of network-based engagement differ by advocacy issue areas.
The findings of this research are expected to provide theoretical and methodological insights for
nonprofit scholars especially for two reasons. First, this research found that the variables that
have been widely used in previous studies have low predictive power; this implies that further
research needs to be conducted to find which factors influence advocacy nonprofits’ ability to
engage their stakeholders and citizens. Second, this is the first empirical research that illustrates
the impact of advocacy nonprofits’ social media use on civic engagement, especially non-
constituents of organizations.
Background
Advocacy Nonprofit Organizations and the Use of Social Media
Advocacy is a core function of many nonprofit organizations. Through advocacy
activities, nonprofit organizations attempt to influence politics and policies by representing
collective goals and interests (Guo & Saxton, 2014; Schmid, Bar, & Nirel, 2008). Advocacy
nonprofit organizations work as intermediary organizations by linking citizens to the larger
political structure, so they ultimately contribute to democratic governance (Guo & Saxton, 2014;
LeRoux, 2007).
Advocacy nonprofit organizations adopt and use online social media strategically
because it has been considered as an effective tool for influencing and mobilizing civic
77
engagement in socio-political movements by increasing the recognition of an issue, reducing
obstacles to participation, and reaching a multitude of individuals (Carty, 2010; Eaton, 2010;
Hara & Estrada, 2005; Haro-de-Rosario, Sáez-Martín, & del Carmen Caba-Pérez, 2016; Obar,
Zube, & Lampe, 2012). Social media is specialized for information diffusion; the use of social
media facilitates information diffusion by conveying issues and topics to members promptly.
Furthermore, the high speed and reduced costs for communication, the consistent accuracy of the
original message, and the interaction between organizations and the public are key features that
make social media an effective communication tool (Diani, 2000; Obar, Zube, & Lampe, 2012).
Moreover, social media facilitates public participation in advocacy activities. Various social
media platforms provide intuitive and easy interfaces that enhance accessibility to a wider
audience and many-to-many communication; not only do they help overcome time and spatial
constraints, but they also lower costs for communication and participation (Carty, 2010; Guidry,
Waters, & Saxton, 2014; Klenin, 1999; Park, Reber, & Chon, 2016). In addition, social media
enables groups to reach out to new constituents, engage with them, and influence them (Diani,
2000). Social media platforms, such as Twitter and Facebook, allow advocacy nonprofits to
spread their messages beyond members and supporters, to members’ acquaintances who may
have never heard of the organization before (Obar, Zube, & Lampe, 2012). These advantages
make social media widely used by advocacy nonprofits, which rely on efficient and effective
communication with their stakeholders.
Literature Review on Nonprofit Organizations and Social Media
In the nonprofit literature, there are two main research streams on social media use
among nonprofit organizations. The first research stream involves the adoption of social media
among nonprofit organizations and the factors facilitating or hindering the utilization of social
78
media (Briones, Kuch, Liu, & Jin, 2011; Curtis et al., 2010; Gálvez-Rodríguez, Caba-Pérez, &
López-Godoy, 2016; Kim, Chun, Kwak, & Nam, 2014; Nah & Saxton, 2013). Nah and Saxton
(2013) explored factors affecting the utilization of social media by nonprofit organizations. They
found that funding expenses, web capability, membership structure, board size, and
organizational efficiency are important determinants of social media adoption and usage. Gálvez-
Rodríguez, Caba-Pérez, and López-Godoy (2016) conducted research on a similar question as
Nah and Saxton (2013). They found that the highest degree of dependence on donors is a
significant determinant of Twitter usage for one-way communication, whereas an organization’s
small size and an online community’s large size are significant determinants for two-way
communication; one-way communication refers to unidirectional information dissemination from
an organization to stakeholders, while two-way communication indicates interactive dialogue
between an organization and stakeholders by receiving and providing information in both
directions (Gálvez-Rodríguez, et al., 2016; Waters & Jamal, 2011).
The second research stream involves a content analysis of social media messages (Guo
& Saxton, 2014; Kim, et al., 2014; Lovejoy, Waters, & Saxton, 2012). Lovejoy and Saxton
(2012) categorized social media messages by three forms of communication: 1) an “information”
category, whose messages contain information about an organization and its activities, 2) a
“community” category, whose messages attempt to build relationships with stakeholders by
interacting and conversing with them, and 3) an “action” category, whose messages ask their
members to do something for the organization. By analyzing Twitter messages based on this
typology, Guo and Saxton (2014) found that the majority of tweets from nonprofit organizations
belong to the “information” category, akin to the findings of Lovejoy and Saxton (2012)’s
research. This implies that nonprofit organizations’ social media utilization constitutes
79
predominantly one-way communication.
A literature review reveals that previous studies mainly focused on “what organizations
do with social media, not on how target audiences respond” (Guidry, et al., 2014, p.241).
Questions about “how the public responds to messages or which types of messages elicit more
favorable responses and higher levels of engagement” remain unanswered (Guidry, et al., 2014,
p.241). In order to fill this gap, researchers recently began to explore the topic of public response
(Carboni & Maxwell, 2015; Guidry, et al., 2014). Based on the criticism that previous research
rarely focused on the impacts of social media on promoting stakeholder engagement, Carboni
and Maxwell (2015) explored the types of Facebook posts that are most likely to influence user
engagement (e.g. post liking, commenting, and sharing). They found that posts with photos and
organizational spending on advertising are positively associated with stakeholder engagement in
Facebook. Despite recent efforts, a better understanding of the effectiveness of social media in
mobilizing civic engagement is needed. Therefore, this research aims to shed light on the
effectiveness of social media for promoting civic engagement by analyzing Twitter user
engagement with advocacy organizations.
Research Questions
Civic Engagement with Advocacy Organizations vial Social Media
The substantial benefit of using social media is that it effectively elicits civic
engagement, defined by “the process of involving individuals and groups that either affect or are
affected by activities of an organization” (Sloan, 2009, p.26). Many forms of engagement are
available via social media. One-way information diffusion from an organization to its
stakeholders is considered a weak form of engagement. Strong engagement can be achieved
when stakeholders can respond to an issue in any forms of involvement on an organization’s
80
social media pages (Devin & Lane, 2014). For example, on social media, stakeholders engage
with the organization by joining in the conversation embedded in an organization’s social media
pages (Men & Tsai, 2014). Actions on social media pages, such as the “like,” “favorite,”
“retweet,” and “comment” functions of Facebook or Twitter, are considered forms of
engagement (Lovejoy, et al., 2012; Men & Tsai, 2014; Watkins, 2017). Effective stakeholder and
public engagement via social media enhances the organization-public relationship, and the
relationship can drive the public to have supportive attitudes toward the organization (Men &
Tsai, 2014). Moreover, stakeholders can voice their opinions by interacting with an
organization’s social media page; the organization may deal with specific issues, if it knows the
interests of its stakeholders (Devin & Lane, 2014). Although there is much discussion on social
media being adopted for promoting civic engagement, research on this topic remains limited
(Haro-de-Rosario, Sáez-Martín, & del Carmen Caba-Pérez, 2016).
Network Effects of Social Media
A social network is the key feature embedded in social media platforms that enables
rapid and wide information diffusion, as well as effective mobilization of civic engagement. A
social network refers to “relationships among social entities and the patterns of these
relationships” (Wasserman, & Faust, 1994, p. 3). The concept of social networks is important
because “relationship linkages between social entities are channels for transfer or flow of
resources (either material or nonmaterial)” (Wasserman, & Faust, 1994, p. 4).
With social media, users can choose to be a “subscriber” or a “follower” of other users;
by following (i.e. subscribing) users or being followed (i.e. being subscribed to) by other users,
users build networks and relationships (Huberman, Romero, & Wu, 2008). The social network
relationship on social media is the way that users get information; followers or subscribers get
81
notified when the users they follow post a new message (Huberman, Romero, & Wu, 2008). Not
only do they do they passively receive information from the users they follow, users also actively
participate by using key functions of social media, such as “liking,” “favoriting,” “retweeting,”
and “commenting.”
Figure 4.1
Types of Information Diffusion Structure
Source: Goel, Anderson, Hofman, & Watts, 2015, p. 181.
Social networks embedded in social media enable viral diffusion of information. Figure
4.1 describes two types of information diffusion structures. Type 1 in Figure 4.1 shows a
broadcast diffusion structure, which means an organization’s information is only shared among
its members. On the other hand, Type 2 in Figure 4.1 shows a viral diffusion structure, which
indicates that an organization’s information is shared among its members, then passed on to non-
members through the social networks of members; ultimately, the information can reach a much
larger population (Bakshy, Rosenn, Marlow, & Ademic, 2012; Goel, Aderson, Hofman, & Watts,
2015; Obar, Zube, & Lampe, 2012). The viral spreading goes beyond merely reaching a large
population; it also signifies rapid and large-scale diffusion driven largely by peer-to-peer
spreading through social networks (Goel, et al., 2015). Nonprofit advocacy organizations use
82
social media due to the belief that social media effectively disseminate large amounts of
information to a vast number of people promptly through social networks (Laer & Aelst, 2010).
However, this topic has received little attention in the empirical nonprofit literature.
In addition, according to social contagion theory, social networks serve as mechanisms
that expose individuals to information as well as behavior of others; this exposure increases the
likelihood that network members will develop beliefs and attitudes similar to others in their
networks (Monge & Contractor, 2003). For instance, Twitter and Facebook allow users to
observe activities and opinions of their friends even without direct meetings or conversation.
Moreover, a personal connection is an important prerequisite for effective mobilization
(Gladwell, 2010; Obar, Zube, & Lampe, 2012); researchers found that people tend to participate
in social events in which their friends or family members are involved (Fisher & Boekkooi,
2010). In sum, social media has the ability to create “awareness” and increase “exposure”
through social networks, so that it reinforces outreach efforts and mobilizes civic engagement
effectively (Obar, et al., 2012). Therefore, nonprofit advocacy organizations expect that social
media will enable them to reach out to their members along with those who have never heard of
them before (Briones, et al, 2011; Obar, et al., 2012). However, the efficacy of using social media
for attracting the public and promoting their engagement has remained unexplored.
Information science researchers have attempted to investigate information diffusion
structures in social media’s online networks. During a study of Facebook’s message-sharing
patterns, Bakshy, Rosenn, Marlow, and Adamic (2012) found that individuals who are exposed to
signals about friends’ sharing behavior are more likely to share the same information and share
sooner than those who are not exposed. Bakshy, Hofman, Mason, and Watts (2011) analyzed the
patterns of information diffusion on Twitter according to user attributes and content. They found
83
that the largest information cascades tend to be generated by users who have a large number of
followers. They analyzed the information diffusion accounting for the nature of content being
shared. They found that content that is interesting and elicits positive feelings is more likely to be
shared. Moreover, they also found that some types of content (e.g., “lifestyle,” “technology”)
spread more than others (e.g., “business and finance,” “sports”).
Research about information diffusion throughout social network services has suffered
from difficulties surrounding observation and measurement (Bakshy, et al., 2011). This topic has
not been widely investigated in nonprofit literature due to the challenges of data collection and
the analysis of social media data.
Twitter as an Effective Tool for Civic Engagement
Twitter is the most popular social networking and microblogging site, where users can
broadcast short messages called ‘tweets’ to a global audience. A tweet refers to a short message
of 280 characters
38
or less that can be read by and shared with many other people (Park, Reber,
& Chon, 2016). One of the main differences between Twitter and other social network platforms
is the way that users connect. Twitter users can choose to subscribe to other users’ tweets by
becoming their “followers”—in comparison, Facebook relationships are reciprocal. On Twitter,
the users who follow an organization or an individual are called followers, while the people
being followed are called friends; the relationship between followers and friends is not
necessarily mutual (Bonzanini, 2016). Twitter users build personal social networks by becoming
followers and friends of other users. A user’s tweets are shared on his or her followers’ Twitter
pages, and the tweets are exposed to the followers of his or her followers again if users respond
38
Tweets were originally restricted to 140 characters, but the limit doubled to 280 characters in November 2017
(https://en.wikipedia.org/wiki/Twitter).
84
to the tweets. As a result, a user’s tweets can be read and diffused rapidly through his or her
personal network on Twitter. Due to these features, Twitter has been widely used among
nonprofit organizations, as it is an effective tool for keeping their members informed
continuously and creating relationships with individuals who had no previous connection to them
(Briones et al., 2011; Gálvez-Rodríguez, et al., 2016).
Kent and Taylor (1998) provided principles for successful dialogue with the public via
the Internet; the principles include the creation of a dialogic loop for allowing two-way
communication, the provision of useful information, the generation of return visits, and an
intuitive interface. Twitter provides user-friendly interfaces that permit two-way
communications, such as following, retweeting, replying, and mentioning. Considering that
nonprofit advocacy organizations have widely used, as an optimal way for disseminating
information and engaging with the public (Lovejoy, Water, & Saxton, 2012; Park, Reber, &
Chon, 2016), this research focuses on Twitter usage to analyze the engagement patterns of the
followers and non-followers of advocacy organizations.
Research Hypotheses
Resource Dependence Theory and the Use of Social Media
Resource dependence theory posits that organizations require an ability to acquire and
maintain resources for organizational survival and growth. As a result, managerial skills and
organizational strategies are critical especially for nonprofit organizations that heavily rely on
external stakeholders for obtaining resources. Therefore, nonprofit organizations implement
various tactics for effective communication and building strong relationships with stakeholders
(Hodge, & Piccolo, 2005; Guo, 2007; Pfeffer & Salancik, 2003; Verbruggen, Christiaens, &
Milis, 2011; Yeon, Choi, & Kiousis, 2007).
85
From this perspective, researchers have attempted to investigate factors affecting
stakeholder engagement with nonprofit organizations’ social media. Nah and Saxton (2013)
examined organizational characteristics influencing the adoption and use of social media among
nonprofit organizations. They found that organizational strategy (fundraising and lobbying
expenses, program service revenue), capacity (asset size, website age, the number of website
visitors), governance structure (membership basis, board size, program spending ratio), and
external environment (donor and government dependence) play a part in social media use and
utilization outcomes. Gálvez-Rodríguez, Caba-Pérez, and López-Godoy (2016) investigated
factors affecting the use of Twitter as a mechanism for disclosing information and establishing
dialogue with stakeholders. Among factors that include donor dependence, fundraising expenses,
organizational age, organizational size, online community size, social media activity, and board
size, they found that dependence on donor is significantly associated with Twitter use for one-
way communication, and organizational size and online community size are significantly
associated with Twitter use for one-way and two-way communication.
Applying resource dependence theory and considering previous studies, this research
distinguishes seven factors that are expected to be significantly associated with the engagement
of followers or non-followers of nonprofit advocacy organizations on Twitter: donor dependence,
fundraising expenses, organizational age, organizational size, online community size, social
media activity, and advocacy issue areas.
Factors Affecting Civic Engagement on Social Media
Donor dependence. Contributions, such as individual donations and foundation grants,
are one of the primary revenue sources for nonprofit organizations (Hodge & Piccolo, 2005).
When nonprofit organizations rely heavily on external stakeholders for procuring funding, they
86
actively use websites for disclosing information (Gálvez-Rodríguez, Caba-Pérez, & López-
Godoy, 2014). Thus, it is expected that advocacy organizations with greater donor dependence
utilize social media for effective communication with donors and to build strong relationships
with them (Gálvez-Rodríguez, et al., 2014; Nah & Saxton, 2013), and the public is more likely to
engage with nonprofits’ social media. From this perspective, this research assumes that donor
dependence is positively associated with an organization’s civic (i.e. followers and non-
followers) engagement on Twitter.
Fundraising expenses. Gálvez-Rodríguez, Caba-Pérez, and López-Godoy (2016)
assumed that nonprofits reduce their fundraising expenses to reach the public with more cost-
effective tools. Although their empirical model did not confirm this assumption, it was confirmed
by research conducted by Nah and Saxton (2013). Similar to their study, this research expects
that organizations that have high levels of fundraising expenses do not fully utilize social media
and are less likely to put effort toward promoting stakeholder engagement via social media. From
this perspective, this research expects that the fundraising expense ratio (i.e. % of expenses) is
negatively associated with an organization’s civic (i.e. followers and non-followers) engagement
on Twitter.
Organizational age and size. Organizational age and size have been used as significant
indicators of organizational reputation and capacity (Nah & Saxton, 2013; Saxton & Wang,
2014). According to the previous studies, older and larger organizations have higher capacity to
build close relationships with stakeholders in an online community (Gálvez-Rodríguez, Caba-
Pérez, & López-Godoy, 2016). Furthermore, they are considered to have greater expertise and
trustworthiness, which are significant factors influencing civic engagement (Liu, Liu, & Li,
2012). Therefore, this research expects that organizational age and size are positively associated
87
with an organization’s civic (i.e. followers and non-followers) engagement on Twitter.
Online community size and activity. Online community size and activity are expected
to have positive impacts on civic engagement with advocacy organizations via social media (Zhu
et al., 2011). If many users follow an organization’s tweets, the information has a higher chance
of being read and shared by a large number of Twitter users (Gálvez-Rodríguez, Caba-Pérez, &
López-Godoy, 2016). In addition, an organization that actively uses social media cultivates
loyalty in stakeholders (Liu, Liu, & Li, 2012). From the findings of previous studies, this
research expects that online community size and activity are positively associated with advocacy
organizations’ civic (i.e. followers and non-followers) engagement on Twitter.
Advocacy issue areas. In addition to above six factors, this research assumes that
advocacy issue areas are significantly associated with an advocacy organization’s civic (i.e.
followers and non-followers) engagement on Twitter. The issue areas of nonprofit advocacy
organizations have been used as control variables in previous studies on the use of social media
among nonprofit organizations (Nah & Saxton, 2013; Waters, Burnett, Lamm, & Lucas, 2009).
However, Saxton and Wang (2014), Waters (2007), and Waters, Burnett, Lamm, and Lucas
(2009) found that nonprofit organizations’ issue areas are significant factors affecting the
utilization of online tactics because nonprofits that advocate certain issue areas are more likely to
interact with stakeholders for the purposes of education and promoting donations. Therefore, this
research expects that civic (i.e. followers and non-followers) engagement with advocacy
organizations on Twitter differs across advocacy issue areas.
Method
Sample and Data Collection
Following the sampling strategy of Guo and Saxton (2014), sample organizations were
88
selected from Charity Navigator, which is the most utilized independent evaluator for nonprofit
organizations in the United States. Charity Navigator provides rating systems for more than
9,000 U.S. 501(c)(3) charitable organizations by assessing their financial health, accountability,
and transparency.
From the complete set of charities, the human and civil rights organizations whose
scope of work is at the national level and who hold a rating score of 4 stars (i.e. the highest rating
score), are selected for this research. As Table 4.1 shows, 70 of the 73 organizations that met the
listed conditions in December 1027 use Twitter. Consequently, these 70 human and civil rights
nonprofit organizations were included in the sample.
Table 4.1
Twitter Presence of the 73 Human and Civil Rights Organizations, December 2017
Size of organization
39
Twitter presence
Total
Yes No
Up to 3.5 M 29 1 30
3.5-13.5M 28 1 29
13.5M and up 13 1 14
Total 70 3 73
Python code was written for downloading the Twitter data of the 70 sample
organizations. Twitter offers “a series of application programming interfaces (API) to provide
programmatic access to Twitter data, including reading tweets, accessing user profiles, and
posting content” (Bonzanini, 2016, p. 52). By using Twitter API, user profiles and basic statistics
(tweet and follower counts, etc.) of the sample organizations as well as a list of their followers, a
39
Charity Navigator divides the size of organization into three categories based on organizational assets: 1) Up to
3.5 million, 2) 3.5 to 13.5 M, and 3) 13.5M and up.
89
list of followers of their followers, and the number of times their tweets have been favorited or
retweeted were collected in December 2017. User profiles and basic statistics (tweet and
follower counts, etc.) of the sample organizations are accessible without limitations. However,
the Twitter API limits access for downloading users’ tweets; only users’ most recent 3,200 tweets
are accessible, therefore the number of retweets and favorites and the lists of users who retweet a
user’s tweets are calculated with the most recent 3,200 tweets of the sample organizations. Other
organizational level data, including revenue from contributions, fundraising expenses, and the
organization’s founding year, were obtained from their tax form 990 and Charity Navigator.
Variables
Dependent variables. In order to promote civic engagement effectively, three
conditions—increasing the recognition of an issue, reducing obstacles to participation, and
enabling involvements in issues—need to be satisfied (Guidry, Waters, & Saxton, 2014). Twitter
meets all these conditions through its key features: 1) “tweets,” allowing information-sharing
through personal networks, 2) “mention” and “reply,” allowing direct communication between
users, and 3) “retweet” and “favorite,” allowing engagement of users (Guidry, Waters, & Saxton,
2014). On Twitter, users participate in dialogue with other users by favoriting and retweeting a
particular tweet (Bonzanini, 2016).
“Retweet” is the act of reposting someone else’s tweet for sharing with his or her own
followers (Golbeck, Grimes, & Rogers, 2010). A key feature, retweeting enables viral
information diffusion through users’ social networks by copying and republishing tweets, so that
shared information can reach any number of users simultaneously (Bakshy, Rosenn, Marlow, &
Adamic, 2012; Liu, Liu, & Li, 2012; Park, Reber, & Chon, 2016). On Twitter, a tweet is seen
only by the followers of users when a user posts it. If followers of a user believe that information
90
in a tweet is worthwhile to be passed on, the followers retweet that tweet; consequently,
followers of the user’s followers have an opportunity to read and spread information when the
followers of the user retweet it. The information in the tweet is diffused virally and rapidly
through users’ social networks, even if the author of the original tweet does not pay attention to
the information spread (Liu, Liu, Li, 2012; Xiong et al., 2012). Retweeting gives Twitter the
most powerful mechanism for diffusing information via social media (Xiong et al., 2012).
In this research, four dependent variables represent civic engagement with advocacy
organizations on Twitter: 1) the number of retweets per tweet, 2) the number of favorites per
tweet, 3) the ratio of retweeters (i.e. users who retweet a user’s tweet) among the followers of an
advocacy nonprofit, and 4) the ratio of non-followers among the retweeters. The first and second
dependent variables measure the levels of public engagement (both followers and non-followers)
with advocacy organizations. The higher the number of retweets and favorites, the more people
are engaging in dialogue with an advocacy organization via Twitter. The third dependent
variable, the ratio of retweeters among followers, attempts to examine how many stakeholders
actively participate in an advocacy nonprofit’s Twitter pages. The last dependent variable is the
ratio of non-followers among retweeters for measuring the levels of network-based engagement
on Twitter. As described, a user’s tweet is seen by only followers of the user at first; non-
followers have a chance to read it when their friends favorite or retweet it. Without social
networks, non-followers rarely have a chance to learn about the tweet and then retweet it. Thus,
the high ratio of non-followers among retweeters indicates that Twitter is effective in reaching
out to an organization’s non-constituents; ultimately, it elicits non-members’ engagement with
the organization through Twitter. This research is the first empirical study in the nonprofit
literature to use this variable, so this research is expected to contribute to the literature by being
91
the first to examine the impact of social media use to promote non-constituents of civic
organizations. Previous studies did not differentiate between engagement of constituents and
non-constituents of organizations, but this research investigates them separately in the empirical
analyses.
Independent variables. There are seven independent variables in this research. Donor
dependence is measured by the ratio of contributions to total revenue of an advocacy
organization. Fundraising expense ratio is measured by the ratio of fundraising expenses to an
organization’s total expenses. Organizational age is measured by the number of years since an
organization was founded. Organizational size is measured by the natural logarithm of an
organization’s net assets. Online community size is measured by the natural logarithm of the
total number of followers. Social media activity is represented by the number of tweets per
month. After reviewing the advocacy mission areas of the sample organizations on their websites
and Twitter pages, these issue areas are classified into 10 categories: 1) civil rights (overall
issues), 2) LGBT, 3) children, 4) democracy, 5) digital, 6) advocating for a specific interest
group, 7) religion, 8) reproductive, 9) special issues, and 10) women.
Analytic Strategies
In order to investigate the factors influencing the engagement patterns of followers and
non-followers with an advocacy nonprofit on Twitter, this research conducted multiple regression
analyses of the four dependent variables on the seven independent variables. Two dependent
variables, the number of retweets and favorites per tweet were analyzed by using ordinary least
squares regressions, while the other dependent variables, the ratio of retweeters among followers
and the ratio of non-followers among retweeters, were analyzed by generalized linear models
because they are proportional variables. After running regression models, contrast analyses
92
across advocacy issue areas were conducted for testing the significance of the differences
between advocacy issue areas.
Results
Descriptive Statistics on Twitter Use of Nonprofit Advocacy Organizations
Table 4.2 presents descriptive statistics about the organizational characteristics and
Twitter usage and engagement patterns of the 70 sampled advocacy organizations. According to
Table 4.2, the patterns of Twitter usage and civic engagement vary widely across the
organizations.
Table 4.2
Descriptive Statistics of the Sample (N=70)
Variable Mean
Standard
deviation
Minimum Maximum
Number of tweets 13,212 10,905 84 47,564
Tweets per month 151 115 1 508
Number of followers 64,490 199,889 253 1,429,525
Number of friends 4,626 19,466 0 163,807
Twitter start year 2010 1.72 2006 2015
Ratio of retweeted tweets 28.52% 14.81% 0.08% 59.15%
Retweets per tweet 28.21 125.60 0.15 1045.22
Favorites per tweet 46.18 231.32 0.13 1926.35
Ratio of retweeters among
followers
10.27% 4.50% 3.00% 28.09%
Ratio of non-followers among
retweeters
56.51% 10.99% 36.33% 82.61%
Users reached by 1-degree
connection
181,800,000 340,000,000 576,802 1,880,000,000
Organizational net assets $16,300,000 $32,400,000 $308,085 $246,000,000
Fundraising expenses 7.51% 3.85% 0.60% 17.30%
Donor dependence 87.98% 16.43% 27.10% 100%
Organizational age (months) 356.06 128.74 132 600
The average number of tweets per month ranges from 1 to 508. The average number of
93
followers is 64,490; this number is higher than 707 followers, which is the average number of
followers of 95 million Twitter users
40
. Among tweets of an organization, 28.52% of tweets are
retweeted on average. The average number of retweets per tweet is 28.21 and the average
number of favorites per tweet is 46.18 on average. The average ratio of retweeters among the
followers is 10.27%, which means that not all followers actively engage in an organization’s
Twitter activity. The average ratio of non-followers among retweeters is 56.51%, which means
that information in tweets are diffused through the social networks of users and ultimately elicits
engagement of non-followers without a previous connection to the organization. Furthermore,
more than half of the retweeters are non-followers, implying that non-followers more actively
participate in nonprofit advocacy organizations’ Twitter dialogue than the organization’s own
followers.
Factors Affecting Civic Engagement with Nonprofit Advocacy Organizations on Twitter
Table 4.3 provides a correlation analysis among the independent variables; it shows the
weak linear relationships between the independent variables. According to the correlation
analysis, organizational size and the number of followers are significantly correlated, evidence
that larger organizations are more likely to have more followers on social media. Moreover, the
Twitter activity of an organization is also significantly correlated with the number of followers,
which indicates that organizations actively using Twitter are more likely to have more followers.
40
https://kickfactory.com/blog/average-twitter-followers-updated-2016/
94
Table 4.3
Correlations of the Explanatory Variables
Donor
dependence
Fundraising
expense
Organizational
age
Organizational
size
# of
Followers
(log)
Tweets
per
month
Donor dependence 1
Fundraising expenses 0.057
1
Organizational age -0.324 *** -0.117
1
Organizational size 0.017
0.026
0.266 ** 1
# of followers (log) 0.172
0.225 * -0.068
0.384 *** 1
Tweets per month -0.015 0.073 0.009 0.174 0.639 *** 1
Notes: *p<0.1; **p<0.05; ***p<0.01
Table 4.4 presents information about the patterns of Twitter usage and civic engagement
by advocacy topic areas. According to Table 4.4, the number of total tweets, the number of
friends, the number of users reached by 1-degree connection, and the ratio of non-followers
among retweeters significantly differ across the issue areas. Table 4.4 provides important
findings; advocacy organizations related to LGBT, civil rights, religion, reproductive rights, and
women actively post more tweets compared to the other organizations related to children, digital,
interest groups, and special issues. Nonprofit advocacy organizations specializing in LGBT
issues actively follow other Twitter users, and they can reach a larger number of users by 1-
degree connection, compared to the other advocacy nonprofits in this sample; but this is because
they have the largest number of followers.
41
The results in Table 4.4 also show that non-
followers of civil rights advocacy organizations retweet organizations’ tweets more often and
participate in disseminating information, while information about other specific issues, such as
LGBT, children, digital issues, and religion, are less likely to be shared by non-followers.
41
The average number of followers was the highest in advocacy organizations that focus on democracy, suggesting
that the followers of these particular nonprofits are more influential users.
95
Table 4.4.
Twitter Activity and Engagement across Advocacy Topic Areas (Average Numbers by Advocacy Topic Areas)
Issue
Areas
# of
tweets**
Tweets
per
month
# of
followers
Average # of
friends***
Retweets
per tweet
Favorites
per tweet
Users reached
by 1-degree
connection***
Ratio of
retweeters
among followers
Ratio of
non-followers
among retweeters*
LGBT
(N=3)
19,431 23.76 364,677 56,068 48.35 79.60 952,900,000 4.70% 48.05%
Children
(N=5)
6,061 21.85 13,638 642 13.84 4.19 47,025,131 9.58% 46.27%
Democracy
(N=6)
12,752 23.63 24,279 2,606 5.43 4.09 137,900,000 10.14% 57.58%
Digital
(N=3)
8,764 25.67 143,494
531
35.71 33.13 280,900,000 9.10% 45.55%
Interest group
(N=16)
10,405 25.07 20,620 2,022 9.57 9.76 107,100,000 11.07% 57.29%
Civil rights,
overall (N=9)
15,654 24.38 179,755 2,529 125.77 138.11 306,600,000 9.24% 61.72%
Religion
(N=5)
16,425 25.36 27,165 1,659 9.90 13.05 85,325,084 9.15% 49.77%
Reproductive
rights (N=6)
16,516 20.36 38,087 2,816 25.24 20.87 152,500,000 10.75% 60.29%
Special issues
(N=13)
9,770 29.73 15,002 2,211 7.04 5.27 73,573,669 12.96% 59.90%
Women
(N=4)
29,457 27.68 67,215 6,602 10.76 13.31 296,800,000 7.39% 59.18%
Total
(N=70)
13,212 25.23 64,490 4,630 28.21 29.93 181,800,000 10.27% 56.51%
Notes: *p<0.1; **p<0.05; ***p<0.01 (p-Value by Pearson χ
2
test)
96
Table 4.5 presents the results of multivariate regression analyses, which only partially
confirmed the hypotheses of the research. The ratio of fundraising expenses is significantly
associated only with the ratio of retweeters among followers. Organizational age is positively
associated only with the number of favorites per tweet. The number of followers is significantly
associated with all dependent variables; however, the directions of the associations are different
across the variables. The numbers of retweets and favorites are positively associated with the
number of followers, which indicates that tweets of an organization with more followers are
more likely to be retweeted or favorited. On the other hand, the ratio of retweeters among
followers and the ratio of non-followers among retweeters are negatively associated with the
number of followers; it is important to note that a large number of followers does not necessarily
imply high levels of engagement (Haro-de-Rosario, Sáez-Martín, & del Carmen Caba-Pérez,
2016). The results imply that the number of active followers does not dramatically increase, even
as the number of the organization’s followers increases.
The differences in civic engagement patterns across the advocacy issue areas are also
partially confirmed. First, tweets of advocacy organizations related to interest groups, religion,
reproductive rights, special issues, and women are less likely to be retweeted and favorited than
tweets of civil rights advocacy nonprofits. Second, tweets of LGBT advocacy nonprofits are less
likely to be retweeted than tweets of civil rights advocacy nonprofits. Third, LGBT advocacy
nonprofits are likely to have a smaller ratio of retweeters among followers, which means that
only a small proportion of followers actively participates in LGBT advocacy nonprofits’ Twitter
dialogue. Lastly, civil rights advocacy nonprofits are likely to have a larger percentage of non-
followers among retweeters than LGBT, children, digital, and religious advocacy nonprofits.
97
Table 4.5
Factors Affecting Network-based Engagement on Twitter
Model I Model II Model III Model IV
Retweeted per tweet Favorited per tweet
Ratio of retweeters
among followers
Ratio of non-followers
among retweeters
Coefficient (Std. Err.) Coefficient (Std. Err.) Coefficient (Std. Err.) Coefficient (Std. Err.)
Level of dependency on donors
Donor dependence -3.311 (96.677)
-7.506 (109.110)
0.214 (0.290)
0.290 (0.303)
Fundraising expenses
Ratio of fundraising expenses -1.568 (3.841)
-1.724 (4.335)
-0.025 (0.012) ** -0.005 (0.010)
Organizational age
Age (months) 0.211 (0.127)
0.247 (0.143) * 0.000 (0.000)
0.000 (0.000)
Organizational size
Net asset (log) 19.643 (12.342)
21.492 (13.929)
-0.038 (0.036)
0.007 (0.044)
Online community size
Followers (log) 39.425 (14.024) *** 44.532 (15.827) *** -0.110 (0.056) ** -0.122 (0.045) ***
Social media activity
Tweets per months -0.131 (0.179)
-0.168 (0.202)
0.000 (0.001)
0.000 (0.000)
Advocacy issue area
(Reference group: civil rights)
LGBT -142.058 (76.101) * -130.554 (85.888)
-0.442 (0.234) * -0.359 (0.196) *
Children -53.733 (65.198)
-70.602 (73.582)
-0.269 (0.304)
-0.845 (0.294) ***
Democracy -59.504 (59.315)
-66.239 (66.943)
0.039 (0.168)
-0.263 (0.212)
Digital -125.149 (78.569)
-146.288 (88.673)
0.102 (0.219)
-0.518 (0.186) ***
Interest groups -110.516 (49.109) ** -123.010 (55.424) ** 0.097 (0.176)
-0.249 (0.196)
Religion -157.460 (64.707) ** -171.002 (73.028) ** 0.049 (0.207)
-0.496 (0.223) **
Reproductive rights -122.158 (60.161) ** -141.972 (67.897) ** 0.046 (0.156)
-0.121 (0.228)
Special issues -89.510 (51.760) * -100.466 (58.417) * 0.269 (0.199)
-0.180 (0.187)
Women -133.065 (72.189) * -143.306 (81.472) * -0.110 (0.231)
-.028 (0.216)
Constant -604.310 (183.216) *** -672.101 (206.778) *** -0.775 (0.549)
1.353 (0.708)
Number of Observations 70
70
70
70
Adjusted R-squared 0.2389 0.2350
Notes: *p<0.1; **p<0.05; ***p<0.01
98
After regression analyses, this research conducted contrast analyses that compared the
differences of civic engagement patterns across advocacy issue areas while controlling for other
independent variables. Table 4.6 shows the contrast comparison results from the grand mean of
dependent variables across advocacy issue areas. According to Table 4.6, advocacy issue areas
are jointly significant in explaining only the ratio of non-followers among retweeters. In the case
of organizations advocating for civil rights and women, their tweets are more likely to be
retweeted by non-followers than the other organizations, while tweets of organizations
advocating for children and digital are more likely to be shared by their followers. The results
also indicate that civil rights advocacy organizations are more likely to have higher levels of
civic engagement on Twitter than nonprofit advocacy organizations in other issue areas; the
tweets of civil rights advocacy organizations are more likely to be retweeted and favorited
compared to the grand mean of the sample organizations. LGBT advocacy organizations have
smaller proportions of retweeters among followers, while nonprofits advocating for special
issues have larger proportions of retweeters among followers than the average of the other
advocacy organizations. These findings imply that tweets about issues that are relevant to more
of the general population are more likely to be diffused through social networks on Twitter, while
tweets about specific groups and issues are less likely to be shared by non-followers.
99
Table 4.6
Contrast Analyses of the Patterns of Engagement on Twitter across Advocacy Issue Areas
Retweeted per
tweet
Favorited per
tweet
Ratio of retweeters
among followers
Ratio of non-
followers among
retweeters**
Grand mean
contrast
Grand mean
contrast
Grand mean
contrast
Grand mean
contrast
Advocacy issue area
LGBT -42.743
-21.210
-0.421 ** -0.053
Children 45.582
38.741
-0.247
-0.539 **
Democracy 39.811
43.105
0.061
0.043
Digital -25.833
-36.944
0.124
-0.212 **
Interest group -11.201
-13.666
0.119
0.057
Civil rights,
overall
99.315 *** 109.344 ** 0.022
0.306 **
Religion -58.145
-61.658
0.071
-0.190
Reproductive -22.843
-32.628
0.068
0.185
Special issues 9.805
8.878
0.291 ** 0.126
Women -33.749 -33.962 -0.088 0.277 **
Notes: *p<0.1; **p<0.05; ***p<0.01 (p-Value by F-statistics)
Discussion and Conclusion
This chapter aimed to investigate the efficacy of advocacy organization’s social media
usage by analyzing the factors affecting civic engagement with the organizations on Twitter. By
analyzing 70 nonprofit advocacy organizations’ Twitter data, this research found that civic
engagement patterns with the organizations on Twitter vary considerably. This research found
that the organizations’ tweets virally diffused to a large audience through social networks of
users on Twitter; both followers and non-followers of the organizations engage with advocacy
nonprofits’ Twitter accounts. The information in tweets that are retweeted and favorited are read
and shared not only by the followers of an organization but also by followers of the followers
who retweeted or favorited the original tweet; this is also confirmed by the finding of this
research that 56.51% of retweeters are non-followers, on average.
100
Although the findings of this research provide empirical evidence supporting the
effectiveness of social media for disseminating information and mobilizing engagement, this
research provided only limited support for the hypotheses about factors affecting civic
engagement on Twitter. Fundraising expenses had a negative association only with the ratio of
retweeters among followers, and organizational age had a positive association with the number
of favorite tweets. Although the number of followers had significant impacts on all dependent
variables, the directions of the association varied. The numbers of retweets and favorited tweets
are positively associated with the number of followers. This result is consistent with the findings
of Bakshy, Hofman, Mason, and Watts (2011), who demonstrated that large information cascades
tend to be generated by users who have a large number of followers. However, the number of
followers is negatively associated with the ratio of retweeters among followers and the ratio of
non-followers among retweeters; this implies that a large number of followers does not
necessarily result in higher levels of engagement. In sum, the empirical findings indicate that the
independent variables of this study have low predictive power; this implies that the variables that
have been used in previous studies do not effectively explain the civic engagement patterns on
Twitter.
This research also aimed to investigate the differences of civic engagement across
advocacy issue areas. Results indicate that civil rights nonprofit advocacy organizations tend to
have higher levels of civic engagement than other organizations while controlling for other
factors. The findings imply that the public is more likely to engage in advocacy issues covering
more general topics, and those issues are more likely to be diffused, while tweets about issues
relating to the advocacy work of special interest groups are more likely to be shared only by
members.
101
Twitter has been considered an effective tool for educating the public on topics related to
the nonprofit advocacy organizations’ missions (Guo & Saxton, 2014). If advocacy organizations
effectively facilitate engagement and ultimately influence the public, they can achieve their goals
and missions more easily (Guidry, et al., 2014). This study provides evidence that nonprofits
should devote resources for using social media strategically to promote civic engagement
(Carboni & Maxwell, 2015; Saffer, Sommerfeldt, & Taylor, 2013). However, the findings of the
analyses only partially confirmed the research hypotheses. The predicter variables of this
research do not effectively explain patterns of civic engagement with the organizations on
Twitter. In Gálvez-Rodríguez, Caba-Pérez, and López-Godoy (2016), these variables also did not
significantly explain Twitter usage patterns. Therefore, further research should be conducted to
explore the factors affecting civic engagement on social media. For example, Bakshy and his
colleagues (2011) argued that the nature of the contents being shared is significantly associated
with engagement of users on social media. They found that content that is more interesting and
sharable is more likely to be spread (e.g. promote user engagement on Twitter). In addition,
Child and Grønbjerg (2007) argued that the field of activity in which advocacy organizations
operate is a predictor of nonprofit advocacy activities. Similar to their argument, this research
found that advocacy issue areas are associated with the levels of civic engagement on Twitter.
The types of social media messages, i.e., whether they are informative or communicative
functions, also can be a significant factor in promoting civic engagement on social media (Guo &
Saxton, 2014). These are potential factors that may affect the patterns of civic engagement on
social media, and future studies need to be conducted with these variables.
The small sample size limits generalization of the findings. Future research should be
conducted with a larger sample size. Moreover, this research focused on the quantitative
102
measures of civic engagement, but future research is needed to determine what content drives
this engagement. Despite some limitations, this research provides empirical evidence about the
effectiveness of using social media by testing the viral information diffusion on social media. In
the nonprofit literature, this topic has rarely explored because of limitations of data and
methodology. This research contributes to the nonprofit literature both theoretically and
empirically by answering the effectiveness of using social media for promoting civic engagement
from social media data mining and analyzing techniques.
103
CHAPTER 5. CONCLUSION
Summary of Findings
The purpose of this dissertation is to examine the functions of intermediary nonprofit
organizations and how they support the sector and society. Nonprofit intermediaries are
brokering organizations that support other organizations in achieving their goals in effective and
efficient ways, helping them increase managerial capacity, mobilize resources, and build
collaborative networks. Nonprofit intermediaries have three main functions: capacity building,
coordination of collaborative networks, and mobilization of political networks. The preceding
chapters explained how these functions contribute to strengthening both the nonprofit sector and
society at large.
Chapter 2 investigated the capacity-building function of nonprofit intermediaries by
focusing on nonprofit management support organizations. Specifically, the chapter examined
these organizations’ evolution and the factors influencing their specialization based on the
theoretical frameworks of organizational ecology and resource partitioning. The empirical
analysis found that the number of nonprofit capacity-building intermediaries has dramatically
grown during the last 20 years as the demands for their services in the nonprofit sector have
increased. Nonprofit capacity-building intermediaries help other nonprofits overcome a lack of
managerial and organizational skills by providing management and technical assistance services.
According to the analysis, there are about 10 nonprofit capacity-building intermediaries for every
1,000 nonprofit organizations. It turns out that the number of these intermediaries is still very
small, and growth has slowed down despite a need for their existence. The findings showed that
high competition among intermediaries is hindering their further growth. In order to avoid high
competition, many of capacity building intermediaries specialize their service areas and focus on
104
helping others in that specific service area; consequently, capacity intermediaries are diversified
into two types, generalists and specialists.
Chapter 3 investigated nonprofit intermediaries’ coordinating function by focusing on
local workforce investment boards, which serve as the coordinating agencies for implementing
and managing job assistance and training programs in local areas. This chapter empirically
examined the role of policy-implementing intermediaries and the impact of their organizational
forms on service equity and effectiveness. Local workforce investment boards take two different
organizational forms: government-controlled nonprofit organizations and government agencies.
The empirical findings of the research reveal that the equity (i.e., whether clients have equal
opportunities to get job training services) and effectiveness (i.e., employment rate, retention rate,
earnings, and earnings change) of job training services significantly differ by policy
implementing intermediaries. In this research, the organizational form of policy-implementing
intermediaries is not significantly associated with the equity of the program, but it is significantly
associated with program effectiveness. Nonprofit workforce investment boards were more
effective in increasing employment rate and retention rates, but they were less effective in
increasing earnings than governmental workforce investment boards.
Chapter 4 investigated the political network mobilizing function of nonprofit
intermediaries by focusing on nonprofit advocacy organizations. This chapter examined the
functions of nonprofit advocacy intermediaries and the ways they use social media for
mobilizing civic engagement in sociopolitical movements. The empirical analysis showed that
advocacy intermediaries actively use social media for communicating with their stakeholders and
reaching out to new constituents. Analyzing 70 advocacy intermediaries’ Twitter use, this
research found that advocacy intermediaries benefit from using social media; their organizational
105
information and activities are diffused to a large audience through the social network, and their
messages are read and shared not only by their direct stakeholders but also friends of the
stakeholders. These findings indicate that advocacy intermediaries’ strategic use of social media
effectively elicit civic engagement on social media, which implies that advocacy intermediaries
are successful political network mobilizers.
Viewed holistically, this dissertation provides evidence that nonprofit intermediaries play
significant roles in strengthening the sector and the greater society. Nonprofit capacity-building
intermediaries bolster the nonprofit sector by providing management support services to other
nonprofits. Policy-implementing intermediaries produce different outcomes according to their
organizational form, and they play a major role in effective social service provision. Advocacy
intermediaries effectively promote civic engagement by using social media. Although
intermediaries contribute to the sector and society in different ways, they help organizations
effectively collaborate with other social actors and reduce transaction costs. Nonprofit capacity-
building intermediaries enable funders and grantees to save time searching for information and
resources. They bridge funders and grantees by helping the former manage and monitor the
latter; they also help grantees to connect with funders by improving their organizational capacity.
Policy-implementing intermediaries also connect and broker partner organizations in a social
service delivery structure. Advocacy intermediaries, which provide platforms for individuals who
share similar interests, promote civic engagement. In summary, the analyses of this dissertation
imply that the functions of nonprofit intermediaries facilitate effective collaboration between
social actors, so they ultimately strengthen the nonprofit sector and society.
106
Contributions and Implications for the Field
This dissertation makes several theoretical contributions to the field of nonprofit and
organizational studies. First, it contributes to the theoretical literature by offering a
comprehensive understanding of nonprofit intermediaries. Previous studies often focused on one
function of nonprofit intermediaries descriptively (Benjamin, 2010; de Souza Briggs, 2003;
Lopez et al., 2005; Shea, 2011; Smith, 2007; Szanton, 2004; Wynn, 2000). However, through
investigating three main functions of nonprofit intermediaries, this dissertation attempted to
provide better understanding about nonprofit intermediaries comprehensively. Second, this
dissertation contributes to the organizational literature by emphasizing the mediating, bridging,
and brokering roles of intermediary organizations. Increases in environmental complexity and
uncertainty require organizations to collaborate with various social members, heightening the
significance of nonprofit intermediaries. However, intermediary organizations have received
little attention in the literature. The findings of this dissertation underscore the importance of
intermediary organizations.
Each chapter of this dissertation also makes theoretical and empirical contributions in
various topic areas. Chapter 2 found that the organizational population of nonprofit management
support organizations (i.e. nonprofit capacity-building intermediaries) has diversified from
sector-wide management organizations assisting a variety of nonprofits in multiple domains, to
nonprofit management support organizations serving a field occupied by nonprofits operating in
a specific service area. Although the diversification of nonprofit management support
organizations is a prominent trend, few studies explained this process. This research contributes
to the nonprofit literature by being the first empirical research to explore two different types of
nonprofit management support organizations and the factors affecting their proliferation from
107
resource partitioning and organizational ecology perspectives.
Chapter 3 found that the organizational form of local workforce investment boards,
whether they are government-controlled nonprofits or governmental agencies, is significantly
associated with the effectiveness of the job training programs of the Workforce Investment Act
Adult program. This research contributes to the public management literature by empirically
investigating the performance differences between nonprofit and government agencies, which
has been a significant research topic in the literature.
Chapter 4 investigated the efficacy of advocacy nonprofit organizations’ social media
use for promoting civic engagement. Despite their wide use, the efficacy of social media use has
rarely been examined empirically in the nonprofit literature due to limitations of data and
methodology. This research provided empirical evidence on the effectiveness of social media for
disseminating information and mobilizing engagement by adopting social media data-mining and
analyzing techniques, so it is expected to contribute to the nonprofit literature both theoretically
and empirically. This research found the low predictive power of the variables that have been
widely used in previous studies; the findings of this research suggest the need for future research
to identify the factors influencing advocacy nonprofits’ ability to engage their stakeholders and
citizens. Moreover, this research is the first empirical study to examine the engagement of non-
followers (i.e. non-constituents) of advocacy nonprofits, and provides concrete evidence that
social media promotes participation not only of constituents but also non-constituents of
advocacy organizations.
Limitations of the Study and Future Research Directions
This dissertation consists of three essays covering three functions of intermediary
nonprofit organizations. Since the chapters have different topics, data, and methods,
108
comprehensive discussions on nonprofit intermediaries are limited. Instead of drawing
synthesized implications about nonprofit intermediaries, this dissertation focused on the
functions of nonprofit intermediaries separately in each chapter. Despite the prevalence of
intermediary organizations, there has been limited information about them. Specially, there is no
clear comprehensive theoretical frameworks about nonprofit intermediaries, so more in-depth
theoretical discussion about them should be conducted in the future.
The chapters of the dissertation also have their own limitations. The chapters mainly rely
on quantitative methods for addressing the research questions, so they are limited to answering
thoughtful questioning about the topics. For example, Chapter 2 empirically explored the
diversification of nonprofit management support organizations. However, the chapter does not
discuss differences in the services provided by generalist and specialist management support
organizations, in practice. Furthermore, this research does not focus on the effectiveness of
nonprofit management support organizations in promoting the managerial capacity of nonprofits.
Future research should be conducted for answering the role of nonprofit management support
organizations in strengthening the sector, in practice.
Chapter 3 provided the empirical evidence that nonprofit policy implementing
intermediaries were more effective in increasing employment and retention rates, but they were
less effective in increasing earnings and earning changes. However, the research does not
investigate why and how organizational forms produce different outcomes. Furthermore, these
findings imply that nonprofit intermediaries may focus on the quantity of job placements instead
of their quality, but there is no clear evidence that this is intentioned. Future research should be
conducted on this issue.
Chapter 4 also empirically explored the efficacy of nonprofit advocacy organizations’
109
use of social media, but future research should investigate what types of messages on social
media are more likely to promote civic engagement. While the chapters of this research offered
initial empirical evidence around their research questions, in-depth qualitative exploration also
needs to be conducted in the future.
110
REFERENCES
Abramson, A. J., & McCarthy, R. (2012). Infrastructure organizations. The state of nonprofit
America.
Aldrich Howard, E., & Ruef, M. (2006). Organizations evolving. Thousands Oaks CA: Sage.
Anderson, K. H., Burkhauser, R. V ., & Raymond, J. E. (1993). The effect of creaming on
placement rates under the Job Training Partnership Act. ILR Review, 46(4), 613-624.
https://doi-org.libproxy2.usc.edu/10.1177/001979399304600402
Archibald, M. E. (2007). An organizational ecology of national self-help/mutual-aid
organizations. Nonprofit and Voluntary Sector Quarterly, 36(4), 598-621. https://doi-
org.libproxy2.usc.edu/10.1177/0899764006297666
Arya, B., & Lin, Z. (2007). Understanding collaboration outcomes from an extended resource-
based view perspective: The roles of organizational characteristics, partner attributes, and
network structures. Journal of management, 33(5), 697-723.
https://doi.org/10.1177/0149206307305561
Backer, T. E., Bleeg, J. E., & Groves, K. (2004). The expanding universe: New directions in
nonprofit capacity building. Alliance for Nonprofit Management.
Bakshy, E., Hofman, J. M., Mason, W. A., & Watts, D. J. (2011, February). Everyone's an
influencer: quantifying influence on twitter. In Proceedings of the fourth ACM international
conference on Web search and data mining (pp. 65-74). ACM. Retrieved January 10, 2018,
from http://snap.stanford.edu/class/cs224w-readings/bakshy11influencers.pdf
Bakshy, E., Rosenn, I., Marlow, C., & Adamic, L. (2012, April). The role of social networks in
information diffusion. In Proceedings of the 21st international conference on World Wide
Web (pp. 519-528). ACM. Retrieved January 10, 2018, from
111
https://arxiv.org/pdf/1201.4145.pdf
Balassiano, K., & Chandler, S. M. (2010). The emerging role of nonprofit associations in
advocacy and public policy: Trends, issues, and prospects. Nonprofit and Voluntary Sector
Quarterly, 39(5), 946-955. https://doi.org/10.1177/0899764009338963
Barnow, B. S., & Gubits, D. B. (2003). Review of recent pilot, demonstration, research, and
evaluation initiatives to assist in the implementation of programs under the Workforce
Investment Act. US Department of Labor, Employment and Training Administration.
Retrieved 09, January, 2018 from
https://wdr.doleta.gov/research/FullText_Documents/ETAOP%202003-
10%20Review%20of%20Recent%20Pilot,%20Demonostration,%20Research,%20and%20E
valuation%20Initiatives.pdf
Barnow, B. S., & King, C. T. (2005). The workforce investment act in eight states. US
Department of Labor, Employment and Training Administration, Nelson A. Rockefeller
Institute of Government. Retrieved January 04, 2018, from
http://raymarshallcenter.org/files/2008/07/2005-02-
the_workforce_investment_act_in_eight_states.pdf
Baum, J. A., & Rowley, T. J. (2002). Companion to organizations: An introduction. The
Blackwell companion to organizations, 1-34. Retrieved January 2018 from
https://www.researchgate.net/profile/Joel_Baum/publication/265158485_Companion_to_Org
anizations_An_Introduction/links/58cb3fa7aca2727749ed91e5/Companion-to-Organizations-
An-Introduction.pdf
Baum, J. A., & Singh, J. V . (1996). Dynamics of organizational responses to competition. Social
forces, 74(4), 1261-1297. https://doi-org.libproxy2.usc.edu/10.1093/sf/74.4.1261
112
Beaulieu, L. J. (1999). Improving Job Opportunities for Low-Income People: The Hope of the
Workforce Investment Act of 1998. SRDC Information Brief: A Special Series on Welfare
Reform in the South. Retrieved January 04, 2018, from https://eric-ed-
gov.libproxy1.usc.edu/?id=ED433164
Becker, C. J. (2010). Self-Determination, Accountability Mechanisms, and Quasi-Governmental
Status of Business Improvement Districts in the United States. Public Performance &
Management Review, 33(3), 413-435. https://doi.org/10.2753/PMR1530-9576330306
Bell, S. H., & Orr, L. L. (2002). Screening (and creaming?) applicants to job training programs:
the AFDC homemaker–home health aide demonstrations. Labour Economics, 9(2), 279-301.
https://doi.org/10.1016/S0927-5371(02)00006-4
Benjamin, L. M. (2010). Mediating accountability: How nonprofit funding intermediaries use
performance measurement and why it matters for governance. Public Performance &
Management Review, 33(4), 594-618. DOI: 10.2753/PMR1530-9576330404
Besharov, D. J., & Cottingham, P. H. (2011). The Workforce Investment Act: Implementation
Experiences and Evaluation Findings. WE Upjohn Institute.
Bielefeld, W. (1994). What affects nonprofit survival?. Nonprofit management and
leadership, 5(1), 19-36. http://dx.doi.org/10.1002/nml.4130050104
Bills, D. B. (2003). Credentials, signals, and screens: Explaining the relationship between
schooling and job assignment. Review of educational research, 73(4), 441-449. https://doi-
org.libproxy1.usc.edu/10.3102/00346543073004441
Blau, F. D., & Kahn, L. M. (2007). The gender pay gap have women gone as far as they
can?. The Academy of Management Perspectives, 21(1), 7-23.
http://www.jstor.org/stable/4166284
113
Bonzanini, M. (2016). Mastering social media mining with Python. Packt Publishing Ltd.
Boris, Elizabeth T. (2001). Next Steps for Building Capacity in Nonprofit Organizations. In
Carol De Vita & Cory Fleming (Eds.), Building Capacity in Nonprofit Organizations.
Washington, D.C.: Urban Institute Press.
Bradley, D. H. (2013). The Workforce Investment Act and the one-stop delivery system.
Washington, DC: Congressional Research Service. Retrieved January 08, 2018 from
http://digitalcommons.ilr.cornell.edu/key_workplace/1058
Briones, R. L., Kuch, B., Liu, B. F., & Jin, Y . (2011). Keeping up with the digital age: How the
American Red Cross uses social media to build relationships. Public relations review, 37(1),
37-43. https://doi.org/10.1016/j.pubrev.2010.12.006
Brown, L. D., & Kalegaonkar, A. (2002). Support organizations and the evolution of the NGO
sector. Nonprofit and Voluntary Sector Quarterly, 31(2), 231-258. https://doi-
org.libproxy2.usc.edu/10.1177/0899764002312004
Brown, T. L., & Potoski, M. (2003). Transaction costs and institutional explanations for
government service production decisions. Journal of Public Administration research and
theory, 13(4), 441-468. https://doi.org/10.1093/jpart/mug030
Buck, M. L. (2002). Charting New Territory: Early Implementation of the Workforce Investment
Act. Field Report Series. Retrieved 09, January 2018 from https://files-eric-ed-
gov.libproxy2.usc.edu/fulltext/ED463444.pdf
Cairns, B., Harris, M., & Young, P. (2005). Building the capacity of the voluntary nonprofit
sector: Challenges of theory and practice. Intl Journal of Public Administration, 28(9-10),
869-885. https://doi-org.libproxy1.usc.edu/10.1081/PAD-200067377
Carboni, J. L., & Maxwell, S. P. (2015). Effective social media engagement for nonprofits: What
114
matters?. Journal of Public and nonprofit Affairs, 1(1), 18-28.
http://dx.doi.org/10.20899/jpna.1.1.18-28
Carroll, G. R. (1985). Concentration and specialization: Dynamics of niche width in populations
of organizations. American journal of sociology, 90(6), 1262-1283.
https://doi.org/10.1086/228210
Carroll, G. R., & Swaminathan, A. (2000). Why the microbrewery movement? Organizational
dynamics of resource partitioning in the US brewing industry. American Journal of
Sociology, 106(3), 715-762. https://doi.org/10.1086/318962
Carroll, G. R., Dobrev, S. D., & Swaminathan, A. (2002). Organizational processes of resource
partitioning. Research in organizational behavior, 24, 1-40. https://doi.org/10.1016/S0191-
3085(02)24002-2
Carty, V . (2010). New information communication technologies and grassroots
mobilization. Information, Communication & Society, 13(2), 155-173.
https://doi.org/10.1080/13691180902915658
Cebrien, E. (2012). The Workforce Investment Act (Economic Issues, Problems and
Perspectives). New York: Nova Science.
Chrisinger, C. K. (2013). Earnings Progression and the Workforce Investment Act: Evidence
from Washington State. Industrial Relations: A Journal of Economy and Society, 52(4), 853-
877. http://dx.doi.org/10.1111/irel.12038
Cohen, A., Timmons, J. C., & Fesko, S. L. (2005). The Workforce Investment Act: How policy
conflict and policy ambiguity affect implementation. Journal of Disability Policy
Studies, 15(4), 221-230. https://doi-org.libproxy1.usc.edu/10.1177/10442073050150040401
Connolly, P., York, P., Munemitsu, S., Ruiz-Healy, C., Sherman, A., & Trebb, C. (2003). Building
115
the capacity of capacity builders. New York, NY: TCC Group. Retrieved February 10, 2018
from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.180.3338&rep=rep1&type=pdf
Connor, J. A., Kadel‐Taras, S., & Vinokur‐Kaplan, D. (1999). The role of nonprofit management
support organizations in sustaining community collaborations. Nonprofit Management and
Leadership, 10(2), 127-136. http://dx.doi.org/10.1002/nml.10202
Courty, P., & Marschke, G. (2007). Making government accountable: Lessons from a federal job
training program. Public Administration Review, 67(5), 904-916.
http://dx.doi.org/10.1111/j.1540-6210.2007.00777.x
Curtis, L., Edwards, C., Fraser, K. L., Gudelsky, S., Holmquist, J., Thornton, K., & Sweetser, K.
D. (2010). Adoption of social media for public relations by nonprofit organizations. Public
Relations Review, 36(1), 90-92. https://doi.org/10.1016/j.pubrev.2009.10.003
Child, C. D., & Grønbjerg, K. A. (2007). Nonprofit advocacy organizations: Their characteristics
and activities. Social Science Quarterly, 88(1), 259-281. https://doi-
org.libproxy1.usc.edu/10.1111/j.1540-6237.2007.00457.x
D'Amico, R. & Salzman, J. (2004). Implementation Issues in Delivering Training Services to
Adults Under WIA. In Job Training Policy in the United States, Christopher J. O'Leary,
Robert A. Straits, and Stephen A. Wandner, eds. Kalamazoo, MI: W.E. Upjohn Institute, pp.
101-134.https://doi.org/10.17848/9781417549993.ch4
D’Amico, R., Dunham, K., Henderson-Frakes, J., Kogan, D., Koller, V ., Mack, M., ... &
Weissbein, D. (2004). The Workforce Investment Act after five years: Results from the
national evaluation of the implementation of WIA. ETA Occasional Paper, (2004-05).
Retrieved January 14, 2018 from https://www.doleta.gov/reports/searcheta/occ/papers/SPR-
WIA_Final_Report.pdf
116
Darity, W. A., & Mason, P. L. (1998). Evidence on discrimination in employment: Codes of
color, codes of gender. Journal of Economic Perspectives, 12(2), 63-90.
https://doi/10.1267/jep.12.2.63
Davis, G. F., & Cobb, J. (2010). Resource Dependence Theory: Past and future. Research in the
Sociology of Organizations, 28 21-42. http://dx.doi.org/10.1108/S0733-
558X(2010)0000028006
de Souza Briggs, Xavier. (2003). Working the middle: roles and challenges of intermediaries.
Cambridge, MA: The Community Problem-Solving Project. Retrieved in January 2018 from
http://web.mit.edu/cpsproject/images/artsci_tool_intermedaries_web_0603.pdf
De Tray, D. (1982). Veteran status as a screening device. The American Economic Review, 72(1),
133-142. http://www.jstor.org.libproxy1.usc.edu/stable/1808580
De Vita, C. J. & Fleming, C. (Eds.). (2001). Building capacity in nonprofit organizations. Urban
Institute. Retrieved January 20, 2018 from
http://research.urban.org/uploadedPDF/building_capacity.PDF
Decker, P. T., & Berk, J. A. (2011). Ten years of the Workforce Investment Act (WIA):
Interpreting the research on WIA and related programs. Journal of Policy Analysis and
Management, 30(4), 906-926. http://dx.doi.org/10.1002/pam.20597
Devin, B. L., & Lane, A. B. (2014). Communicating engagement in corporate social
responsibility: A meta-level construal of engagement. Journal of Public Relations
Research, 26(5), 436-454. http://dx.doi.org/10.1080/1062726X.2014.956104
Diani, M. (2000). Social movement networks virtual and real. Information, Communication &
Society, 3(3), 386-401. https://doi-org.libproxy1.usc.edu/10.1080/13691180051033333
Dobrev, S. D., & Kim, T. Y . (2006). Positioning among organizations in a population: Moves
117
between market segments and the evolution of industry structure. Administrative Science
Quarterly, 51(2), 230-261. https://doi-org.libproxy1.usc.edu/10.2189/asqu.51.2.230
Eaton, M. (2010). Manufacturing community in an online activist organization: The rhetoric of
MoveOn. org's e-mails. Information, Communication & Society, 13(2), 174-192. https://doi-
org.libproxy1.usc.edu/10.1080/13691180902890125
Fisher, D. R., & Boekkooi, M. (2010). Mobilizing Friends and Strangers: Understanding the role
of the Internet in the Step It Up day of action. Information, Communication & Society, 13(2),
193-208. https://doi-org.libproxy1.usc.edu/10.1080/13691180902878385
Fleissig, A. R. (2014). Return on investment from training programs and intensive
services. Atlantic Economic Journal, 42(1), 39-51. https://doi-
org.libproxy1.usc.edu/10.1007/s11293-013-9394-y
Galaskiewicz, J., & Bielefeld, W. (1998). Nonprofit organizations in an age of uncertainty: A
study of organizational change. Transaction Publishers.
Gálvez-Rodríguez, M. D. M., Caba-Pérez, C., & López-Godoy, M. (2014). Drivers for the
proactive online disclosure of information in the NGO sector: The Colombian case. Online
Information Review, 38(6), 769-787. https://doi.org/10.1108/OIR-05-2014-0113.
Gálvez-Rodríguez, M. D. M., Caba-Pérez, C., & López-Godoy, M. (2016). Drivers of Twitter as
a strategic communication tool for non-profit organizations. Internet Research, 26(5), 1052-
1071. https://doi.org/10.1108/IntR-07-2014-0188
Gilmer, S., & Hughes, C. (2013). Human Resource Capacity Building and Nonprofit
Organizations: Addressing Potential Leadership Shortages. Journal of the North American
Management Society, 7(1), 119-127. Retrieved February 10, 2018 from
https://www.researchgate.net/publication/282660240_Human_resource_capacity_building_a
118
nd_nonprofit_organizations
Gladwell, M. (2010) Small change: why the revolution will not be tweeted, The New Yorker, 4
October. Retrieved February 13, 2018, Retrieved from
https://sites.tufts.edu/mitznereng1fall2017/files/2017/09/Small-Change-The-New-Yorker.pdf
Goel, S., Anderson, A., Hofman, J., & Watts, D. J. (2015). The structural virality of online
diffusion. Management Science, 62(1), 180-196. https://doi-
org.libproxy1.usc.edu/10.1287/mnsc.2015.2158
Golbeck, J., Grimes, J. M., & Rogers, A. (2010). Twitter use by the US Congress. Journal of the
Association for Information Science and Technology, 61(8), 1612-1621. https://doi-org
/10.1002/asi.21344
Grønbjerg, K. A., & Paarlberg, L. (2001). Community variations in the size and scope of the
nonprofit sector: Theory and preliminary findings. Nonprofit and Voluntary Sector
Quarterly, 30(4), 684-706. https://doi.org/10.1177/0899764001304004
Guidry, J., D. Waters, R., & D. Saxton, G. (2014). Moving social marketing beyond personal
change to social change: Strategically using Twitter to mobilize supporters into vocal
advocates. Journal of Social Marketing, 4(3), 240-260. https://doi.org/10.1108/JSOCM-02-
2014-0014
Guo, C., & Acar, M. (2005). Understanding collaboration among nonprofit organizations:
Combining resource dependency, institutional, and network perspectives. Nonprofit and
voluntary sector quarterly, 34(3), 340-361. https://doi.org/10.1177/0899764005275411
Guo, C., & Saxton, G. D. (2014). Tweeting social change: How social media are changing
nonprofit advocacy. Nonprofit and Voluntary Sector Quarterly, 43(1), 57-79. https://doi-
org.libproxy1.usc.edu/10.1177/0899764012471585
119
Hager, M. A., Galaskiewicz, J., & Larson, J. A. (2004). Structural embeddedness and the liability
of newness among nonprofit organizations. Public Management Review, 6(2), 159-188.
https://doi-org.libproxy2.usc.edu/10.1080/1471903042000189083
Hannan, M. T., & Freeman, J. (1977). The population ecology of organizations. American
journal of sociology, 82(5), 929-964. https://doi.org/10.1086/226424
Hannan, M. T., & Freeman, J. (1993). Organizational ecology. Harvard university press.
Hansmann, H. (1987). Economic theories of nonprofit organization. The nonprofit sector: A
research handbook, 1, 27-42. Yale University Press.
Hara, N., & Estrada, Z. (2005). Analyzing the mobilization of grassroots activities via the
internet: a case study. Journal of Information Science, 31(6), 503-514.
https://doi.org/10.1177/0165551505057013
Haro-de-Rosario, A., Sáez-Martín, A., & del Carmen Caba-Pérez, M. (2016). Using social media
to enhance citizen engagement with local government: Twitter or Facebook?. New Media &
Society, 1461444816645652. https://doi.org/10.1177/1461444816645652
Heckman, J. J., & Smith, J. A. (2004). The determinants of participation in a social program:
Evidence from a prototypical job training program. Journal of Labor Economics, 22(2), 243-
298. https://doi.org/10.1086/381250
Hegewisch, A., & Luyri, H. (2010). The Workforce Investment Act and Women’s Progress: Does
WIA Funded Training Reinforce Sex Segregation in the Labor Market and the Gender Wage
Gap?. Briefing Paper C, 72. Retrieved January 10, 2017 from
https://www.researchgate.net/publication/277285952_The_Workforce_Investment_Act_and_
Women%27s_Progress_Does_WIA_Funded_Training_Reinforce_Sex_Segregation_in_the_
Labor_Market_and_the_Gender_Wage_Gap
120
Heinrich, C. J. (2007). False or fitting recognition? The use of high performance bonuses in
motivating organizational achievements. Journal of policy analysis and management, 26(2),
281-304. http://dx.doi.org/10.1002/pam.20244
Heinrich, C. J., & Lynn Jr, L. E. (2000). CHAPTER Governance and Performance: 3 The
Influence of Program Structure and Management on Job Training Partnership Act (JTPA)
Program Outcomes. Governance and performance: New perspectives, 68. Retrieved 09,
January 2018 from
https://www.researchgate.net/publication/23551070_Governance_And_Performance_The_In
fluence_of_Program_Structure_and_Management_on_Job_Training_Partnership_Act_JTPA
_Program_Outcomes
Heinrich, C., Mueser, P. R., Troske, K. R., Jeon, K. S., & Kahvecioglu, D. (2009). New estimates
of public employment and training program net impacts: A nonexperimental evaluation of the
Workforce Investment Act program. Retrieved January 08, 2018 from
https://mospace.umsystem.edu/xmlui/bitstream/handle/10355/2389/NewEstimatesPublicEmp
loyment.pdf?sequence=1&isAllowed=y
Hodge, M. M., & Piccolo, R. F. (2005). Funding source, board involvement techniques, and
financial vulnerability in nonprofit organizations: A test of resource dependence. Nonprofit
Management and Leadership, 16(2), 171-190. https://doi-
org.libproxy1.usc.edu/10.1002/nml.99
Holcomb, P., & Barnow, B. S. (2004). Serving People with Disabilities through the Workforce
Investment Act’s One-Stop Career Centers. Washington, DC and Baltimore, MD. The Urban
Institute and John Hopkins University. Retrieved August, 14, 2006. Retrieved 09 January
2018 from https://www.issuelab.org/resources/1549/1549.pdf
121
Hollenbeck, K. (2009). Workforce Investment Act (WIA) net impact estimates and rates of
return. Retrieved January 16, 2018,
http://research.upjohn.org/cgi/viewcontent.cgi?article=1005&context=confpapers
Hollenbeck, K., Schroeder, D., King, C. T., & Huang, W. J. (2005). Net Impact Estimates for
Services Provided through the Workforce Investment Act.” Employment and Training
Administration Occasional Paper (ETAOP. Retrieved 17, January 2018 from
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.596.2344
Holzer, H. J., Raphael, S., & Stoll, M. A. (2003). Employment barriers facing ex-
offenders. Urban Institute Reentry Roundtable, 1-23.
http://webarchive.urban.org/UploadedPDF/410855_holzer.pdf
Honig, M. I. (2004). The new middle management: Intermediary organizations in education
policy implementation. Educational Evaluation and Policy Analysis, 26(1), 65-87.
https://doi.org/10.3102/01623737026001065
Howells, J. (2006). Intermediation and the role of intermediaries in innovation. Research
policy, 35(5), 715-728. https://doi.org/10.1016/j.respol.2006.03.005
Huberman, B. A., Romero, D. M., & Wu, F. (2008). Social networks that matter: Twitter under
the microscope. arXiv preprint arXiv:0812.1045. Retrieved in January 2018 from
https://arxiv.org/pdf/0812.1045.pdf
Hwang, H., & Powell, W. W. (2009). The rationalization of charity: The influences of
professionalism in the nonprofit sector. Administrative Science Quarterly, 54(2), 268-298.
https://doi-org.libproxy1.usc.edu/10.2189/asqu.2009.54.2.268
Javar, J. O., and Wandner, S. A. (2004). The Use of Service Providers and Brokers/Consultants in
Employment and Training Programs. In Job Training Policy in the United States, Christopher
122
J. O'Leary, Robert A. Straits, and Stephen A. Wandner, eds. Kalamazoo, MI: W.E. Upjohn
Institute, pp. 135-176. https://doi.org/10.17848/9781417549993.ch5
Johansen, M., & LeRoux, K. (2013). Managerial networking in nonprofit organizations: The
impact of networking on organizational and advocacy effectiveness. Public Administration
Review, 73(2), 355-363. https://doi.org/10.1111/puar.12017
Kent, M. L., & Taylor, M. (1998). Building dialogic relationships through the World Wide
Web. Public relations review, 24(3), 321-334. https://doi.org/10.1016/S0363-8111(99)80143-
X
Kim, D., Chun, H., Kwak, Y ., & Nam, Y . (2014). The employment of dialogic principles in
website, Facebook, and Twitter platforms of environmental nonprofit organizations. Social
science computer review, 32(5), 590-605. https://doi.org/10.1177/0894439314525752
Klein, H. K. (1999). Tocqueville in cyberspace: Using the Internet for citizen associations. The
Information Society, 15(4), 213-220. https://doi.org/10.1080/019722499128376
Laer, J. V ., & Aelst, P. V . (2010). Internet and social movement action repertoires. Information,
communication & society, 13(8), 1146-1171. https://doi-
org.libproxy1.usc.edu/10.1080/13691181003628307
LeRoux, K. (2007). Nonprofits as civic intermediaries: The role of community-based
organizations in promoting political participation. Urban Affairs Review, 42(3), 410-422.
https://doi.org/10.1177/1078087406292257
Liu, Z., Liu, L., & Li, H. (2012). Determinants of information retweeting in
microblogging. Internet Research, 22(4), 443-466.
https://doi.org/10.1108/10662241211250980
Lopez, M. E., Kreider, H., & Coffman, J. (2005). Intermediary organizations as capacity builders
123
in family educational involvement. Urban Education, 40(1), 78-105.
https://doi.org/10.1177/0042085904270375
Lovejoy, K., & Saxton, G. D. (2012). Information, community, and action: How nonprofit
organizations use social media. Journal of Computer-Mediated Communication, 17(3), 337-
353. https://doi.org/10.1111/j.1083-6101.2012.01576.x
Lovejoy, K., Waters, R. D., & Saxton, G. D. (2012). Engaging stakeholders through Twitter:
How nonprofit organizations are getting more out of 140 characters or less. Public Relations
Review, 38(2), 313-318. https://doi.org/10.1016/j.pubrev.2012.01.005
MacIndoe, H., & Whalen, R. (2013). Specialists, generalists, and policy advocacy by charitable
nonprofit organizations. Journal of Sociology and Social Welfare, 40(2), 119.
http://scholarworks.wmich.edu/jssw/vol40/iss2/8
Marco, B., Almandsmith, S., & Hague, M. (2003). Creating Partnerships for Workforce
Investment: How Services are Provided Under WIA: Revised Final Report for"
Understanding the Role of Intermediaries Under WIA". Berkeley Policy Associates.
Mead, J., & Warren, K. (2016, September). Quasi-Governmental Organizations at the Local
Level: Publicly-Appointed Directors Leading Nonprofit Organizations. Nonprofit Policy
Forum, 7(3), 289-309. https://doi.org/10.1515/npf-2014-0044
Men, L. R., & Tsai, W. H. S. (2014). Perceptual, attitudinal, and behavioral outcomes of
organization–public engagement on corporate social networking sites. Journal of Public
Relations Research, 26(5), 417-435. http://dx.doi.org/10.1080/1062726X.2014.951047
Mendel, S. C., & Brudney, J. L. (2012). Putting the NP in PPP: The role of nonprofit
organizations in public-private partnerships. Public Performance & Management
Review, 35(4), 617-642. https://doi.org/10.2753/PMR1530-9576350403
124
Millesen, J. L., Carman, J. G., & Bies, A. L. (2010). Why engage?. Nonprofit Management and
Leadership, 21(1), 5-20. http://dx.doi.org/10.1002/nml.20009
Milward, H. B., & Provan, K. G. (2000). Governing the hollow state. Journal of Public
Administration Research and Theory, 10(2), 359-380.
https://doi.org/10.1093/oxfordjournals.jpart.a024273
Minzner, A., Klerman, J. A., Markovitz, C. E., & Fink, B. (2014). The impact of capacity-
building programs on nonprofits: A random assignment evaluation. Nonprofit and Voluntary
Sector Quarterly, 43(3), 547-569. https://doi-
org.libproxy1.usc.edu/10.1177/0899764013491013
Monge, P. R., & Contractor, N. S. (2003). Theories of communication networks. Oxford
University Press, USA.
Moore, R. W., & Gorman, P. C. (2009). The impact of training and demographics in WIA
program performance: A statistical analysis. Human Resource Development Quarterly, 20(4),
381-396. http://dx.doi.org/10.1002/hrdq.20029
Mosley, J. E. (2014). Collaboration, public-private intermediary organizations, and the
transformation of advocacy in the field of homeless services. The American Review of Public
Administration, 44(3), 291-308. https://doi.org/10.1177/0275074012465889
Moss, T. (2009). Intermediaries and the governance of sociotechnical networks in
transition. Environment and Planning A, 41(6), 1480-1495. https://doi-
org.libproxy2.usc.edu/10.1068/a4116
Murnane, R. J., & Willett, J. B. (2010). Methods matter: Improving causal inference in
educational and social science research. Oxford University Press.
Nah, S., & Saxton, G. D. (2013). Modeling the adoption and use of social media by nonprofit
125
organizations. New Media & Society, 15(2), 294-313. https://doi-
org.libproxy2.usc.edu/10.1177/1461444812452411
NCCS, (2006), Guide to Using NCCS Data, Retrieved January 10, 2018 from
http://nccsdataweb.urban.org/kbfiles/468/NCCS-data-guide-2006c.pdf
NCCS, (2007), National Taxonomy of Exempt Entities-core Codes, Retrieved January 10, 2018
from http://nccsdataweb.urban.org/kbfiles/322/NTEE-CC-Manual-2007a.pdf
Nilsen, S. R. (2002). Workforce Investment Act: Improvements Needed in Performance
Measures To Provide a More Accurate Picture of WIA's Effectiveness. Report to
Congressional Requesters. Retrieved January 08, 2018 from https://eric-ed-
gov.libproxy2.usc.edu/?id=ED463448
O'Shea, D. P., & King, C. T. (2001). The Workforce Investment Act of 1998: Restructuring
workforce development initiatives in states and localities (No. 12). Albany, NY: Nelson A.
Rockefeller Institute of Government.
http://raymarshallcenter.org/files/2001/04/wia_final_2001.pdf
Obar, J. A., Zube, P., & Lampe, C. (2012). Advocacy 2.0: An analysis of how advocacy groups in
the United States perceive and use social media as tools for facilitating civic engagement and
collective action. Journal of information policy, 2, 1-25. doi:10.5325/jinfopoli.2.2012.0001
Pager, D., Bonikowski, B., & Western, B. (2009). Discrimination in a low-wage labor market: A
field experiment. American sociological review, 74(5), 777-799. https://doi-
org.libproxy1.usc.edu/10.1177/000312240907400505
Park, H., Reber, B. H., & Chon, M. G. (2016). Tweeting as health communication: health
organizations’ use of twitter for health promotion and public engagement. Journal of health
communication, 21(2), 188-198. https://doi.org/10.1080/10810730.2015.1058435
126
Pfeffer, J., & Salancik, G. R. (2003). The external control of organizations: A resource
dependence perspective. Stanford University Press. Chapter 1 & 3.
Reis, S., Negro, G., Sorenson, O., Perretti, F., & Lomi, A. (2012). Resource partitioning
revisited: evidence from Italian television broadcasting. Industrial and Corporate
Change, 22(2), 459-487. https://doi-org.libproxy2.usc.edu/10.1093/icc/dts020
Renz, D. O. (2008). The US nonprofit infrastructure mapped. Nonprofit Quarterly, 15(4), 17-21.
Saffer, A. J., Sommerfeldt, E. J., & Taylor, M. (2013). The effects of organizational Twitter
interactivity on organization–public relationships. Public Relations Review, 39(3), 213-215.
https://doi.org/10.1016/j.pubrev.2013.02.005
Saxton, G. D., & Wang, L. (2014). The social network effect: The determinants of giving through
social media. Nonprofit and Voluntary Sector Quarterly, 43(5), 850-868. https://doi-
org.libproxy1.usc.edu/10.1177/0899764013485159
Schmid, H., Bar, M., & Nirel, R. (2008). Advocacy activities in nonprofit human service
organizations: Implications for policy. Nonprofit and Voluntary Sector Quarterly, 37(4), 581-
602. https://doi-org.libproxy1.usc.edu/10.1177/0899764007312666
Scott, W. R. (2004). Reflections on a half-century of organizational sociology. Annu. Rev.
Sociol., 30, 1-21. https://doi.org/10.1146/annurev.soc.30.012703.110644
Shea, J. (2011). Taking Nonprofit Intermediaries Seriously: A Middle‐Range Theory for
Implementation Research. Public Administration Review, 71(1), 57-66.
https://doi.org/10.1111/j.1540-6210.2010.02306.x
Singh, J. V ., & Lumsden, C. J. (1990). Theory and research in organizational ecology. Annual
review of sociology, 16(1), 161-195. https://doi-
org.libproxy2.usc.edu/10.1146/annurev.so.16.080190.001113
127
Sloan, P. (2009). Redefining Stakeholder Engagement: From Control to Collaboration. Journal
of Corporate Citizenship, (36). http://www.jstor.org.libproxy1.usc.edu/stable/jcorpciti.36.25
Smith, B. C. (2008). The sources and uses of funds for community development financial
institutions: The role of the nonprofit intermediary. Nonprofit and Voluntary Sector
Quarterly, 37(1), 19-38. https://doi.org/10.1177/0899764007301282
Soule, S. A., & King, B. G. (2008). Competition and resource partitioning in three social
movement industries. American Journal of Sociology, 113(6), 1568-1610.
https://doi.org/10.1086/587152
Sternlieb, F., Bixler, R. P., & Huber-Stearns, H. (2013). A question of fit: reflections on
boundaries, organizations and social–ecological systems. Journal of environmental
management, 130, 117-125. https://doi.org/10.1016/j.jenvman.2013.08.053
Szanton, P. (2004). Toward more effective use of intermediaries. The Foundation Center.
Retrieved in January 2018 from
https://folio.iupui.edu/bitstream/handle/10244/16/TowardMoreEffectiveUse2004.pdf?sequen
ce=1
Taylor, P. E., & Walker, A. (1994). The ageing workforce: Employers' attitudes towards older
people. Work, employment and society, 8(4), 569-591. https://doi-
org.libproxy1.usc.edu/10.1177/095001709484005
TCC Group. (2010). Fortifying LA’s nonprofit organizations: Capacity-building needs and
services in Los Angeles County (A study commissioned by the Weingart
Foundation). Retrieved from the Weingart Foundation website http://www. weingartfnd.
org/files/Capacity-Report-Final. pdf.
Tucker, D. J., Singh, J. V ., & Meinhard, A. G. (1990). Organizational form, population dynamics,
128
and institutional changes: The founding patterns of voluntary organizations. Academy of
Management Journal, 33(1), 151-178. https://doi.org/10.2307/256356
Turner, M. A., Fix, M., & Struyk, R. J. (1991). Opportunities denied, opportunities diminished:
Racial discrimination in hiring. The Urban Institute.
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(2), 211-
235. https://doi-org.libproxy2.usc.edu/10.1177/0899764003032002003
Van der Meulen, B., Nedeva, M., & Braun, D. (2005, October). Intermediaries organisation and
processes: theory and research issues. In PRIME Workshop (V ol. 6, p. e7). Retrieved January
2018 from
https://www.researchgate.net/profile/Barend_Van_Der_Meulen2/publication/254552743_Inte
rmediaries_Organisation_and_Processes_theory_and_research_issues/links/55d6cb4608aed6
a199a4ec60/Intermediaries-Organisation-and-Processes-theory-and-research-issues.pdf
Van Slyke, D. M. (2003). The mythology of privatization in contracting for social
services. Public Administration Review, 63(3), 296-315. http://dx.doi.org/10.1111/1540-
6210.00291
Van Slyke, D. M., & Roch, C. H. (2004). What do they know, and whom do they hold
accountable? Citizens in the government–nonprofit contracting relationship. Journal of
Public Administration Research and Theory, 14(2), 191-209. https://doi-
org.libproxy1.usc.edu/10.1093/jopart/muh013
Van Thiel, S. (2004). Trends in the public sector: why politicians prefer quasi-autonomous
organizations. Journal of Theoretical Politics, 16(2), 175-201. https://doi-
org.libproxy1.usc.edu/10.1177/0951629804041120
129
Verbruggen, S., Christiaens, J., & Milis, K. (2011). Can resource dependence and coercive
isomorphism explain nonprofit organizations’ compliance with reporting standards?.
Nonprofit and Voluntary Sector Quarterly, 40(1), 5-32. https://doi-
org.libproxy1.usc.edu/10.1177/0899764009355061
Wagner, M. M., & Blackorby, J. (1996). Transition from high school to work or college: How
special education students fare. The Future of Children, 103-120.
http://www.jstor.org.libproxy1.usc.edu/stable/1602496
Wandner, S. (2002). Targeting Employment Services under the Workforce Investment
Act. Targeting Employment Services, Kalamazoo, Michigan: WE Upjohn Institute, 1-25.
Retrieved January 04, 2018, from
http://research.upjohn.org/cgi/viewcontent.cgi?article=1186&context=up_bookchapters
Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications (V ol. 8).
Cambridge university press.
Waters, R. (2007). Nonprofit organizations' use of the internet: A content analysis of
communication trends on the internet sites of the philanthropy 400. Nonprofit Management &
Leadership., 18(1), 59-76. https://doi-org.libproxy1.usc.edu/10.1002/nml.171
Waters, R. D., & Jamal, J. Y . (2011). Tweet, tweet, tweet: A content analysis of nonprofit
organizations’ Twitter updates. Public Relations Review, 37(3), 321-324.
https://doi.org/10.1016/j.pubrev.2011.03.002
Waters, R. D., Burnett, E., Lamm, A., & Lucas, J. (2009). Engaging stakeholders through social
networking: How nonprofit organizations are using Facebook. Public relations review, 35(2),
102-106. https://doi.org/10.1016/j.pubrev.2009.01.006
Watkins, B. A. (2017). Experimenting with dialogue on Twitter: An examination of the influence
130
of the dialogic principles on engagement, interaction, and attitude. Public Relations
Review, 43(1), 163-171. https://doi.org/10.1016/j.pubrev.2016.07.002
Weisbrod, B. A. (1998). Institutional form and organizational behavior. Private action and the
public good, 69-84. Yale University Press.
Weiss, A. (1995). Human capital vs. signalling explanations of wages. Journal of Economic
perspectives, 9(4), 133-154. DOI: 10.1257/jep.9.4.133
Wholey, D. R., & Brittain, J. W. (1986). Organizational ecology: Findings and
implications. Academy of Management Review, 11(3), 513-533. DOI: 10.2307/258308
Williamson, O. E. (1981). The economics of organization: The transaction cost approach.
American journal of sociology, 548-577. https://doi.org/10.1086/227496
Wimberley, T., & Rubens, A. (2002). Management support organizations and university
partnership in nonprofit education, training and consultation. The Social Science
Journal, 39(1), 129-136. https://doi-org.libproxy2.usc.edu/10.1016/S0362-3319(01)00168-9
Winch, G. M., & Courtney, R. (2007). The organization of innovation brokers: An international
review. Technology analysis & strategic management, 19(6), 747-763.
https://doi.org/10.1080/09537320701711223
Workforce Innovation and Opportunity Act, Public Law. 113-128 (2014).
Workforce Investment Act; Final rules, 20 CFR Part 652; Part 660-671. (2000). Retrieved
January 04, 2018, from https://www.doleta.gov/regs/statutes/finalrule.htm
Wrigley, H. S., Richer, E., Martinson, K., Kubo, H., & Strawn, J. (2003). The Language of
Opportunity: Expanding Employment Prospects for Adults with Limited English Skills.
Center for Law and Social Policy. Retrieved January 2018 from https://files-eric-ed-
gov.libproxy1.usc.edu/fulltext/ED481481.pdf
131
Wynn, J. R. (2000). The Role of Local Intermediary Organizations in the Youth Development
Field: Summary. Chapin Hall Center for Children at the University of Chicago.
Xiong, F., Liu, Y ., Zhang, Z. J., Zhu, J., & Zhang, Y . (2012). An information diffusion model
based on retweeting mechanism for online social media. Physics Letters A, 376(30-31), 2103-
2108. https://doi.org/10.1016/j.physleta.2012.05.021
Yeon, H. M., Choi, Y ., & Kiousis, S. (2007). Interactive communication features on nonprofit
organizations' webpages for the practice of excellence in public relations. Journal of Website
Promotion, 1(4), 61-83. https://doi-org.libproxy1.usc.edu/10.1300/J238v01n04_06
Zhu, J., Xiong, F., Piao, D., Liu, Y ., & Zhang, Y . (2011, October). Statistically modeling the
effectiveness of disaster information in social media. In Global Humanitarian Technology
Conference (GHTC), 2011 IEEE (pp. 431-436). IEEE. 10.1109/GHTC.2011.48
Abstract (if available)
Abstract
The purpose of this dissertation is to develop a better understanding of intermediary nonprofit organizations and their roles in supporting the sector and society. Intermediary organizations are middlemen that connect various social actors and help them collaborate in an effective and efficient manner in order to solve intense social, political, and economic problems in our society. Intermediaries are especially significant in the nonprofit sector where many organizations have limited organizational capacity and resources. Despite their growing importance, nonprofit intermediaries remain invisible in the literature. Filling that gap, this dissertation consists of three essays that cover nonprofit intermediaries’ three main functions: capacity building, coordinating policy implementing networks, and mobilizing political networks. ❧ The first study aims to investigate the growth and evolution of nonprofit management support organizations and analyze the factors affecting their diversification. Nonprofit management support organizations provide capacity-building services for other nonprofits and play a contributing role in improving the sector’s performance. By analyzing panel data on the population of these organizations from 1990 to 2013, this research found the number of nonprofit management support organizations has grown dramatically during the last 20 years, as the demands for capacity building services have increased. This research determined that the organizational population of nonprofit management support organizations has evolved from sector-wide management organizations assisting a variety of nonprofits in multiple domains to management support organizations serving nonprofits operating in a specific service area. The empirical analysis suggests that high competition between nonprofit management organizations hinders their further growth. ❧ The second study explores the function of nonprofit intermediaries as coordinators of collaborative policy implementing networks. In particular, this research investigates local workforce investment boards, which coordinate US job assistance programs in local areas, and the efficacy and equity of their varied organizational structures. Specifically, these boards exist as two types: government-controlled nonprofit organizations and governmental agencies. By analyzing the client data of the US Workforce Investment Act Adult program in 17 states, this research found that the organizational form of local workforce investment boards is significantly associated with program outcomes. The empirical analysis indicates that nonprofit workforce investment boards were more effective in increasing their clients’ employment but less effective in growing their earnings. This implies that nonprofit boards may focus on the quantity, rather than the quality, of job placements, demonstrating the adverse effects of using non-governmental agencies to deliver public services. ❧ The third study investigates nonprofit advocacy organizations, which serve as political network mobilizers, and the effectiveness of their social media use for mobilizing civic engagement. Looking at how 70 advocacy intermediaries use Twitter, this research found that advocacy intermediaries benefit from using social media
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The institutionalization of nonprofit management: emergence, development, and legitimization
PDF
A study of Chinese environmental NGOs: policy advocacy, managerial networking, and leadership succession
PDF
Institutional contestation, network legitimacy and organizational heterogeneity: interactions between government and environmental nonprofits in South Korea
PDF
Strategy and ideology in nonprofit advocacy organizations
PDF
Looking to the stars: perceptions of celebrity influence in the nonprofit sector
PDF
Governing regional collaboratives: institutional design, management and leadership
PDF
Mission match, reputation, and bureaucratic motivation
PDF
Goal setting and performance in federal agencies
PDF
Why go green? Cities' adoption of local renewable energy policies and urban sustainability certifications
PDF
Three essays on the causes and consequences of China’s governance reforms
PDF
Nonprofit organizations and the environmental policy outcomes: a systematic inquiry into the role of different types of nonprofits to influence the processes and outcomes of the environmental policy
PDF
Governance networks: internal governance structure, boundary setting and effectiveness
PDF
Managers’ roles in supporting employee engagement in Jewish nonprofit organizations
PDF
Fundraising in small health and human service nonprofit organizations: an evaluation study
PDF
Does collaborative R&D policy work? The effect of collaborative R&D policy on innovative activities of firms in Korea
PDF
The risks and rewards of city management: how city managers evaluate the nature of the job and compensation
PDF
Intradepartmental collaboration in the public organizations: implications to practice in an era of resource scarcity and economic uncertainty
PDF
The economy-environment trade-off in China's air pollution policies
PDF
A nonprofit study evaluating board chair onboarding practices for effective governance
PDF
Collaboration: is it worth it? The Magnolia Community Initiative from the perspective of initiative partner participants
Asset Metadata
Creator
Cho, Yusun
(author)
Core Title
The functions of the middleman: how intermediary nonprofit organizations support the sector and society
School
School of Policy, Planning and Development
Degree
Doctor of Philosophy
Degree Program
Public Policy and Management
Publication Date
07/30/2018
Defense Date
05/14/2018
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
intermediary organizations,nonprofit intermediaries,nonprofit management support organizations,OAI-PMH Harvest,Workforce Investment Boards, social media, advocacy organizations
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Robertson, Peter J. (
committee chair
), Painter, Gary D. (
committee member
), Tang, Shui Yan (
committee member
)
Creator Email
cloudfog@hanmail.net,yusuncho@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-41914
Unique identifier
UC11671481
Identifier
etd-ChoYusun-6448.pdf (filename),usctheses-c89-41914 (legacy record id)
Legacy Identifier
etd-ChoYusun-6448.pdf
Dmrecord
41914
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Cho, Yusun
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
intermediary organizations
nonprofit intermediaries
nonprofit management support organizations
Workforce Investment Boards, social media, advocacy organizations