is. 5.4!.
3. .y...

 

. .
hwl.
v.15“...

If .n

 

w:
H
a.
n

HHS...

A: 2.... am

e...) .32.:
:l.

4.0; 0r. .,

.w‘t.'

.1
52...?!
2!! a... ?

mgr
. .

 

1.4-1.1..3.‘ ‘ . 2:5«13‘5. .
V - n v . ov‘igyu: ‘01’.’ )l—n.‘ ‘ it! , >o“ , v‘ggflfifia-3| f.

r. .;E E 23.5.3 cgfifé . gwgg

. . 1o. .5 .3: .9 .1. . .\

 

LIBRARY
Michigan State

5 70/ a 7/7 97’ University

This is to certify that the
thesis entitled

UNDERSTANDING THE BRIDGING ROLES OF
NEIGHBORHOOD-BASED INTERMEDIARY
ORGANIZATIONS: AN EXPLORATORY ANALYSIS OF THE
NEIGHBORHOOD NETWORK CENTER MODEL

presented by
MICHELLE L. BERKOWITZ

has been accepted towards fulfillment
of the requirements for the

degree In PSYCHOLOGY

 

 

 

/,...... 9:10er - gt. W -
Major Professor’s Signature
9/ 3 , 05/

Date

 

MSU is an Affinnative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.

1'0 AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 c:/CI!RC/DateDue.inddop.15

 

.1 ...._. -.

a.

UNDERSTANDING THE BRIDGING ROLES OF NEIGHBORHOOD-BASED
INTERMEDIARY ORGANIZATIONS: AN EXPLORATORY ANALYSIS OF THE
NEIGHBORHOOD NETWORK CENTER MODEL
By

Michelle L. Berkowitz

A THESIS

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

MASTER OF ARTS
Department of Psychology
2005

ABSTRACT
UNDERSTANDING THE BRIDGING ROLES OF NEIGHBORHOOD-BASED
INTERMEDIARY ORGANIZATIONS: AN EXPLORATORY ANALYSIS OF THE
NEIGHBORHOOD NETWORK CENTER MODEL
By
Michelle L. Berkowitz

Neighborhood-based intermediary organizations and related organizational
structures are gaining increased attention for their potential to facilitate both
community change efforts and comprehensive service delivery in local
communities by bridging or “brokering” relationships among various segments of
the community. However, the extent to which such entities actually play these
roles has gone largely unexplored; such work is needed as a basis for building
theory and practice around what makes for successful implementation of these
efforts. In this study, social network analysis techniques were applied towards
understanding the roles Neighborhood Network Centers - one form of
neighborhood-based intermediary or broker organization - play in one community
system. Findings Show that the three Network Centers analyzed showed different
levels, patterns, and modes of brokerage activity. In addition, findings suggest

the differences in how bridging roles unfold given both the capacity of

intermediary organizations and the neighborhoods they serve.

ACKNOWLEDGEMENTS

I would like to formally recognize some of the many people who have
helped to make this thesis possible. First, I would like to thank my advisor, Dr.
Pennie Foster-Fishman, and my committee members, Dr. Ken Frank and Dr.
Frank Fear. I am also grateful to the MSU Families and Children Together
(FACT) coalition for funding this research. My heartfelt thanks go to Brad Wallet
for working with me on an earlier version of the R-language program that helped
so greatly to round out my analysis, and for giving me the impetus I needed to
learn R for myself. In addition, I am extremely grateful to Brandy Nowell for
sharing with me both the true friendship and intellectual colleagueship that on
bad days help to see me through and on good days remind me of why we got
into this business in the first place.

While I will omit their names here, I also wish to express my deepest
gratitude to my community partners in Lansing, including the Neighborhood
Network Center staff, board, and especially the citywide coordinator. As a
community scholar, I have been privileged to work with and learn from people
with so much wisdom, dedication, talent, and experience in community practice.
You are truly an inspiration.

Finally, I wish to thank my family. I am grateful to my parents for their
love, support, and unwavering belief in me. And, above and beyond my gratitude
for his patience and support through the latter part of this process, I wake every
morning feeling blessed to be sharing my life with Zachary Chartkoff.

Thank you all.

iii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................ viii
LIST OF FIGURES ..................................................... ' .......................................... x
CHAPTER 1
INTRODUCTION .................................................................................................. 1
1.1 Overview .............................................................................................. 1
1.2 The Neighborhood Network Center Model ........................................... 5
1.2.1 Functions of the Neighborhood Network Centers ................... 5
1.2.2 History .................................................................................... 8
1.2.3 Funding .................................................................................. 9
1.2.4 Structure and Governance ..................................................... 9
1.3 Literature Review ............................................................................... 12
1.3.1 Overview .............................................................................. 12
1.3.2 Network Centers as Bridging Agents: Conceptual
Underpinnings .................................................... g ........................... 12
1.3.3 Network Centers as Service Models: Comparative
Approaches ................................................................................... 30
1.3.4 Summary .............................................................................. 32
1.4 Research Questions ........................................................................... 33
CHAPTER 2
METHODS .......................................................................................................... 35
2.1 Study Overview .................................................................................. 35
2.2 Phase l—Social Network Analysis of the Community System ........... 36
2.2.1 Objective .............................................................................. 36
2.2.2 Bounding the Network and Defining the Sample .................. 37
2.2.3 Instrument ............................................................................ 42
2.2.4 Procedures ........................................................................... 47

iv

2.2.5 Data Entry and Management ................................................ 49

2.2.6 Preliminary Analysis ............................................................. 50
2.2.7 Understanding Bridging Through Formal Brokerage. ........... 54
2.3 Phase Il—Qualitative Interviews ........................................................ 68
CHAPTER 3
RESULTS ........................................................................................................... 71
3.1 Overview ............................................................................................ 71
3.2 Identification of Formal Brokerage Roles ........................................... 71
3.2.1 Brokerage at the Neighborhood Level .................................. 73
3.2.2 Brokerage Between Neighborhood and City-Level
Organizations ................................................................................ 83
3.2.3 Brokerage Among City-Level Organizations ......................... 90
3.3 One Model, Three Manifestations ...................................................... 90
3.3.1 Network Center A: “The Convener" ...................................... 93
3.3.2 Network Center B: "The Entrepreneurial Advocate" ........... 100
3.3.3 Network Center C: "The Interface” .......... I ........................... 107
3.4 Summary .......................................................................................... 112
CHAPTER 4
DISCUSSION .................................................................................................... 114
4.1 Overview .......................................................................................... 114
4.2 Interpretation of Brokerage Patterns ................................................ 115
4.2.1 Differences Across Network Centers .................................. 115
4.2.2 Difference Between Information Sharing and Joint Planning
and Programming ........................................................................ 123
4.3 Implications ...................................................................................... 126
4.3.1 Importance of Trust/Credibility ............................................ 126
4.3.2 Process-Product Integration ............................................... 127

4.3.3 Sustainability ...................................................................... 131
4.4 Limitations ........................................................................................ 133
4.4.1 Methodological Limitations of the Social Network Analysis 133

4.4.2 Methodological Limitations of the Qualitative Follow Up

Interviews .................................................................................... 1 37
4.4.3 Overall Limitations. ............................................................. 137
4.5 Directions for Future Research ........................................................ 138
4.6 Conclusion ....................................................................................... 142
APPENDIX A Participants by Organizational Role .......................................... 145
APPENDIX B Invitation to Participate — Phase I ............................................. 147
APPENDIX C Phase I Informed Consent ........................................................ 150
APPENDIX D Phase I Pre-Interview Information ............................................ 153
APPENDIX E Phase I (Social Network) Survey .............................................. 156
APPENDIX F Reported Tie Change Data ....................................................... 167
APPENDIX G Transitive vs. Brokered Triads at Varying Cutoffs .................... 170
APPENDIX H Phase II Interview Protocol ....................................................... 173
APPENDIX l Calculation of Standardized Brokerage Scores ......................... 175
APPENDIX J Brokered Triad Classification Routine ....................................... 179
APPENDIX K Brokerage Score Tables ........................................................... 184
APPENDIX L Reported Ties and Tie Strengths by Region/
Organization Type ............................................................................................. 191
APPENDIX M Network Maps .......................................................................... 194
APPENDIX N Neighborhood Census Data ..................................................... 208
APPENDIX 0 Basic Network Measures ......................................................... 210
APPENDIX P Sample Qualitative Analysis Summary Tables ......................... 213
APPENDIX Q Qualitative Analysis of ChangeData ........................................ 219

vi

REFERENCES ................................................................................................ 224

vii

LIST OF TABLES

Table 1. Participation rates by region ................................................................. 42
Table 2. Approximate tests of concentration of ties 4 Kliquefinder subgroup
algorithm. ............................................................................................................ 52
Table 3. Analysis of response rates to relational change measure items. ......... 53
Table 4. Number of organizations and participation rate by type and Network
Center partition. .................................................................................................. 63
Table 5. Standardized brokerage scores by Network Center, relationship type,
and brokerage type ............................................................................................. 73
Table 6. Key for combinations of organizations shown in Figure 2. ................... 75
Table 7. Key for combinations of organizations shown in Figure 3 .................... 85
Table 8. Comparison of key features of Network Centers and surrounding
neighborhood areas. ........................................................................................... 92
Table 9. Participants by organizational role ..................................................... 146

Table 10. Degrees of reported changes in relationships by type and region of
reporting and cited organizations, aggregating new/stronger reported ties and
. weaker/defunct reported ties .................................................. , ........................... 1 68

Table 11. Brokerage scores for neighborhood organizations within subnetworks,
by neighborhood and tie type ............................................................................ 185

Table 12. City-level organizations with significant positive brokerage scores for
information sharing between neighborhood-level organizations
(consultant/itinerant brokerage), by neighborhood subnetwork. ....................... 188

Table 13. City-level organizations with significant brokerage scores for joint
planning and programming between neighborhood-level organizations
(consultant/itinerant brokerage), by neighborhood subnetwork. ....................... 188

Table 14. City-level organizations with significant brokerage scores for
information sharing between city-level and neighborhood-level organizations
(gatekeeper/representative brokerage), by neighborhood subnetwork. ............ 189

Table 15. City-level organizations with significant brokerage scores for joint
planning and programming between city-level and neighborhood-level
organizations (gatekeeper/representative brokerage), by neighborhood
subnetwork. ...................................................................................................... 190

viii

Table 16. Reported ties and ties strengths by organization type and region... 192

Table 17. Economic status of Network Center primary service areas (per 2000
Census Data) .................................................................................................... 209

Table 18. Basic Network Measures for Sub-Networks - Information Sharing Ties
by Network Center ............................................................................................ 211

Table 19. Basic Network Measures for Sub-Networks - Joint Planning &
Programming Ties by Network Center .............................................................. 212

Table 20. Sample qualitative analysis summary table: Network Center roles..214
Table 21. Sample qualitative analysis summary table: “what makes it work”...215

Table 22. Core themes in content analysis of open—ended survey items. ........ 220

ix

LIST OF FIGURES

Figure 1. Types of Formal Brokerage Measures as'Det‘Ined by Gould &
Fernandez (1989) and Applied in this Study ....................................................... 58

Figure 2. Patterns of direct and brokered ties among neighborhood-level
organizations by Network Center for lnforrnation Sharing (Info) and Joint
Planning and Programming (JPP) ....................................................................... 76

Figure 3. Patterns of direct and brokered ties between neighborhood and city-
level organizations by Network Center for Information Sharing (Info) and Joint
Planning and Programming (JPP) ....................................................................... 86

Figure 4. Transitive vs. brokered triads at varying cutpoints for information
Shaflng .............................................................................................................. 171

Figure 5. Transitive vs. brokered triads at varying cutpoints — joint planning and
programming ..................................................................................................... 172

Figure 6. Neighborhood-Level Network Of Neighborhood A - Information
Sharing. Node size is proportional to coordinator brokerage score. ................ 196

Figure 7. Neighborhood- -Level Network Of Neighborhood A- Joint
Planning/Programming. Node size is proportional to coordinator brokerage
score. ................................................................................................................ 197

Figure 8. Neighborhood-Level Network of Neighborhood B - Information
Sharing. Node size is proportional to coordinator brokerage score. ................ 198

Figure 9. Neighborhood-Level Network of Neighborhood B — Joint Planning and
Programming. Node size is proportional to coordinator brokerage score. ....... 199

Figure 10. Neighborhood-Level Network of Neighborhood C — Information
Sharing. Node Size is proportional to coordinator brokerage score. ................ 200

Figure 11. Neighborhood-Level Network of Neighborhood C - Joint Planning
and Programming. Node size is proportional to coordinator brokerage score. 201

Figure 12. Full Subnetwork for Network Center A (Neighborhood + Citywide) —
Information Sharing. Node size is proportional to gatekeeper/representative
brokerage score. ............................................................................................... 202

Figure 13. Full Subnetwork for Network Center A (Neighborhood + Citywide) -
Joint Planning and Programming. Node Size is proportional to
gatekeeper/representative brokerage score. .................................................... 203

Figure 14. Full Subnetwork for Network Center B (Neighborhood + Citywide) -—
Information Sharing. Node size is proportional to gatekeeper/representative
brokerage score. ............................................................................................... 204

Figure 15. Full Subnetwork for Network Center B (Neighborhood + Citywide) —
Joint Planning and Programming. Node size is proportional to
gatekeeper/representative brokerage score. .................................................... 205

Figure 16. Full Subnetwork for Network Center C (Neighborhood + Citywide) -
Information Sharing. Node size is proportional to gatekeeper/representative
brokerage score. ............................................................................................... 206

xi

CHAPTER 1
INTRODUCTION

1.1 Overview

Over the past two decades, a number of trends have emerged in the way
local communities structure their interactions with and services to local
neighborhoods. These have included a significant emphasis on community-
based and community-driven approaches to change and service delivery
(Chaskin, 2001 ). For example, communities have increasingly implemented a
range of initiatives such as community development intermediary organizations,
comprehensive community-building initiatives, and a multitude of efforts at
collaboration, coalition-building, and service system integration.

All of these trends have placed a widespread emphasis on various models
of service coordination and collaboration as means to improve functioning across
‘ organizations in the community system. At their core, all of these trends have
much to do with increasing the capacity of the local system to respond to
community needs by changing the way organizations relate to each other and to
neighborhood sub-systems to make decisions, solve problems, coordinate
strategies, and implement programs and services in order to achieve greater
impact.

The impetus for these shifts comes from several related developments.
First, in many cities across the country, there is an increasing recognition among
community institutions, local policy makers, and funders that traditional,
fragmented modes of service delivery fail to support healthy community

functioning and are inadequate for the complex array of social, economic, and

physical issues facing communities today. That is, although health and human
service functions have traditionally been compartmentalized into distinct service
sectors with separate funding streams, philosophies, and intervention strategies
(Kagan & Neville, 1994; Roemer, Kramer, & Frink, 1975), many of the problems
affecting communities are what Ackoff (1974) dubbed "messes”—issues too big
and too complex to be solved by any single organization or sector (Bruner, 2000;
Kagan & Neville, 1994).

This call for service coordination has been further emphasized with
growing support for comprehensive and place-based (and particularly
neighborhood-based) approaches to community change and service delivery, as
service providers have recognized that much of their target populations are
concentrated in inner-city neighborhoods that often lack access to service
resources. As many social issues tend to correlate with poverty, this trend has
been exacerbated with the movement of jobs and economic activity to the
suburbs (Katz, 1998), creating further isolated pockets of concentrated poverty
and entrenched social issues. A focus on servicing low-income neighborhoods
carries the aim of concentrating resources targeted where they are needed most
(Aspen Roundtable, 1998) and promoting more effective service delivery by
making services easier for low-income people to access.

At the same time, federal and state policies over the past two decades
have created a parallel demand for increased coordination at the local levels, as
federal funding cutbacks and devolution of authority to state and local agencies

have forced human service agencies to attempt to ”do more with less” by using

coordinated approaches to service provision to reduce duplication of services
and improve performance (Holcomb, Seefeldt, Trutko, Bamow, & Nightingale,
1993; Stone, 1994). ‘

Ironically, these same funding pressures have created a policy
environment that has encouraged growing reliance on contracting public services
to private nonprofit entities, as public agencies seek to minimize their own
service costs (Milward, 1994; Mirabella, 2001; Smith & Lipsky, 1993). This has
resulted in services that are further decentralized (Milward & Provan, 2000) and
delivered by a broader array of smaller organizations; as a result, ”public services
in the community are jointly produced, requiring increased levels of collaboration
and cooperation among workers in these sectors.” (Mirabella, 2001) Thus,
community service systems face the dual challenges of simultaneously pulling
services apart and putting them back together across organizational boundaries
in constellations that efficiently and effectively meet a complex and
interdependent array of community needs.

With these combined trends has come an interest in and movement
toward structures and processes that can foster these critical linkages among
organizations, including the growth of comprehensive, community-based
approaches that support improved linkages between neighborhoods and
community service systems to promote positive change for families,
neighborhoods, and community systems (Aspen Roundtable, 1998). While there
are many approaches to developing these structures, one strategy of such efforts

has been to establish or support what Chaskin et al (2001; Chaskin, Brown,

Venkatesh, & Vidal, 2001) has called "broker" organizations-organizations
endowed with the role of fostering partnerships and networks among existing
organizations in the community:
”By creating broker organizations, community capacity-building
efforts attempt to provide a mediating institution embedded in the
community that can acts as an ongoing mechanism for problem
solving, resource development and acquisition, and, possibly, a
degree of community-level decision making and governance.
Fundamentally, broker organizations serve as organizational

mechanisms through which relations may be created, fostered, or
negotiated." (Chaskin, 2001, p. 149)

The purpose of this study is to look at whether and how one type of
intermediary structure in one community does this. This study presents and
examines one community's strategy for developing community capacity through
the creation of a system of Neighborhood Network Centers, neighborhood-based
intermediary organizations (a form of "broker organization" as described by
Chaskin, 2001) that are designed to foster relationships at the neighborhood
level and between neighborhoods and the larger, citywide system of local
government, health and human service agencies, and nonprofit organizations.
This model is described in detail in the following section. While such approaches
are designed to facilitate needed partnerships, the extent to which they actually
do promote relationships and strengthen the systems they operate in has been
broadly described but not closely, specifically, and empirically examined. The
purpose of this study is to generate an empirical understanding of the roles that
neighborhood-based intermediaries in this community system play in facilitating
interorganizational relationships and thereby enhancing the organizational

infrastructure of the local community system.

1.2 The Neighborhood Network Center Model
1.2.1 Functions of the Neighborhood Network Centers

Neighborhood Network Centers as developed 'in Lansing, Michigan
represent one implementation of the ”broker organization" concept. These
entities attempt to serve an intermediary role by combining neighborhood-based
assessment, planning, service coordination, and resident-focused activities into
an approach for integrating the activities of neighborhood-based and citywide
organizations and fostering partnerships between these organizations for the
betterment of their focus neighborhoods. Neighborhood Network Centers take a
comprehensive approach to neighborhood issues, seeking to address the
interdependent health, education, safety, social service, and housing issues of
neighborhood residents through neighborhood driven planning and programming.
These five domains constitute the primary focus areas of the Network Centers.
This model operates at multiple levels as described below. '

Comprehensive Planning. First, Network Centers serve as a local base
for comprehensive planning of programs and special initiatives to meet resident
needs and identify and achieve neighborhood improvement goals. This planning
occurs to varying degrees and in varying forms across the Network Centers, but
all of these processes incorporate some level of resident input into the needs and
issues facing the neighborhood and the collaborative involvement of both
neighborhood-level and service-system-level (i.e., city-wide) stakeholders.
Through these processes, Network Centers are actively engaged in creating a

”nexus" between their individual neighborhoods and the greater urban system.

Comprehensive Prggramming. Second, network centers strive to develop
the individual capacities of neighborhood residents through activities and
programs that build knowledge and skills (e.g., parenting classes, GED
completion programs, after-school and summer activities for kids, health
education programs, workshops on neighborhood identified topics, etc.) and
opportunities for leadership and volunteer service. In addition, Network Centers
provide information and referral services to connect residents with resources to
meet their needs. Some of the programs run at network centers are developed
and conducted in-house by network center staff and volunteers. Other programs
and activities are coordinated through community partners such as the school
district, social service agencies, and other non-profit organizations; however,
even these external programs often use the neighborhood centers as sites for
service delivery due to the sites' local accessibility and familiarity to
neighborhood residents. 2

Neighborhood Spac_e_._ Third, Network Centers provide the neighborhoods
they serve with community-friendly space for meetings and activities.
Neighborhood space, communal facilities, and other settings of informal public
life have been emphasized by scholars for their importance as vital mechanisms
for community networking (Gilchrist, 2000) by providing venues for citizens to
interact, socialize, and discuss issues of common concern (Oldenburg, 1989).
Thus, Network Centers also provide a venue for neighborhood residents to build
stronger associations with each other, through participation in common activities

and opportunities for both formal and informal socializing. Perhaps most

 

importantly, neighborhood network centers provide a venue where community
members-no matter what issues they may be facing—are treated as neighbors
rather than numbers or clients. Thus, applying the perspective of the Center for
the Study of Social Policy, Network Centers constitute a dual locus: they serve as
a locus around which service agencies provide needed support to neighborhood
residents, and they serve as a home base for resident gathering, advocacy, and
mobilization. (Center for the Study of Social Policy, 1996)

Through these three core aspects of their design, Neighborhood Network
Centers are positioned to act as intermediaries or "brokers" on behalf of the
neighborhoods. By occupying a functional role that logically places them at the
center of much neighborhood activity as well as in high levels of contact with
extra-neighborhood service providers, Network Centers are theoretically able to
facilitate collaboration and mediate among neighborhood organizations and the
larger, citywide community service1 system. This mediating function is
recognized and valued by community stakeholders. In key informant interviews
and informal conversations with community stakeholders, relationship building
was identified as the most critical function of the network center model. However,
this function has heretofore been implicitly recognized but not explicitly analyzed

within this community.

 

1 For the sake of clarity and brevity, the term “community service system” will be used throughout
this paper as a blanket reference to the sum total of local organizations and institutions which
serve to support community residents and community functioning. These include local
government entities, law enforcement, health and social service agencies, nonprofits, schools,
faith-based organizations, neighborhood associations, et cetera.

1.2.2 History

The "Network Center" model was originally developed in response to what
major human service agencies in the area perceived as a dire lack of
coordination among their respective service operations—to paraphrase the
explanation of one agency staff member, "Case managers became aware that
the only time they communicated with each other was when they ran into each
other in a client's driveway." This growing realization provided the impetus for
creating the first Network Center in the early 1990's, which served as a
neighborhood-based agency hub, providing a common base of operations for
case management staff from multiple agencies to serve clients in one
neighborhood. Although this center closed after a few years, key community
leaders among city government and the health and human services domains
maintained a vision of creating an enhanced model of comprehensive
neighborhood-based service planning and delivery.

In 1996, one of the lead entities (a local hospital) joined forces with the
umbrella neighborhood organization serving the neighborhoods in one part of the
city to convene a Neighborhood Health Summit. In step with the national trends
towards comprehensive focus described above, the rationale for this "summit
process" stemmed from the perception that genuine "community health"
improvement required the adoption of a broad definition of health including

issues such as housing, jobs, income, safety and education.

As a result of the success of this event, these lead entities secured initial
funding to launch five Network Centers1 distributed throughout the city in its most
disadvantaged neighborhoods. These were fully launChed over the course of
1998-1999. In addition, Summit processes were continued in the original
neighborhood area and expanded to two other parts of town in 1999-2001 with
the aid of a private foundation grant.

1.2.3 Funding

The Network Centers are funded by a veritable patchwork of local funds
and outside grants. Staff positions are funded through the county Family
Independence Agency, the City of Lansing, and the Corporation for National and
Community Services (VISTA program). Facilities are funded by a combination of
city and county monies and in-kind space provided by the community action
agency that shares a building with one of the Network Centers. Operating costs
and program funds are provided by a wide range of grants and contributions from
federal and state agencies, city and county health and social service
departments, community foundation grants, and lesser donations from local
hospitals, service organizations, for-profit organizations, churches, and
nonprofits.

1.2.4 Structure and Governance

In practice, individual NNCS vary somewhat in their structure and
governance. However, each operates through a combination of paid and

volunteer staff, and each has a neighborhood-level Network Center Advisory

 

1 Three of these Network Centers are still in active existence; the other two were active at the
launch of this study but have since become defunct for reasons unrelated to performance.

Board which it reports to and which provides general direction for the Center's
priorities and activities. In addition, an umbrella board and a citywide Network
Center system coordinator provide overall guidance and oversight for the system
of centers as a whole. These aspects are discussed in greater detail below.

Director and Staff. Each Network Center is led by a paid director, with
support for administering specific programs and activities provided by a
combination of paid (primarily Americorps VISTA workers) and volunteer staff.
Most staff and volunteers are residents of the neighborhoods they serve or have
strong local ties to those neighborhoods. A citywide Network Center system
coordinator serves a technical assistance and resource development role for the
centers themselves, troubleshooting partnership and management issues as they
arise, helping network center staff overcome system barriers to their work,
securing funding from local partners, and working with directors to write grants to
extra-community funding sources. 1

Governance Bodies. Neighborhood Advisory Boards and committees are
primarily composed of representatives from the surrounding neighborhood(s)
with some additional representation from local schools, social service agencies,
police, churches, and other institutions. These Neighborhood Advisory Boards
provide guidance to Network Center staff, establishing the overall direction and
focus for each center, helping to identify needs and set priorities, and engaging in
action planning processes and strategizing around neighborhood issues.

Through a collaborative umbrella structure composed of representatives

from the key community support institutions across the city (the Citywide Network

10

Center Advisory Board or CNCAB for short), the network centers serve both as a
focal point for collaboration and as a mediating organization for the
neighborhoods. Network Center staff serve as representatives to the Citywide
Board, identifying community needs and strategies through their day-to-day
presence in the neighborhoods and the direction of their neighborhood advisory
boards and bringing them to the collective table where they can be shared with
the local government, health department, welfare and mental health agencies,
hospitals, school district, police department, and other neighborhood based
organizations. The CNCAB serves to support the neighborhood network center
efforts by recommending programs and activities, identifying system resources
that can be applied toward Network Center programs and plans, and providing
updated information on the activities, policies and programs available through
service system entities. In addition, NNC staff are able to provide agency and
institutional representatives to the CNCAB helpful information regarding what the
pressing needs in the neighborhoods are, how well programs and services are
being received by neighborhood residents, and what changes in programming
are required to have the greatest impact.

Thus, through their day-to-day operational processes (i.e., planning,
programming, and service coordination activities) and their broader connections
to the community through advisory board structures, Network Centers are
positioned to promote greater sharing of information and resources across these
various organizations, and facilitate collective planning and programming efforts

for more effective service delivery.

11

1.3 Literature Review
1.3.1 Overview

As discussed above, Neighborhood Network Centers represent one model
for the creation of a neighborhood-based intermediary role in a local community
system. Ideally, these neighborhood-based intermediaries are engaged not just
in developing services for neighborhood residents but serve an important
systemic function as well, by brokering between organizations in the system.
This function evolves from the creation of an interorganizational social structure
which enables neighborhood-based entities and city and county-wide agencies
and institutions to work effectively together, identifying needed changes and
untapped resources and implementing innovative solutions. As such, Network
Centers represent both service implementation and structural intervention for the
local community system.

Therefore, an inquiry into a model such as the Neighborhood Network
Centers is best approached and understood within the context of both related
models of neighborhood development and service implementation and the
conceptual bases for their role as bridging agents. This literature review is
organized around these two areas.

1.3.2 Network Centers as Bridging Agents: Conceptual Underpinnings
Communig Capacitvaj Commu_nitv Organizational Infra_structure. The

creation of many organizational structures common to current organizational
change efforts (e.g., intermediaries, coalitions, etc.) is premised, in part, on the
recognition that many communities lack the capacity to mobilize for change,

effectively solve problems or leverage resources. Thus, despite the various forms

12

acquired by present-day approaches to community development and service
coordination, at their core they typically share an orientation toward enhancing
the capacity of the community system to address the needs of local residents.
Community capacity has been broadly described as both: (a) the general ability
of a community to identify and mobilize around community issues and problems
and develop, implement, and sustain actions to address those issues and
promote community well-being and functioning (e.g., Mayer, 1994; Rubin &
Rubin, 1986; Smith, Littlejohns, 8. Thompson, 2001); and (b) as well as the
characteristics of communities that affect that ability. Frameworks for these
characteristics (Chaskin & Abunimah, 1999; Goodman et al., 1998) have
included:

1. A sense of community among community members, including a
threshold level of connectedness and collectively held values and
norms, a sense of belonging to and ownership of the community, and a
recognition of shared circumstances;

2. A level of commitment on the part of individuals, groups, or
organizations who view themselves as active stakeholders in the
collective well-being of the neighborhood (i.e., active participation and
leadership on the part of both residents and partner institutions);

3. Mechanisms of problem solving through which issues can be
addressed, priorities identified, and problems solved, as well as the
Skills to assess those community needs, and implement identified

solutions; and

13

4. Access to varied resources both within and beyond the neighborhood,

primarily through social and interorganizational networks.

Chaskin and Abunimah (1999) sum these factors up as ”the interaction of
human, organizational, and social capital existing within a given community that
can be leveraged to solve collective problems and improve or maintain the well-
being of a given community" (p. 4).

Community capacity building is important to the comprehensive view of
community change because it represents an attempt to create an ongoing ability
within a community to identify and address complex issues rather than implement
one-shot or piecemeal solutions to particular problems. Interventions which
incorporate and build upon an existing foundation of community capacity have
greater chance for success than those which ignore—or, more dangerously,
ignore the absence of—these capacity factors. (Goodman et al., 1998; Smith et
al., 2001). Community capacity building is likewise important because it places
attention on creating the ability for a community to involve multiple levels of
stakeholders—including community residents—in identifying needs and crafting
solutions. This involvement allows efforts to build a constituency for change, to
implement decisions with support from and legitimacy in the community, and to
do so with the benefit of information and resources that can be essential to
effectively assessing and planning social services (Cordero-Guzman, 2001).
Capacity-building initiatives thus seek to transform residents from passive
recipients of services to active participants in a community change process (Finn

& Checkoway, 1998).

14

It is essential to note that all of these aspects of community capacity have,
as a potential and often perhaps critical component, an implied grounding in the
social and civic infrastructure of a community—that is, the presence and strength
of—and connections between—organizations at varying levels of the system with
an interest in and commitment to the well-being of neighborhoods and
neighborhood residents. For example, sense of community is often manifested
in and maintained through organized activities and groups such as neighborhood
organizations, faith communities, and other voluntary associations. Commitment
is most tangibly realized through the active mobilization of groups and
organizations operating both within a particular community and on behalf of that
community. Mechanisms of problem solving that can attain community-level
legitimacy are practically by definition embedded in organized group processes.
Finally, accessing and negotiating resources that can be leveraged on a
significant scale for a neighborhood is also an organizational issue. That is,
although organizations are made up of individuals and particular individuals
within a community may be leaders or othenrvise key players in actualizing these
capacities, it is most often the organizations themselves that become the focal
units for community change activity. This concurs with Knoke's (1990)
assessment that organizations are the principal actors in community systems.

It therefore follows that community capacity is, in large part, embodied in
the ability of organizations at varying levels of the system to work together to
foster positive change. This has led scholars such as Chavis (1995) and

Chaskin (e.g., Chaskin, 2001) to suggest that focusing on the organizational

15

infrastructure of a community—by developing partnerships between
organizations and changing the ways that community-serving organizations
relate to one another—is a crucial aspect of building community capacity for
change. In essence, organizational infrastructure forms the backbone of
community capacity, and relationships are key to the realization of this capacity.
Commu_nitv Organizational Infrastructure and Social Capital. The above
view is supported by social capital theory, which has highlighted the importance
of relationships in enhancing access to tangible and intangible resources (and
achieving corresponding outcomes) at multiple levels of analysis. As suggested
in Chaskin and Abunimah's (1999) framework, building social capital is one of the
primary features of developing community capacity, since it represents building
the relationships needed at both the neighborhood level and between
neighborhoods and the larger service system to promote effective change. Since
much of the work of Neighborhood Network Centers involves both fostering
relationships among neighbors and nurturing partnerships to bring diverse
resources together for the benefit of the neighborhoods on an ongoing and often
informal basis, social capital provides a particularly valuable framework through
which to understand the processes and impacts of these initiatives. Provan and
Milward (2001) suggest that social capital is an important lens for evaluating
interorganizational networks at the community level, since relationships built
among organizations constitute one outcome of interorganizational partnerships

that extend beyond the success or failure of any one particular initiative.

16

Social capital, generally speaking, refers to the resources (human,
economic, or informational) embedded in or accessible through networks of
social relationships, whether casual or close (Briggs, 1997). Social capital is
grounded in the premise that the social relationships between actors in a system
(whether it is, for example, a system of community residents or organizations
operating within a larger community service system) make it easier for those
actors both to obtain access to each other's resources to share these resources
towards common purposes. In community service systems, the prime currencies
can generally be construed of as information and knowledge resources, human
resources (e.g., staff, volunteers, or community members and their skills),
financial resources, and political resources (e.g., clout and influence). These
resources provide the critical bases for organizational action within community
service systems, but are distributed to varying degrees across the whole system
and can be difficult to obtain. In situations such as this, social capital can be
leveraged by organizations and individuals to find and access these resources
when needed. This facilitated access and resource mobilization promote the
development of shared values and collective decision making (Briggs, 1997),
thus facilitating coordinated and collective actions (Coleman, 1990; Putnam,
Leonardi, & Nanetti, 1993), promoting cooperation and collaboration in the
community (Committee for Economic Development, 1995), supporting integrated
planning across organizations, and leading ultimately to systemic change

(Briggs, 1997; Murray, 2000).

17

As such, social capital can generally be viewed as a composed of and
manifesting the linkages between three core constructs that have received
varying levels of priority by different theorists: the structure of relationships—that
is, the patterns of ties connecting actors (e.g., Burt, 1992, 2000; Coleman, 1990),
the qualities imbued in those relationships such as levels of trust and norms of
reciprocity and cooperation (e.g., Coleman, 1990; Nahapiet, 2000; Fukuyama,
1995), and the actual and potential resources that can be leveraged through
those relationships—that is, what can be accomplished by virtue of these
connections (e.g., Lin, 1999; Portes, 1998; Nahapiet & Ghoshal, 1998).

Unfortunately, at a community level of analysis, social capital has primarily
been narrowly examined as an aggregate of norms, trust, and the presence of
community venues that might facilitate the development of these via civic
participation (e.g., Putnam, 1993), largely ignoring the overall availability and
distribution of resources in communities (e.g., Stoecker, 2004) as well as the
actual relational ties among organizations at the community level. In other
words, community-level analyses of social capital have emphasized what has
been termed the cognitive dimension of social capital (Nahapiet & Ghoshal,
1998; Bolino et al., 2001) and paid little attention to understanding how
community ties are structured or how flows of resources in community systems
serve in practice to enhance capacity and facilitate action and change.

More recently, however, community psychologists have described social
capital at the community level as "the physical presence and mix of institutions

and organizations—together with their networks of relationship" (Perkins,

18

Hughey, & Speer, 2003, p. 44), promoting the idea of social capital as a resource
grounded in actual relationships among community organizations. The result of a
healthy network of such relationships should be the efficient and effective flow of
information, resources, and efforts across organizations within the system for the
benefit of community residents. The benefit of applying this more structurally-
oriented framework to the development of community capacity is that it clarifies
potential avenues for intervention through understanding and enhancing the
presence, pattern, and quality of interorganizational relationships in the
community as well as the inherent capacity and service resources of the
organizations within it.

Additionally, social capital theory points to different types of social
capital—and corresponding targets for intervention—that are relevant to
promoting a healthy community network. ”Bonding" social capital (e.g, Putnam,
1999; Gittel & Vidal, 1998) is formed by close ties in dense networks. This has
been described in relation to close-knit neighborhoods, representing the kinds of
local social networks that can provide social support and resources that residents
can share to help each other ”get by" in a crisis (Briggs, 1998). Because
Network Centers are neighborhood-based settings for resident activities and
other forms of voluntary association, they serve as a venue for these types of
relationships to form (Putnam et al., 1993; Putnam, 1999).

Importantly, however, neighborhoods also need vertical linkages that can
work to bring the external resources embedded in community service systems

into the neighborhoods (McKnight & Kretzmann, 1984). According to Narayan

19

(1999, p. 13), "When power between groups in asymmetrically distributed, it is
cross-cutting ties, the linkages between social groups, that become critical to
both economic opportunity and social cohesion." Without this access, even
though neighborhood residents can find participation in local social networks
personally enriching, these networks can still coexist with a general sense of
disempowerment (Lappé & Du Bois, 1997). In addition, horizontal linkages
among different neighborhood groups help to knit the smaller cohesive elements
of a neighborhood into a more coordinated whole that then better positioned to
advocate and mobilize for change on a broader scale. Thus, an extensive array
of "weak ties" (Granovetter, 1973) is needed to maintain the overall integration
and coordination of a system across the gaps in social structure -- what Burt
(1992) calls "structural holes —- that occur between more densely-connected
pockets of individuals and organizations. These cross-group linkages foster
"bridging” social capital—enabling, for example, collective action and resource
flow across different neighborhoods or different sectors of the community service
system (Gittell & Vidal, 1998).

As these facets of social capital in community life illustrate, attention to
these structural elements of relationships among segments of a community
system can potentially enhance our understanding of what helps community
systems to leverage resources for change. As such, the present study embraces
the integration of structural approaches into the understanding of community-
level social capital. It is important to recognize that relationships among

organizations are dynamic, and that relational structure and community

20

processes are intertwined rather than connected in a linear, causal relationship
(e.g., Gabbay & Leenders, 2002). Indeed, the widespread implementation of
collaborative initiatives in communities suggests that community systems are
engaged in fostering linkages at the same time that they are facilitated and/or
constrained in action by the existing patterns of these linkages, suggesting the
need to acknowledge the role of intentional action (e.g., Emirbayer & Goodwin,
1994) while attending to community structure.

Intermediary/Broker Organizations as Structugl Interventions. What can
communities do to help create these critical linkages? As described above,
organizational infrastructure is a key component to community capacity and, by
extension, the ability of communities to address the myriad of social and
economic problems facing many urban neighborhoods. Because of this,
communities (and foundations working with communities) are increasingly
investing resources in efforts that can enhance relationships among
organizations and make the existing organizational resources in a community
more effective. As Gilchrist (2000) notes, "If the networks themselves are seen
as a community asset, then social investment must be directed towards initiatives
and infrastructural resources which facilitate dynamic relationships and open
communication" (p. 272).

Although this thesis focuses on one community's implementation of one
type of structural intervention, there are in fact multiple ways to develop this
community organizational infrastructure, and the individual implementations of

these approaches are perhaps as unique as their host communities and

21

participating partners. However, several broad categories have been discussed
in the literature. First, coalitions and interagency teams have grown common in
nearly every community. While these structures typically have missions tied to
launching initiatives, coordinating service delivery, and/or promoting information
exchange among partner organizations within the coalition context, it is important
to note that these partnership structures have been found to promote increased
interaction and collaboration between participating organizations outside this
context as well (e.g., Foster-Fishman, Salem, Allen, & Fahrbach, 2001; Vicary,
Doebler, Bridger, Gurgevich, & Deike, 1996).

Second, communities employ what Laumann, Galaskiewicz, and Marsden
(1978) describe as "federative approaches” —— that is, established but loose-knit
networks of organizations that mobilize as needed. In contrast to forrnalizing
relationships through the "virtual” organization structure of a coalition, "the more
extensive and permanent commitment characteristic of a federative subnetwork
is typically formalized by the creation of a special-purpose organization endowed
with the prerogatives of acting on behalf of the entire set of constituent
organizations." (Laumann et al., 1978, p. 474). These roles are often identified
with formalized public-sector service networks and have been referred to by
multiple names including network administrative organizations (NAOs) (Provan &
Milward, 2001) and network brokers (Lawless & Moore, 1989; Mandell, 1984).

Third, communities have developed a range of "intermediary
organizations" and ”broker organizations." While perhaps similar in function to

NAO role described above, these organizations are not tied to a particular

22

freestanding network, but instead serve to varying degrees to build and support
flexible networks within the communities they serve. As brokers, these
intermediary actors can help to facilitate transactions between other actors in the
system who may lack access to one another (Marsden, 1982). The role of
"broker" or "bridger" is viewed as best filled by persons, organizations, or entities
with a broad social interest in the well-being of the community (Gittell & Vidal,
1998) who can attempt to fill structural holes to produce broader social benefits
for the community at large.

From the perspective of social capital theory and community
organizational infrastructure, approaches such as multi-stakeholder collaboration,
formalized and facilitated interorganizational networks, comprehensive
community initiatives, and intermediary organizations can be viewed as structural
interventions that carry the potential for fostering new connections and dynamics
of interaction among different sectors and levels of the community system. Of
these, the Neighborhood Network Center model represents one type of
neighborhood-based intermediary or broker organization.

Chaskin (2001) describes three primary modes for broker organizations to
perform these bridging functions:

First, as matchmakers, broker organization's serve as neutral third parties
that bring organizations together for limited, manageable partnership projects
(Chaskin, 2001). In addition, they help link community sectors and levels by
convening venues for exchange among stakeholders (Blank et al., n.d.; Wynn,

2000), building partnerships, and brokering resources among organizations

23

(Blank et al., n.d.) Network Centers are positioned to act as this sort of match-
maker through their work in developing comprehensive services and projects for
execution at the neighborhood level, involving a wide variety of partners in the
process. In addition, through various venues (e.g., advisory boards and task
forces) Network Centers create generalized opportunities for organizations and
residents to build partnerships for change.

Second, as clearinghouses, broker organizations may act as an efficient
conduit for information and resources, streamlining communication and resource
distribution and improving neighborhood-level access to these resources
(Chaskin, 2001). According to Chavis et al (1993), ”The most basic function of
an enabling system that intermediary support organizations can provide is the
direct provision of information and resources for referral to other appropriate
resources." Referral ability is paramount, because although a given intermediary
may not have all of the knowledge or capacity to address every need in-house,
the knowledge within the system of where resources exist allows organizations to
be connected with the knowledge they need (Chavis et al., 1993). Network
Centers are designed to act as knowledge clearinghouses in their
neighborhoods, as they are charged with the task of providing outreach services
to local residents and connecting residents with the breadth of resources in the
community service system. By virtue of their breadth of their connections,
Network Centers are positioned to have an extensive grasp of the range of

resources available.

24

Third, as community representatives, broker organizations may develop
and facilitate local decision-making processes that may then confer upon the
broker organization, as the convener of these processes, the legitimacy to speak
on behalf of a neighborhood area and represent its interests to the larger
community (Chaskin, 2001). Network Centers demonstrate this mode through
their frequent roles representing the needs of their neighborhoods to decision
makers at the city level.

In addition to these relational bridging modes, intermediaries typically
serve a range of specific functional roles as well, providing services such as
technical assistance, training, knowledge development and dissemination, and
management assistance (Chavis, Florin, & Felix, 1993; Wynn, 2000). In local
practice, one Network Center provides extensive technical assistance to local
neighborhood organizations. However, all of the Network Centers occupy
functional roles in the sense that they provide a range of tangible services in their
local areas, including varying degrees of direct programming and information
gathering.

The Network Center model is also compatible with Chaskin's emphasis on
the very local nature of these broker organizations and defines them as
necessarily in the community they serve. This localization is essential for helping
the broker to maintain the comprehensive focus and function needed for
neighborhood-based work:

"Across cities, as organizations move further from the

neighborhood in terms of their role and connection, they emphasize

fewer and more specialized functions, particularly focusing on
providing financial or technical assistance or acting as

25

implementation mechanisms for particular programs or the delivery
of particular services. Organizations closer to the neighborhood
level—associations, CBOs, coalitions, and initiatives—tend to
embrace a broader array of roles, particularly stressing outreach
and fostering neighborhood involvement." (Chaskin, 2003)

In sum, the goal of Network Centers as a structural intervention is to serve
as an effective link or broker between organizations at the neighborhood and
city-wide levels, promote partnerships between these various organizations,
provide multiple venues for organizations to interact, and serve as neighborhood-
based ”hubs" for the integration of various programs and services.

Neighborhood Network Centers are designed to promote this bridging function
through the connection of residents and neighborhood-level organizations with
collaborative efforts across organizations in the service system. In particular, the
neighborhood network center model is designed to establish Network Centers as
broker organizations that have legitimacy both within their individual
neighborhoods and to the larger service system. While they provide a venue for
neighborhood residents to develop bonding social capital, they also serve as a
conduit for voicing the needs of the neighborhoods to service system entities,
advocating on behalf of neighborhoods, and identifying creative opportunities for
bringing programs, services, staff, and other resources into the neighborhoods
using the network center facilities as host sites. As agents operating on behalf of
the neighborhoods, network center staff seek to build these bridging relationships
both through partnerships formed with individual agencies or organizations on an
ad hoc basis and through the umbrella collaborative comprising the network
center staff, other neighborhood representatives, and representatives from the

community service system. Thus, the Network Centers are designed to serve as

26

an Intermediate structure that promotes sustained, ongoing interaction between
the social networks of neighborhood residents and the networks of the
community service system. As such, these initiatives are actively involved in
reshaping the social structure of the community to facilitate necessary
connections between neighborhood residents and neighborhood-based
organizations with the larger community support system comprising local
government, police, hospitals, county and state agencies, and other institutions.

None of these approaches described above is exclusive in any
community; in fact, most communities today are likely to have a host of varying
collaborative efforts, loose federations, and intermediaries operating
simultaneously on various issues. However, neighborhood-based intermediaries
that can serve as effective broker organizations in local community systems are
important for several reasons.

First, given the drive to incorporate resident—driven approaches into
neighborhood change and service delivery, it is essential that such efforts are
grounded in a social infrastructure that can support resident engagement without
placing the burden of implementation on the shoulders of residents themselves.
As Goodman et al have suggested, "It may be unrealistic to assume that lay-
people are willing and/or able to take the initiative and lead a community health
promotion effort. Such an effort requires passion for the issues, expertise in
planning and program development, an appreciation for existing community

networks, leadership skills, and, most of all, time. Without accounting for such

27

factors, even the best models are not likely to produce the desired outcomes"
(Goodman, Steckler, Hoover, & Schwartz, 1993, p. 216).

Second, broker organizations can serve as a focal. point for
communication and a stable base for bringing the different "worlds" of
neighborhoods and government, health, and human service entities together. For
example, the existence of a strong broker organization provides organizations
outside the neighborhood with a known, stable contact for processes related to
that neighborhood and provides organizations inside the neighborhood with a
mechanism for coordination and a vehicle for creating a strong, unified presence
with outside organizations, thus creating political strength . In analyzing the
system-level attributes that affect the capacity for interorganizational activity of
different cities, Turk (1973a, b) found that organizations that could serve broker
functions provided important focal points around which coalitions of organizations
could form.

Third, research suggests that broker organizations may make
collaborative partnerships more manageable. For example, Bazzoli et al (2003)
found in their comparative examination of eight community based collaborative
partnerships that more complex partnerships comprising either larger numbers of
partners or greater growth in the number of partners showed reduced likelihood
of completing partnership tasks, but that action-oriented leadership helped to
foster task completion. However, well-connected broker organizations may
potentially be in a good position to mobilize partnerships more efficiently. As

Laumann et al (1978, p. 472) describe: "Through an established network of

28

linkages, the nature of an issues may be redefined from a parochial concern to a
matter in the public interest of a coalition of organizational actors. Within any
interorganizational field, it is likely that certain actors will be especially successful
in stimulating a coalition in this manner, because they are identified with shared
purposes or commonly held values." Thus, by creating relevant and timely foci
for collaborative activities and engaging partners on an ad hoc basis, broker
organizations may help to simplify partnerships and keep them on track and on
task.

Unfortunately, despite the growing rates of implementation of various
intermediary and neighborhood-based service strategies, our understanding of
these efforts—particularly with regard to their relational bridging functions—
remains extremely thin. This may in part be due to the fact that much of this
research is embedded in the "foundation" literature, which has tended to
generate broad-based analyses with well-integrated case studies. A As Ferguson
and Dickens (Ferguson & Dickens, 1999) complain, research is well behind
practice in this domain:

"Despite the importance of alliances, there are no standard

frameworks in urban change or community development studies for

guiding the analysis, design, implementation, or evaluation of

alliance-building processes...Discourse on this aspect of the field's

activity is relatively disorganized.” (p. 590).

In addition, although there is little conceptual (let alone empirical) literature
describing bridging and intermediary functions in community networks, there is
very little work devoted to understanding and assessing network effectiveness in

general (Provan & Milward, 2001). While Provan and Milward advocate for

increased understanding of whether networks work, given the conditions

29

promoting increased need for interaction among these components of the
system, an equally important question seems to be what network forms work and
how they do so. I

However, studies of interorganizational networks at the system level have
focused on specific service domains such as provision of mental health services
(Fried, Johnsen, Starrett, Calloway, & Morrissey, 1998; Morrissey, 1992;
Morrissey, Johnsen, & Calloway, 1997; Morrissey, Johnsen, Calloway, & Ullman,
1996; Rivard, Johnsen, Morrissey, & Starrett, 1999; Rosenheck et al., 2002) and
health care (Provan & Milward, 1995; Provan, Veazie, Teufel-Shone, &
Huddleston, 2004). In contrast, examinations of community systems
incorporating the range of city-wide and neighborhood-focused organizations
applicable to the growing focus on comprehensive and place-based approaches
to community problems have yet to emerge.

In addition, while some of the studies that have been done on community
systems have revealed findings about the structural characteristics of
interorganizational networks that promote system effectiveness (e.g., Provan &
Milward, 1995) and the impact of service coordination initiatives on overall
system integration in these types of networks (e.g., Morrissey et al., 1997),
examinations of the specific roles played by organizations in positions intended
to facilitate interorganizational linkages at any level of the system are also
lacking.

1.3.3 Network Centers as Service Models: Comparative Approaches

One essential difference between the Neighborhood Network Center

model and many other "broker organizations" is the fact that the Neighborhood

30

Network Centers interweave their brokering work with the activities of functioning
neighborhood centers which provide a range of services and programming to
neighborhood residents. Neighborhood-based service delivery is not a
particularly new concept, but it has experienced a resurgence particularly in the
past 15 years.

Settlement houses have been in existence in some communities since the
late 19th century. The Settlement House tradition couples neighborhood space
with a comprehensive array of programs that meet the needs of the particular
population in the neighborhood and some degree of advocacy and community
organizing work with residents. For example, University Settlement House,
located on New York's Lower East Side, provides a range of services such as
adult literacy classes, mental health counseling, group and family day care, case
management for the formerly homeless, recreational and educational activities
for school children, a summer day camp, a senior center, an arts program, and a
credit union. A newer but very similar approach is embodied in what sometimes
called "Family Resource Centers” — neighbor-friendly centers that promote ease
of access to services through co-Iocation of service providers (e.g., see Dunst &
Trivette, 2001; Plant 8. King, 1995; Romualdi & Sandoval, 1997). As another
example, full-service schools seek to locate community-based social services in
neighborhood schools and use schools as a basis for outreach to families
(Dryfoos, 1994; Hendrickson 8. Omer, 1995; Morrison et al., 1997).

As described in Section 1.2.1 above, in their role as neighborhood-based

service providers, Network Centers share much in common with these

31

approaches to service delivery. However, it is important to note that the literature
on these approaches is focused primarily on how these organizations operate
and what services they provide, rather than any intentional or unintentional
bridging between or among neighborhood organizations that might be part of
their work.

1.3.4 Summary

As described above, communities are increasingly implementing
strategies designed to build community capacity for change and increase service
coordination. Given the increasing emphasis on "network" and capacity-building
models for addressing community issues, such as coalitions, CCls,
intermediaries, and "broker" organizations, it is problematic that so little is yet
known about what network structures—and what structural interventions—work
to create the bridges between organizations needed to promote community
capacity. Indeed, there is a dearth of empirical literature that applies to the type
of approach represented by the Network Centers, and the literature that does
exist is largely focused on providing descriptive case studies of these
approaches.

Unfortunately, the broad focus of these reports, while an important
preliminary step in building a base of knowledge about these strategies, have
done little to extend our ability to analyze these strategies more deeply to
facilitate understanding of what precisely makes them successful—or, often,
unsuccessful. Truly, if such initiatives and efforts were universally successful,
our lack of knowledge would be less pressing. However, generalizations about

what roles these organizations should be playing do not further our

32

 

understanding of what roles they actually are playing, between whom, and how
well. At this point, a more empirical and thorough investigation of how and
whether these organizations serve their intended bridging roles is needed.

1.4 Research Questions

This study assesses the community capacity building role of the
Neighborhood Network Center model on the local community service system in
terms of the collaborative relationships between community stakeholders. While
the programmatic activities of the Network Center are tangible, the role of the
Network Centers as an intermediary or bridge between levels of the system has
been described but has not been empirically explored.

Specifically, the primary research question of this study is: To what extent
and how do neighborhood network centers play a bridging role in fostering
connections between neighborhoods and the broader social service system and
across agencies in the social service system? To answer this primary question,
the study addresses the following component research questions1 :

1. To what extent and how do neighborhood network centers play a
bridging role in fostering connections among neighborhood-level

organizations?

 

1 These primary questions were originally proposed to be answered indirectly by answering
secondary questions such as: “How has the level and pattern of collaboration between
stakeholders in the community service system changed since the development of an active
Neighborhood Network Center model in Lansing?” and “What role do stakeholders within the
community system attribute to the Network Center model in shaping this change?“ As explained
in Section 2.2.6 of this document, these secondary questions were not answerable given the
planned analysis strategy, and alternative approaches had to be identified. Therefore, this
document acknowledges and describes the original approaches but reserves the bulk of its focus
for answering the primary research questions, privileging the purpose of the research questions
over faithfulness to the original course of inquiry.

33

2. To what extent do neighborhood network centers play a bridging
role in fostering connections between neighborhood-level
organizations and system-level organizations?

3. To what extent do neighborhood network centers play a bridging
role in fostering connections among city-wide/system-level
organizations?

4. Assuming neighborhood network centers do occupy one or more of
the above bridging roles, how do they act out such role(s)?

As described above, neighborhood-based intermediary organizations such
as the Network Centers can ostensibly serve a crucial role in bridging between
organizations at the neighborhood level and the larger city-wide sphere of
agencies, institutions, local government, and other service providers. In addition,
the governance structures used by the Network Centers combinedwith the
process of bridging and bringing local stakeholders together suggests that these
Network Centers can serve significant roles bridging among organizations within
their own neighborhood area as well. Finally, although Network Centers are
clearly neighborhood-based entities, their special place in the local system may
place them in a position to serve as bridging agents among city-level
organizations as well. However, with little evidence from extant literature to guide
prediction of this role behavior, this study takes an exploratory approach to these

questions.

CHAPTER 2
METHODS

2.1 Study Overview

This study was conceived as an exploratory inquiry into identifying the
roles that Network Centers played as neighborhood-based intermediaries
bridging between organizations at two primary levels of the community system:
their local surrounding neighborhood areas, and the larger citywide service
system. To accomplish this, a mixed-method approach was developed based on
a framework of joint insider-outsider inquiry (Bartunek & Louis, 1996) that
included numerous community stakeholders as collaborative partners in
assessing the networks of relationships in the community system. As described
in the introduction above, the initial focus for the research came out of interviews
with key informants in the service system as well as ongoing involvement with
the Neighborhood Network Centers and citywide advisory board. The joint
insider-outsider inquiry process was supported through a multi-phase, iterative
research design that included Network Center stakeholders as planners,
informants, and interpreters throughout the process. As an outsider, I began my
engagement with the community system and with the Network Centers in
particular in 1999, well before the launch of the study. Since that time and
continuing through the present, I have gained direct experience and familiarity
with this system through working closely with individual Neighborhood Network
Centers, their directors, and the citywide Neighborhood Network Center
coordinator in a consulting and technical assistance capacity and attending

Network Center board meetings. These relationships as well as involvement and

35

consulting roles in other aspects of local community development efforts yielded
a strong base of familiarity with the general issues facing the Network Centers,
the larger political and social context in which the Network Centers in Lansing
operate, and many of the key players in various neighborhoods as well as in the
larger local government and service systems.

As described above, the “insider" roles in this research were filled largely
by direct Network Center staff. In particular, the Citywide Network Center
Coordinator was an essential partner due to her extensive knowledge of the
organizational makeup and “players” in the greater Lansing area. In addition,
Network Center board members were used as sounding boards for the process
and findings both during meetings and in informal conversations.

This research was conducted in two primary phases, which are described
in detail on the following pages. In the first phase, social network analysis
methods were used to generate a quantitative assessment of the community
system for addressing questions 1 through 3. In the second phase, qualitative
interviews with Network Center directors and the Citywide Network Center
director were used to contextualize the knowledge created in the first phase and
to address question 4.

2.2 Phase I—Social Network Analysis of the Community System

2.2.1 Objective

The objective of Phase I was to obtain a quantitative assessment of the
Network Centers' bridging role within the local community service system. For
the purposes of this study, the community service system is conceptualized as a

multilevel construct that includes the local neighborhood infrastructure (e.g.,

36

neighborhood associations, churches, and schools) and the broader service
system which includes city and county government offices, human service
agencies, nonprofit service organizations, and local institutions. Taken together,
these levels comprise the formal system of supports and service delivery for
residents in these neighborhoods.

This assessment was achieved through a social network analysis (SNA) of
organizations, institutions, and key neighborhood groups across the community.
SNA is a powerful statistical tool for assessing and examining social
relationships. Through this analysis, one can assess what types of relationships
organizations have with each other, the character of the relationships at the
systems level (e.g., the degree to which the community service system is
connected or disconnected), and the degree to which individual organizations or
groups are integrated into the service system.

2.2.2 Bounding the Network and Defining the Sample

Since this study is primarily concerned with the relationships between
organizations within the community service system, the community organization,
agency, or group served as the unit of analysis. The first task in an
Interorganizational network study is to define logical boundaries for the system of
relationships to be studied (Laumann, Galaskiewicz, & Marsden, 1978;
Morrissey, 1992; Wasserman & Faust, 1994)—in other words, to select which
organizations' relational data will be collected and used. There are typically two
alternative methods for bounding the network. In the deductive technique
(Morrissey, 1992), the network is bounded prior to data collection. A complete

list of actors of interest is identified first, and each actor on the list is then

37

surveyed about every other actor on the list. In the inductive technique
(Morrissey, 1992), a snowball sampling method is used starting with a group of
core organizations and successively branching out to other organizations named
through respondents free listing of their ties in each wave. Once all data are
collected, the investigator examines the data and sets criteria for inclusion or
exclusion.

In this study, a deductive approach was chosen for several reasons. First,
the organizations of interest in the community were identifiable. Therefore, a
snowball sampling strategy would be unlikely to provide many important
nominations beyond the organization lists already available. In addition, the
inductive sampling/bounding technique is generally more resource intensive and
poses greater risks of having incomplete data (Morrissey, 1992). Further, free
listing can be biasing, in that (1) nominations are subject to participant fatigue
and individual differences in ability and energy to recall a comprehensive listing
of relationships; and (2) free listing processes typically put an arbitrary limit on
the number of relationships recorded, concentrating ties on those relationships
which are most salient to participants at that time. While the latter effect can be
desirable in some study contexts, it was inappropriate for this study because
limiting the ties reported in that way would likely exclude data on connections
with the less-prominent neighborhood-level entities.

According to Wasserman and Faust (1994, p. 19), in order to properly
bound the network to study, "One must be able to argue by theoretical, empirical,

or conceptual criteria that the actors in the group belong together in a more or

38

less bounded set." In the case of the current study, the set of interest is
comprised of those organizations in the community service system that impact
the lives of residents in the neighborhoods served by the five Network Centers,
particularly in the four domain areas of health, human services, education, and
public safety. As the primary focus areas encompassed in the stated mission of
the Network Centers, these are the organizations that the Network Centers are
likely to have the most important connections to as intermediaries.

These organizations exist at two primary levels. The first level consists of
local, neighborhood-based organizations such as churches, schools,
neighborhood associations, and neighborhood-based organizations. The second
level includes such major community-wide agencies and institutions such as the
county welfare agency, the health department, the police department, local
government offices, large nonprofits, et cetera.

Given the scope of this study, it would be unrealistic to capture every
organization that might affect the lives of neighborhood residents in the city.
Therefore, in order to gain an appropriate snapshot of the community system as
it relates to the Network Center neighborhoods, l constructed a broadly
representative purposive sample of 95 organizations at both neighborhood and
citywide levels. Names of major agencies, institutions, government departments,
and non-profit organizations to include in the sample were generated through
rosters of major local collaborative bodies (such as the county's Human Service
Advisory Committee) and records of organizations that had been active in

seeking Community Development Block Grant monies and other funds

39

administered through the city. This roster was further developed through a key
informant interview with the Citywide Network Center Coordinator, a longtime
community resident who has an extensive history with and grasp of the local
service system. Because the large size and scope of some of these
organizations suggested that top leadership might be inaccessible and/or not
fully familiar with the less-prominent partnership activities engaged in, this key
informant also aided in the identification of the most appropriate and
knowledgeable contact person to represent each organization. This approach
followed examples such as Morrissey, Johnsen, and Galloway (1997) in
identifying individuals occupying “boundary spanner" roles (Aldrich & Herker,
1977) as informants in larger organizations.

For the community-wide level, the sample included all of the major health,
human service, and education agencies serving the city, key departments and
aspects of city government, and major nonprofit organizations and service
providers operating in the community. In total, 48 organizations were included at
this level of the sample.

Names of organizations at the neighborhood level were generated through
city records of registered neighborhood associations; school district listings of
local elementary, middle; and high schools; and local directory listings of faith-
based organizations. From these lists, organizations which fell within Network
Center service boundaries were chosen. In order to focus the sample on the
faith-based organizations with the greatest significance as potential partners, key

informant interviews with the Network Center directors were used to identify the

40

faith-based organizations that had active outreach and programming in at least
one Network Center service area. In order to capture organizational intermediary
roles, the sample also included the Network Centers and staff from local
"Neighborhood Summits” — neighborhood-based planning processes in various
stages of development at the time data was collected. At the neighborhood level,
leaders of each type of organization (i.e., school principals, clergy of faith-based
organizations, and presidents of neighborhood organizations) were chosen as
contacts to represent their organizations.

The total sample for this phase of the study (included in the network
survey) was 95 organizations, including 48 citywide and 47 neighborhood-level
organizations, including 5 Network Centers. Of this number, 70 (73.68%) chose
to participate in the survey. However, network data for the 25 organizations who
did not respond was available via the unconfirmed ties reported by the
organizations who did participate. Participation rates by neighborhood vs. city-
wide areas are shown in Table 1. Organizational roles of the contacts identified

across organizations are presented in Appendix A.

41

Table 1. Participation rates by region.

 

 

Region N Participation rate
Citywide 48 77.08%
Neighborhood A 13 61.54%
Neighborhood B 10 90.00%
Neighborhood C 14 64.29%
Neighborhood D 4 50.00%
Neighborhood E 6 83.33%
Total N 95 73.68%

 

2.2.3 Instrument

The survey instrument used for this phase of the study (attached as
Appendix E) was developed to capture relevant dimensions of interorganizational
relationships between actors at the time of the survey as well as assess which
relationships were new, had grown stronger, had grown weaker, or had dissolved
within the past few years. The rationale for specific decisions in the development
of this survey, including determination of which dimensions of interorganizational
relationships to assess, and the use of valued ties are discussed in the following
sections.

D_efin_inq Relational Ties of Interest Scholars have described
interorganizational relationships as falling into two general types: linkages based
on transfer of resources (both tangible and intangible — e.g., information) and
those based on interpenetration of organizational boundaries (Laumann et al.,

1978), such as in joint planning and programming (e.g.,Aiken & Hage).

42

Correspondingly, collaborative activities are marked by mutual effort and
commitment on the part of stakeholders, most commonly in the form of sharing
tangible and/or informational resources and in joint activity.

Indeed, the sharing of resources in the form of information, money, or
staff, is so common in collaborative efforts that it has been incorporated into
some definitions of collaboration (e.g.,Gray, 1985). In addition, collaboration can
involve a wide range of activities such as partnership meetings, client referrals,
and integrated planning processes. In consultation with community stakeholders,
the three dimensions of information sharing (e.g., informational resources such
as knowledge about needs, program opportunities, or clients), resource sharing
(e.g., money, staff, or other tangible resources), and joint planning/programming
were identified as the most salient with regard to the collaborative activities that
Network Centers facilitate, promote, or bridge. Examples of information sharing
include sharing information about clients, about services offered, future plans,
issues, etc., whether during advisory board or coalition meetings, one-on-one or
small—group meetings, or on the telephone. Examples of resource sharing include
sharing financial resources with another organization, sharing staff time, sharing
volunteers, or sharing space arrangements. Examples of joint planning and
programming include working together with another organization on programs,
planning activities or strategies together, conducting activities (such as outreach)
on behalf of another organization, or helping another organization to write grants

or do other activities. These dimensions have received extensive use in other

43

interorganizational network studies (e.g., Fried et al., 1998; Morrissey et al.,
1997)

Ugaf V_a_lued Ties. Another major decision in sOciaI network analysis
studies is the use of dichotomous versus valued ties (Wasserman & Faust,
1994). In dichotomous ties, the measured relationship type is either present or
absent. In valued ties, the measured relationship type is construed to have some
level of degree or value associated with it—i.e., frequency of contact, intensity of
relationship, et cetera. In the case of shared information, shared resources, and
joint planning/programming between organizations, there are clearly many
degrees of intensity, frequency, etc. that could be applied. For example,
organizations in the service system under study could share general updates
once a year or exchange detailed information about programs, activities, or
developments with clients or the neighborhoods on a monthly or even weekly
basis. Their joint planning and programming efforts could consist of periodic and
minor mutual assistance or could comprise ongoing and tightly coupled
programming and planning activities. Therefore, in situations such as this it is
important to use scale ratings for each type of tie in order to attempt to capture
this variability across the network. The scale used (see survey instrument in
Appendix E) ranged from a rating of zero (for no relationship along this
dimension) to 5 (representing that the participant organization relates to another
organization "to a very great extent" along this dimension). Use of this scale over
dichotomous ties offered the greatest chance of identifying meaningful

differences in interorganizational relationships, as well as the greatest flexibility in

analysis. Although scale data can be easily recoded as a simpler, dichotomous
variable during analysis, dichotomous data can never be expanded to capture
the nuance of the scale. A

Smev structure. These components were brought together in the survey
instrument, which listed organizations within the sample down the left-hand side
and provided corresponding columns for recording the respondents’ ratings for
their relationships with each of the sampled organizations along the three
relational tie types chosen. For each organization listed, respondents were
asked to indicate separately the extent to which their organization presently
engaged in information sharing, resource sharing, and joint planning and
programming with that organization (i.e., “To what extent does your organization
engage in: Information sharing? Resource sharing? Joint planning and
programming?")

In addition, the survey included additional columns for recording a
secondary respondent name1 and for identifying whether the relationship with a
given organization was new, had grown stronger, had grown weaker, or had
become defunct “within the past few years". For example, at the beginning of the
survey, interviewers told respondents:

“Now as I go through each of these organizations, I’d also like you

to mention whether your organization’s relationship with that

organization is new within the past few years, has gotten

significantly stronger, or is significantly weaker than it used to be.

You can also let me know if there is an organization that you used

to have a relationship with but don’t anymore. I’ll ask you more
about these organizations later on in the survey..."

 

1 For use in case the participant reported that they did not know enough about their organization's
relationship with a particular organization listed to respond. No participants gave secondary
contact names for follow up.

45

The purpose of this item was to identify change in organizations’
relationships overtime, with the aim of creating a basis for retrospectively
inferring the role the Network Centers had played in thoSe shifts.

The listing of organizations was grouped into blocks of similar type to help
orient participants, streamline the survey administration, and provide natural
pauses in the questioning to prevent the survey experience from becoming too
tedious for respondents. Each block of organizations was preceded with
anchoring text that described the category (e.g., “city agencies”) of the
organizations that would be asked about next. This constituted the “main”
section of the survey.

In addition, before conducting this network assessment section, the survey
instrument contained a preliminary set of items intended to collect general
background about the participants and their organizations. Background collected
about organizations included type of organization (e.g., human services agency,
local government entity, etc.), primary focus areas of the organization (e.g.,
health, public safety, social service, economic development, education, etc.), and
collaborative groups the organizations participated in. Background collected
about participants included position/role within organization and years with the
organization.

Finally, the survey instrument contained a series of four open-ended
questions designed to follow up on reported changes in the respondent's
relational ties. Each question instructed the interviewer to ask the respondent in

further detail about the organizations with whom they reported having new,

46

stronger, weaker, or defunct relationships (e.g., “You mentioned x, y, and 2 as
organizations that are new relationships for your organization over the past few
years. Thinking back over the list, are there any other relationships you noted
that are new for your organization? Can you tell me a bit about why these new
relationships have formed?") This line of questioning was repeated for each of
the types of change assessed in the earlier part of the survey, with appropriate
shifts in question language to correspond to the type of change being discussed
(e.g., for stronger: “Why have these relationships changed in this way?”; for
weaker: “Why have these relationships grown weaker over time?"; and for
defunct: “Can you tell me what happened there?")

2.2.4 Procedures

Recruitment of Participants. Identified respondents from each selected
organization were first contacted by mail (see Invitation to Participate, Appendix
B). Letters were followed after 1-2 weeks by phone calls from project staff to
assess an organization's willingness to participate and schedule an appropriate
time to administer the survey. To secure the final response rate for this survey, it
was sometimes necessary to contact potential participants four or more times to
schedule these interviews.

Interviews were then scheduled with the identified contact from each
organization. lnforrned consent materials (Appendix C) were sent out to
participants at the time of scheduling and faxed or mailed back by participants
prior to the scheduled time of each interview. In addition, basic information about

the study and interview (see Appendix D) was included with the informed consent

47

form in order to better prepare participants for the survey content and provide
them with a visual reference to aid in efficient response to the survey items.

Interviews. To promote the high response rates needed for valid network
analysis, surveys were administered in structured phone interviews rather than
mailed out to participants. Interviews were conducted by trained undergraduate
staff as well as by myself. Training consisted of a review of the protocol and
interview procedures, several mock interview sessions, and a test of
recording/coding adequacy using a tape-recorded interview. As a final step, the
first interview conducted by each staff member was conducted with myself
present in the room and listening to the exchange, and these interviews were
followed by feedback and clarification of procedures. All staff interviewers
performed well during their supervised interview, and it was determined that
other than remaining available for questions and concerns on an ongoing basis,
no further training was required.

Interviews were structured as per the survey protocol in Appendix E.
Interviewers phoned participants at the scheduled time, and after reading
introductory material about the study as shown, interviewers asked participants
the organizational and participant background questions. Interviewers then read
text intended to orient participants to the social network section of the survey by
framing the ways that organizations might work together, providing examples of
the types of ties measured, and describing the choice points on the scale used.
This material echoed the handouts sent to participants prior to the interview.

Interviews proceeded through the list of organizations, with interviewers providing

48

framing around sets of organizations as shown, prompting participants with tie
types, and providing additional prompting at periodic intervals (at the end of each
section of organizations or the bottom of each page of organizations if there were
no section breaks on a given page) to ask participants if their relationships with
any of the preceding group of organizations had changed. Interviewers then
asked the follow-up questions related to reported changes in relationships,
identified and answered any questions the participant had, thanked the
participant, and ended the call. Interviews generally took 40-45 minutes to
complete.

2.2.5 Data Entry and Management

During the interviews, responses were recorded on paper copies of the
survey. Data entry consisted of entering the responses from the paper surveys
into a series of Microsoft Excel data spreadsheets set up for this purpose.
Worksheets provided a column for participant ID number on the left followed by a
column for each organization included on the survey in the order listed. One
matrix worksheet was created for each relational tie measured (information
sharing, resource sharing, and joint planning/programming) as well as for each of
the recorded change types (new, stronger, weaker, and defunct). Research staff
systematically worked through each of these worksheets, entering for each
survey the tie strengths reported by respondents for each organization. For
change measures, responses were coded in a binary fashion. For example, in
the worksheet recording relational ties that had grown stronger, staff entered
scores of one in columns where a stronger tie was reported and null otherwise.

To check accuracy of data entry, 20% of participating organizations were

49

randomly sampled for rechecking. Data entry accuracy was one hundred
percent on these verified entries.

Data were then copied into new worksheets and entry rows were
reordered to match the ordering of organizations in the matrix columns, creating
an organization-by—organization matrix for each relational measure. This is a
commonly-used format for importing data into social network analysis software.

Because of the brevity of responses to the open ended items on this
survey, specialized software for facilitating qualitative analysis was not necessary
for managing this data, and short answer items were entered into tables in
Microsoft Word for further analysis.

2.2.6 Preliminary Analysis

My intent in analyzing this data was originally threefold: First, I planned to
use identification of cohesive subgroups within the network combined with
analysis of linkages within and between subgroups as a basis for ascertaining
the roles of the Network Centers. The underlying assumption here was that each
Network Center, as a “bridger” in the community system, would be responsible to
some degree or another for connecting links between its own subgroup and other
subgroups of organizations in the system. Second, I planned to use the stronger,
weaker, new and defunct tie data as a means for constructing an approximate
retrospective view of the network, from which change in the network structure
could be identified and follow-up questions about the role of the Network Centers
in fostering such change could be based. Third, based on these first two
analyses I planned to use visual representations of the social networks as a

starting point for interpretation and discussion of both the changes in the

50

community service system and the roles the Network Centers had played in such
changes with selected organizational representatives in qualitative follow up
interviews. I

Unfortunately, this plan withered in the face of the realities my data
presented. First, analysis using Kliquefinder software (Frank, 1995, 1996)
yielded no significant cohesive subgroups in the data for any of the three
relational tie types. This procedures identifies subgroups via a procedure which
iteratively forms and assigns/reassigns actors to subgroups based on an
algorithm which seeks to maximize the odds that two actors assigned to the
same subgroup will have a tie relative to the odds that two actors assigned to
different subgroups will interact (Frank & Yasumoto, 1996). By comparing this
odds ratio (termed “theta1”) with a simulated distribution, Kliquefinder tests
whether the converged subgroup solution is significantly better than what could
be expected from chance assignment of actors to subgroups. As Table 2 below
shows, the theta1 values yielded by this analysis did not approach statistical

significance.

51

Table 2. Approximate tests of concentration of ties - Kliquefinder subgroup
algorithm.

 

TH ETA1 APPROXIMATE APPROXIMATE
SUBGROUP LRT P-VALUE
PROCESSES

 

Information sharing
>= 3
Within 0.19 1.21 0.50
subgroups
based on size
of theta1

Within .16 .82 0.50
subgroups

based on

conservative

predicted value

for theta1

Joint Planning N/A — NO SUBGROUPS FOUND

>=3

 

Second, this lack of subgroups or other clear macro-level patterns in the
local network was visually quite apparent in the early network maps I produced
with both Kliquefinder and the Netdraw routine in UCINET 6 (Borgatti, Everett, &
Freeman, 2002). Presentations of these maps to community partners were met
with good-natured puzzlement: partners wanted to make sense of the dense
webs and locational patterns of nodes in the maps I showed them, but it quickly
became clear that it would be unrealistic to expect any organizational leader to
be able to interpret network-level dynamics in an interview setting, particularly in
the absence of an organizing frame such as subgroup membership.

Finally, the use of the relative change measure in the survey proved
problematic in its own right. As described above, participants were prompted at

the end of each section of organizations in the survey to indicate whether any of

52

their relationships with those organizations had changed significantly in the past
few years. Response rates and distributions for the various relational change

measures are shown in Table 3 below.

Table 3. Analysis of response rates to relational change measure items.

 

Relational Change Measure

New Stronger Weaker Defunct
Organizations citing 26 47 42 20
change (N)
Number of relationship
changes identified by
organizations citing
change
Mean 2.14 6.49 1.76 0.44
Max 48 51 15 3
SD 7.40 10.95 2.41 0.76

 

While examination of the patterns of these ties by neighborhood showed
some slight variability across Network Centers and neighborhoods (see Appendix
F), these differences were not clear enough or sufficient for basing interpretations
of the Network Centers’ roles. In addition, the short open-ended questions asked
of participants in follow-up to reports of changed ties hinted at a wide range of
reasons for underlying shifts in the network‘. For example, ties were created or
strengthened for reasons ranging from increased exposure to other
organizations, increased proximity (e.g., sharing a common location), institutional

pressures and funding incentives, and emergent threats and opportunities.

 

1 Content analysis of these open-ended responses is shown in Appendix Q.

53

Weaker and defunct relationships were attributed to a wide range of similar
influences such as moving offices, staff changes, lack of funds or loss of
incentive/grant funding, and changes in programming priorities. Thus, while
some of these descriptions referenced some Network Center involvement in
relationship development (e.g., gaining familiarity with an organization through
common membership on a Network Center board), it seemed apparent that this
approach would not yield an analysis of the Network Centers' role or influence of
sufficient depth for the study.

Finally, the very generalized nature of the responses requested of
participants on these items made it very difficult - particularly with reports of
“stronger" or “weaker" ties - to know how to interpret these changes in, for
example, interpolating an approximate retrospective network from which to
compare overall changes in the system. That is, while such an approximation
could easily remove the ties reported as “new” by respondents, the ties reported
as “stronger,” “weaker,” or “defunct” raised questions such as: how much of a tie
strength increase would “stronger" represent?

2.2.7 Understanding Bridglng Through Formal Brokerage.

Given the inability to clearly discern network center bridging roles through
an examination of the macro structure of the overall community network,
alternative analytic strategies were sought to examine the role of the network
centers that could address the primary research questions of the study, that is:
To what extent do the Network Centers serve in bridging roles among and
between the neighborhood and city-wide levels of the community service

system?

As a consequence, the brokerage measures conceptualized and
developed by Gould and Fernandez (1989) were identified as an appropriate
means to quantitatively assess the Network Centers' roles in the community
service system network with respect to their functions as intermediaries. This
analysis strategy is described in detail below.

Defining formal brokerage. Marsden (1982, p. 202) defines brokerage as
a process ”by which intermediary actors facilitate transactions between other
actors lacking access to or trust in one another." Gould and Fernandez (1989)
further develop this idea in several ways. First, they identify that brokerage
roles—i.e., roles corresponding to the structure of relationships in a network such
that particular organizations‘ are in a position to broker and facilitate such
transactions — can be formally measured in terms of triadic relationships. A
brokered triad2 consisting of members i, j, and k and represented as r'jk is one in
which ihas a directed tie to j, j has a directed tie to k, and k does not have a
directed tie to i. In this scenario, j is identified as a broker because of the
potential to act as a conduit between the two disconnected organizations, iand k.
Thus, in the conception of Gould and Fernandez, an organization's total
brokerage score consists of the number of brokered triads in which that

organization is in the broker role.

 

1 In actuality, this measure is applied to any type of actor within a network - individual,
organizational, or otherwise. However, for simplicity of explanation, I have framed the definition
and discussion of this measure throughout in terms compatible with its specific application in this
study.

2 Altemately described as an intransitive triple in other literature (See Holland & Leinhardt, 1970,
1978)

55

Second, they introduce the concept and measure of partial brokerage, a
modified measure Of brokerage which assigns organizations "credit" for each
brokered pair in proportion to the number of other organizations also in a position
to broker between that pair. Hence, if three organizations all occupy brokerage
positions between a disconnected pair, each would receive 1/3 point for their
partial brokerage score. Therefore, as Gould and Fernandez write:

“These two measures portray different aspects of the brokerage

roles. Raw brokerage expresses the number of brokerage relations

an actor is capable of mediating whereas partial brokerage

identifies the degree to which an actor controls brokerage relations
in the network.”

Although this sense of “controlling” the brokerage relations is sometimes
used as rationale for examining partial brokerage scores in relation to an
organization’s power within a given network (e.g., Gould & Fernandez, 1994), in
this instance the partial brokerage measures can provide insight in addition to the
rawbrokerage measures by identifying how unique or redundant the Network
Centers' roles might be and, by extension, how important their particular function
is in facilitating connections across the network. Therefore, this study examined
both raw and partial brokerage scores.

Third, Gould and Fernandez propose that all brokerage relations are not
created equal, and go on to distinguish five qualitatively distinct types of
brokerage role based upon the group affiliation (based upon any conceptually
meaningful, mutually exclusive categorization of actors in the network) of the
three parties involved. That is, a brokerage relation may occur: (1) between two
organizations belonging to the same group as the broker, suggesting that the

broker is serving as a coordinator within its group; (2) from an organization

56

outside the broker’s group to an organization within the broker's group,
suggesting that the broker is serving as a gatekeeper of the resources in
question for its group; (3) from an organization within the broker’s group to an
organization outside the broker’s group, placing the broker in a position to serve
as a resource representative of its own group to the external group; (4) between
two organizations that share a common group affiliation apart from that of the
broker, suggesting a role as a consultant (alternately termed cosmopolitan or
itinerant) in the transfer; and (5) between two organizations that are distinct in
group membership from both the broker and from each other. In this instance,
the broker role is construed as that of a liaison, serving to link the two
organizations from an ostensibly neutral position. These five brokerage types
are illustrated in the left column of Figure 1. Correspondingly, any organization’s
total brokerage score can be partitioned into subscores for these brokerage types
depending on the classification of the organizations making up the organization's

brokered triads.

57

FORMAL BROKERAGE FORMAL BROKERAGE - APPLIED TO

 

(GOULD & FERNANDEZ) PRESENT STUDY
’1i;‘\
/ \
, / NC; \ BRIDGING
/ \ WITHIN
/ NEIGHBORHOOD
, r . ,‘\
l (n1, In2/I j LEVEL
l\‘~\ /’ \,_ // //‘

COORDINATOR

\\ BRIDGING
® , BETWEEN

/ NEIGHBORHOOD
/ AND CITY LEVELS
\
GATEKEEPER/
REPRESENTATIVE

 

 

 

 

 

REPRESENTATIVE
o BRIDGING
A AMONG
/,/ \V CITY LEVEL
I 0 0 »
\_—_,//
CONSULTANT CONSULTANT
0 NOT
USED
LIAISON

 

 

Figure 1. Types of formal brokerage measures as defined by
Gould & Fernandez (1989) and applied in this study.

58

Because this measure allows use of subgroups defined by any
conceptually meaningful criteria — whether sociometrically defined clusters or
categorical attributes such as the organization levels of interest in this study -
this conceptualization can readily be applied to this study’s research questions by
appropriate partitioning of the network into subgroups based on whether an
organization operates at the city-wide level or the neighborhood level, and further
identifying neighborhood-level organizations (including each Network Center) by
the Network Center service region (i.e., neighborhood area) in which they are
based‘. By partitioning a network for each Network Center comprising only the
ties among City-level organizations and the neighborhood-level organizations in
that Network Center’s neighborhood (the process of which is described in further
detail below), it becomes possible to use four of the five types of brokerage role
described by Gould and Fernandez to measure the brokerage roles for each
Network Center in three ways, as illustrated in Figure 1. Specifically:

Network Center scores for the coordinator type of brokerage role —
representing, in this instance, brokerage between two neighborhood-level
organizations in their region (indicated as n1 and n2 in the figure) — can help
identify the extent to which the Network Centers are bridging among

organizations at the neighborhood level.

 

1 This categorical representation of the network, in addition to matching the interest of the study,
is particularly appropriate here in light of the absence of cohesive subgroups found in preliminary
analysis. If cohesive subgroups had been found, this would have suggested a need to closely
examine subgroup membership in conjunction with neighborhood or City-wide categorical
affiliation to identify appropriate classifications that would both serve the study focus and allow
appropriate interpretations that respected the network's macrostructure.

59

Network Center scores for the combined values of gatekeeper and
representative roles — representing brokerage relations in which a Network
Center is in a brokerage position between a neighborhood-level organization and
a city-level organization (indicated as n1 and c1 in the figure, respectively) —
address the extent to which Network Centers are in a position to bridge between
organizations at the neighborhood and City levels.

Finally, Network Center scores for the consultantfltinerant type of
brokerage role — representing brokerage relations between two organizations
that both operate at the city-wide level (indicated as c1 and c2 in the figure) -
address the extent to which Network Centers can serve to bridge among city-
level organizations.

Brokerage score calculation procedures. As referenced above,
preliminary steps in calculating the brokerage scores for this data included
establishing group partitions and reshaping the networks to be suitable for
analysis. First, I established non-overlapping subgroups by coding each
organization in the overall sample into a "region" category indicating its status as
either a city-wide organization or an organization operating at a neighborhood
level in one of the five neighborhood areas, with individual codes for each
neighborhood. As organizations that operate and are based in their focus

neighborhoods, the Network Centers were coded as neighborhood-level entities‘.

 

1 It can be noted that although grouping the Network Centers with neighborhood-level
organizations seemed the more conceptually appropriate choice, from a statistical perspective
this particular decision is inconsequential. Grouping the Network Centers with city-level
organizations would have no effect on the brokerage measures themselves, but would simply
switch the application of coordinator and consultant/itinerant brokerage measures with regard to
answering the research questions.

60

Second, for analysis purposes, I split the network into sub-networks representing
each network center's effective scope for bridging relationships of interest using
the UCINET "Subgraphs from Partitions" routine; each Network Center's
subnetwork consisted of the neighborhood-level organizations within a Network
Center’s particular region and the entire set of city-level organizations. This
partitioning allows brokerage scores to be calculated for each Network Center
based upon the actual bridging dynamics of interest (that is, among
neighborhood-level organizations, among city-level organizations, and between
organizations at these two levels) without confounding the scores with
relationships that the network centers might be in a position to broker due to
secondary relationships with organizations based in other neighborhoods.
Subsequent to this stage of analysis, I determined that further analysis
should be limited to three of the original five Network Centers. During the
analysis process, one of the Network Centers was disbanded by the Citywide
Network Center board, and it became evident that another was in a state of
decline‘; in addition, after initial data collection the latter Network Center had
relocated to another part of its geographically large service area. This part of the
City developed much later and in a more dispersed manner than the other
neighborhood areas included in this analysis; as such, organizations are spread
over a large territory. Because of this, the small initial set of organizations that

had been included in the sample as relevant for that Network Center’s work in

 

1 This Network Center is now also defunct.

61

 

the area it most directly served1 and the corresponding ties between those
organizations became appreciably less appropriate and valid for analysis in the
new context. As such, analysis proceeded with the three remaining Network

Centers. These partitioned networks are described in Table 4.

 

1 The catchment area for the elementary school which housed that Network Center’s facility

62

Table 4. Number of organizations and participation rate by type and Network
Center partition.

 

 

 

Network Network Network
Center A Center B Center C
City-level organizations
City/county 15 (73.3%) 15 (73.3%) 15 (73.3%)
government
Agencies/Institutions 14 (92.9%) 14 (92.9%) 14 (92.9%)
Nonprofit 16 (75%) 16 (75%) 16 (75%)
Organizations
Other1 3 (33.3%) 3 (33.3%) 3 (33.3%)
Subtotal: City level 48 (77.1%) 48 (77.1%) 48 (77.1%)
Neighborhood-level
organizations
Neighborhood 6 (50.0%) 2 (100%) 5 (60.0%)
associations
Schools 3 (66.7%) 4 (75.0%) 3 (33.3%)
Faith-based 2 (50.0%) 3 (100.0%) 1 (0%)
organizations
Network Centers 1 (100%) 1 (100%) 1 (100%)
Other2 1 (100%) 0 . 4 (100%)
Subtotal: 13 (61.5%) 10 (90.0%) 14 (64.3%)
Neighborhood level
Total Sub-Network Size 61 (73.8%) 58 (79.3%) 62 (74.2%)

 

1At the city level, “other” organizations comprise one school, one faith-based organization, and
one collaborative group that operated on a cross-cutting level and therefore could not be
categorized into a particular neighborhood region.

2At the neighborhood level, “other” organizations included one other neighborhood-based
collaborative planning organization in each of Neighborhoods A and C, and an additional 3
nonprofit organizations that were found to base their work predominantly in Neighborhood C (one
community center and two housing-focused community development corporations).

 

Several additional decisions were required to implement the Gould and
Fernandez brokerage measures with this data set. First, brokerage measures,
as developed by Gould and Fernandez, are calculated based on the binary

presence/absence of ties between actors in the network. As is the case with

63

other social network measures calculated on binary data, valued tie data can
readily be used but must be dichotomized at a selected cutoff level. The task of
determining an appropriate cut-off value is typically conceptual rather than
statistical in nature. In the present study, the conceptual criteria used was: What
is the minimum tie strength that could be reasonably expected to produce or
facilitate collaborative outputs such as transfer of knowledge and coordinated
action to a significant degree? Correspondingly, what is the tie strength that,
below this level, two organizations could be reasonably expected to benefit from
the connections of a third-party intermediary with stronger connections to both?
Anecdotal evidence from participant responses to the survey suggested that
ratings of 1 (“minimal”) and 2 (“a little") on the relational measures generally
corresponded to rather superficial levels of communication and exchange
between organizations (e.g., exchanging newsletters or attending the same
meetings without additional efforts to collaborate) and intuitively seemed subject
to response bias favoring reporting a minimal tie versus acknowledging lack of a
genuine relationship with another organization. Therefore, a rating of 3
(corresponding to a degree of “somewhat” on the scale of 0 = “none” to 5 = “to a
very great extent”) was selected as an appropriate cutoff value. In addition, a tie
value of 3 was determined to be the level at which a given triad in the overall
network had a similar chance of being either brokered or fully transitive (i.e., all
three organizations connected by a tie) (see Appendix G for calculations of
brokered versus transitive triads at varying dichotomization points); as such it is

reasonable to infer that selecting this value would strike a balance between

64

potentially overstating or understating brokerage in the network. For example,
consider a triad of organizations i, j, and k, with relational tie X measured
between each. At a dichotomization threshold of two, iwould be identified as a
broker if Xij and Xik were each greater than or equal to two and only if Xjk was
less than or equal to one. But Xjk has a higher probability of being greater than
the threshold, resulting in the likelihood that across the network triples will tend to
be fully transitive (all three connected) rather than brokered. On the other hand,
at a dichotomization threshold of five, organization i would be identified as a
broker if Xij and Xik were each rated at strength five, and Xjk were any tie
strength less than five. This condition would both place very stringent criteria on
what tie strength was considered noteworthy and would, in my opinion, under-
account for potential transitivity in the network’s triads.

Second, the brokerage measure as developed by Gould and Fernandez
assumes that ties are directed, allowing Clear designation of two-step paths, Via
brokers, between primary senders and ultimate receivers. However, given that
relations were measured in a nondirectional manner (e.g., “To what extent do
you engage in information sharing with this organization?” and “To what extent
do you engage in joint planning or programming with this organization?”),
brokerage calculations on this network must be performed to assume that ties
are bi-directional - that is, any given reported tie assumes that both actors are
simultaneous senders and receivers. This is accomplished by symmetrizing the
networks using the maximum reported value for each pair. Thus, the resulting

binary, symmetrized sociomatrix for information sharing reflects a tie between a

65

given organization iand organization j if either organization ior j reported sharing
information with the other to an extent warranting a rating of "somewhat" or
higher. Likewise, the resulting binary, symmetrized sociomatrix for joint
planning/programming reflects a tie between a given organization iand
organization j if either organization for j reported engaging in joint planning and
programming with the other to an extent warranting a rating of "somewhat" or
higher.

For this study, raw brokerage scores were calculated using the
“NetworklEgo NetworkslBrokerage” routine in UCINET 6 on the relations of (1)
information sharing and (2) joint planning and programming. To gain an
understanding of both the total capacity of the Network Centers to broker and the
uniqueness of their brokerage roles, both unweighted and partial brokerage
scores were calculated, as defined above.

I From the raw brokerage scores, standardized brokerage scores were
calculated in order to enable comparisons both between the three Network
Centers and between the three types of brokerage roles they could occupy. As
described by Gould and Fernandez (page 102):

"An obvious problem with these measures is that their values

depend on the size of the network being examined and on the

subgroup partition being used. For example, an actor in a

subgroup with twenty members has many more opportunities for

[coordinator] brokerage than an actor in a subgroup with only three

members, because there are many more paths on which the first
actor might lie."

To address this challenge to interpretability, Gould and Fernandez devised
calculations for standardized brokerage scores based on a null model of random

tie distribution given the density of the network. These calculations provide both

66

expected values and estimated variances for the distributions of the various
brokerage scores under the null model, creating a test statistic that assumes a
standard normal distribution. As such, standardized scOres with an absolute
value >= 1.96 are significantly greater than one would expect under chance given
a random distribution of ties in the network. Standardized brokerage scores can
not only be compared across actors in different subgroups in the same network,
but can also be compared across entirely different networks. Given both the
separation of the network into subnetworks for analysis and the difference in the
number of neighborhood-level organizations included in each subnetwork, use of
standardized scores for interpretation is essential to interpretation.

Estimated values and variances to compute standardized brokerage
scores were calculated in the R statistical programming language (see Appendix
I for script) using the formulae presented in Gould and Fernandez (1989), but
doubling the variance of individual measures to allow the null model to account
for complete symmetry in ties as discussed on page 104 of the same article.

Additional analysis of brokerage rofi. While formal brokerage scores can
speak broadly to brokering between organizations in the various defined
subgroups, it cannot illuminate with any further clarity the characteristics of who
is being brokered by whom. To this end, a function was written in the R
statistical programming environment to identify each triad brokered by a Network
Center and classify each triad by the types of organizations brokered. Briefly,
this function builds on the triad Classification function in the R social network

analysis library by generating a set of all possible triadic combinations of actors

67

in the network, testing each for a brokered triadic configuration using the Davis
and Leinhardt (1972) classification of triad types, identifying the broker in each of
these triads, and creating an output file consisting of the identified brokered
triads. From this output file, triads brokered by each of the Network Centers
were extracted and cross-tabulated based on categorizations of organizational
type for the other members of the triads, yielding counts of the number of ties
each Network Center was in a position to broker between each type of
organization in their subnetwork. These figures were examined with summaries
of actual direct ties and total possible ties for each combination of organization
type in their subnetworks to yield comparisons of the potential contribution of
Network Centers’ brokerage capacity to the overall connectivity of different
combinations of organizations in the system.

2.3 Phase II—Qualitative Interviews

Given the analyses described above, the purpose of Phase II was to
collect data that could aid in interpreting and validating the brokerage scores and
to answer research question #4: Assuming neighborhood network centers do
occupy one or more of bridging roles, how do they act out such role(s) ?

In order to accomplish this, I conducted open-ended, semi-structured
qualitative interviews with the Network Center directors and the Citywide Network
Center Coordinator (n=4). These interviews were comprised of three general
sections. In the first section, participants were asked in open-ended fashion to
discuss their role and their work in the community system. In addition, the

concept of bridging roles inherent in this study was explained to participants, and

68

they were asked to identify which, if any, aspects of their work involved
consciously acting in such a bridging role, and how they served in those roles if
they did indeed exist for them. (e.g., “To what extent do you think that the
network center serves in this type of role? Can you give me some examples?
Are there particular types of organizations that you see this role strongly with?
Are there particular areas in which you think this operates very strongly?”)

In the second section of the interview, I presented each participant with
preliminary findings for their Network Center (or, in the interview with the
Citywide Network Center Coordinator, with findings for all of the included
Network Centers) based on the brokerage analyses described above. (e.g., Do
you experience the network center as brokering in these ways? How so? How
not? In what ways are you brokering between these types of orgs? Where do
you think there are missed opportunities?) Together, these two sections served
as a “reality check” on the brokerage score data, by yielding a general sense
(although not by any means a precise measure) of whether and how the capacity
for brokerage illustrated in the quantitative data reflected the way that they
described their own activities in the community.

Finally, the third section of the interview included, as time allowed,
discussion of what served to help that network center in its brokering, what
barriers it encountered, and what advice it would offer to other communities
considering a network center model (e.g., “What helps you to use your role to
effectively bridge orgs? What gets in the way of bridging? What kinds of lessons

would you share about being in this sort of role? What advice would you give to

69

a new network center?" The protocol that was used as a basis for these
interviews is attached as Appendix H. Interviews were conducted in person at
locations of each participant's choice and each lasted between 1 and 2.5 hours.
Interviews were tape recorded and transcribed. Preliminary thematic
coding was performed in ATLAS.ti software for facilitating qualitative analysis.
Preliminary coding had no predefined frame. Upon completion of preliminary
coding, coded sections of data were output and re-examined; through this
process, a frame was developed for second-order processing of case accounts.
This frame included: (1) Examples/validation/case studies of brokering/bridging
(or the contrary); (2) What makes “it” work?; and (3) What's different/unique
about this Network Center? Visual representations of the elements of these
themes were constructed in table format (see Miles 8. Huberman) and
diagrammatic form to facilitate comparison and integration of thedata into
meaningful interpretations (see Appendix P for examples of these). Broad
themes from these analyses were presented at a Citywide Network Center Board
meeting, and additional member Checks were conducted informally with Network
Center staff and two other key informants to authenticate specific emergent

themes.

70

CHAPTER 3
RESULTS

3.1 Overview

The following sections present the results of both quantitative and
qualitative analyses of the three Network Centers’ roles as bridging agents
between and among neighborhood- and city-level organizations in the local
community service system. Section 3.2 presents quantitative results from the
brokerage score analyses coupled with some confirmatory and explanatory detail
from the qualitative follow up interviews. Section 3.3 delves further into the
qualitative data to present case summaries of the three Network Centers, explore
the various ways in which they operate as brokers, and describe aspects of their
functioning that suggest meaningful differences between them.
3.2 Identification of Formal Brokerage Roles

As described in Section 2.2.7, formal brokerage scores for each of the
three Network Centers were calculated and compared to provide a framework for
answering the first three research questions:

1. To what extent and how do neighborhood network centers play a
bridging role in fostering connections among neighborhood-level
organizations?

2. To what extent do neighborhood network centers play a bridging
role in fostering connections between neighborhood-level

organizations and system-level organizations?

71

3. To what extent do neighborhood network centers play a bridging
role in fostering connections among city-wide/system-Ievel
organizations?

In the following sections, I address each of the first three research
questions in turn, drawing from Phase II qualitative interview data to support
and/or question brokerage score findings, where applicable. These brokerage
scores are summarized in Table 5 (tables of brokerage scores for each Network
Center’s subnetwork can be found in Appendix K) and further described and
explained below. This table presents standardized scores for both unweighted
and partial brokerage calculations for the three brokerage types of interest
(coordinator, combined gatekeeper/representative, and consultant/itinerant, as
previously explained, presented by individual network center and type of
relational tie. As described in Section 2.2.7 above, use of standardized
brokerage scores allows clear comparison across organizations in the network
(e.g., comparison of Network Centers to each other and to other organizations in
the community) as well as across types of brokerage role (e.g., determining the
relative significance of one brokerage type over another in the overall role an
organization plays). Standardized brokerage scores outside of +/- 1.96 are
significant at the alpha = .05 level (indicated by a single asterisk in Table 5), and
standardized brokerage scores outside of +/- 2.57 are significant at the .01 level
(indicated by a double asterisk In the table), indicating brokerage scores of
sufficient magnitude to reject the possibility that such scores occurred via the

chance distribution of ties in the network.

72

Table 5. Standardized brokerage scores by Network Center, relationship type,
and brokerage type

 

 

 

 

Coordinator Gatekeeper/ Consultant!
Brokerage1 Representative ltinerant
Brokerage2 Brokerage3
Un- . Un- . Un- .
weighted Pama' weighted Pama' weighted Pam?"
x < Information 6.38“ 12.39“ 22.63““ 41.20" 0.51 0.43
at t: Sharing
O LIJ
E 5 Joint Planning/ 7.01" 10.48““ 17.41“ 25.76“ -0.33 -0.62
g 8 Programming
)6 an Information 1.57 3.54“ 6.29““ 12.85“ -1.03 -0.94
m c: Shanng
0 LU
E '2 Joint Planning] 2.58“ 4.26““ 10.66““ 16.01“" 0.33 -0.48
g (“3 Programming
x 0 Information 5.02" 8.39"" 35.75“ 54.26““ 6.28““ 8.46““
I: o: Shanng
O LIJ
E '2 Joint Planning/ 0.30 0.22 7.06“ 6.94“ 1.84 2.38“"
Lg 8 Programming

 

1 Indicates formal brokerage among neighborhood-level organizations

2 Indicates formal brokerage between neighborhood and city-level organizations
3 Indicates formal brokerage between city-level organizations

* Scores significant at .05 level

** Scores significant at .01 level

3.2.1 Brokerage at the Neighborhood Level

Given the Chosen approach for calculating brokerage scores for this
network, the Network Centers’ scores for coordinator brokerage were examined
to address research question #1: To what extent and how do neighborhood
network centers play a bridging role in fostering connections among
neighborhood-level organizations?

An inspection of these scores indicates that all three Network Centers fill

some brokerage role at this level, in varying forms. Network Center A has the

73

largest brokerage role between neighborhood-level organizations of the three
Network Centers for both information sharing and joint planning and
programming (standardized partial brokerage scores of 12.39 and 10.48,
respectively). Network Center B also had significant brokerage scores of this type
for both relational ties, with standardized partial brokerage scores of 3.54 for
information sharing and 4.26 for joint planning and programming. In contrast to
B, Network Center C had a higher score for information sharing (8.39), but a
nonsignificant score for joint planning and programming (0.22). As Table 5
shows, these patterns hold as well for unweighted brokerage scores. These
numbers suggest that while both Network Centers A and B (and particularly A)
are highly active intermediaries in facilitating both information flows and
coordinated planning among neighborhood-level organizations, Network Center
C’s neighborhood-level role is focused almost entirely around sharing
information, with little role in coordination of planning or programming among
disconnected neighborhood-level organizations.

The analysis of brokerage relations among types of neighborhood-level
organizations (as illustrated in Figure 3) is further illustrative of these differences.
This figure shows patterns of direct ties (darker shading) and brokered ties
(lighter shading) as a percentage of possible ties between each type of
organization (each solid line represents 10%). Each segment of the figure
represents one combination of organizations, as identified by number in the table

below:

74

Table 6. Key for combinations of organizations shown in Figure 2.

 

1

Neighborhood Associations by Neighborhood Associations

 

 

 

 

 

 

2 Neighborhood Associations by Schools

3 Schools by Schools

4 Schools by Faith Organizations

5 Faith Organizations by Faith Organizations

6 Neighborhood Associations by Faith Organizations

 

 

This allows quick visual comparison of direct connectivity and brokerage

capacity across Network Centers and between information sharing and joint

planning and programming ties.

75

 

Figure 2. Patterns of direct and brokered ties among neighborhood-level
organizations by Network Center for Information Sharing (Info) and Joint
Planning and Programming (JPP).

76

 

As this figure shows, both Network Center A and B are consistent in their
relational activities when comparing their information sharing and joint planning
relationships at the neighborhood level. Network Center A's brokerage capacity
is distributed across combinations of brokered organization types, but is greatest
in triads that include neighborhood associations (labels 1, 2, and 6 in Figure 2).
The heavy involvement of Network Center A with local neighborhood
associations as a basis for much of its activity was confirmed by its director:
"You know, we really believe that the sine qua non of neighborhood
improvement is an organized neighborhood—highly networked,

highly social, and savvy about institutional resources and able to
advocate for themselves." - Director A

Network Center B, on the other hand, exhibits less brokerage capacity at
the neighborhood level than Network Center A. However, an examination of the
distributions of direct and brokered ties in that neighborhood suggests that this
may be more related to the higher level of strong direct connections in this
neighborhood — particularly those involving schools and faith-based
organizations‘ - rather than a lack of intensive involvement of Network Center B
itself. It is instructive to note that while Network Center A's advisory board is
composed almost entirely of the presidents of local neighborhood organizations
and the principal of a local elementary school, Network Center B uses a more
diverse array of neighborhood-level stakeholders in its advisory activities and
draws heavily on local faith-based organizations to help administer aspects of the

neighborhood’s social services programs, as described in one interview:

 

1The 100% direct connection between neighborhood associations in this neighborhood may be
slightly misleading, since this figure represents a single tie between only two neighborhood
organizations in that neighborhood area in the sample, versus the much higher number of
available neighborhood associations in Neighborhoods A and C.

78

"[Network Center B] has faith-based institutional members on her
board even, but she has strong enough relationships with them that
when people are in financial crises in the neighborhood, she can
turn to the local missions committee of the churches within her area
and they will provide cash assistance if necessary to help pay a few
bills or buy a tank of gas or provide food from their food bank, or
provide volunteers to assist with a community project or supply
Thanksgiving baskets for families who need them... They also sit
round the table when called and help with the development of
proposals for bringing money into the community to help solve
community problems." - Citywide Coordinator

The difference in Network Center C's brokerage capacity between
information sharing and joint planning relational tie types, however, is quite
dramatic. As illustrated in the comparison plot, the vast majority of Network
Center C’s brokerage relations for information sharing involve ties between
schools and other organizations, and indeed, the only brokerage role played by
this Network Center C for joint planning processes involved schools. Director C
confirmed this high level of involvement with the schools during the qualitative
follow-up phase, and indicated that work with the neighborhood schools and their
students was a focal element of its programmatic activities:

"We work with the schools on a regular daily basis. Like the after

school programs. We have a group of students that come here in

the afternoons from [two of the local schools] and they were getting

failing grades, and now they’re getting all A’s and B’s...because

they're getting after-school training here. If their grades were bad

before they came here, the school certainly wasn’t improving them

because they don’t have the time to spend one on one with the
child like we do here."

One of the differences between the Network Centers’ brokerage roles is
evident in relative prominence of neighborhood associations across the Network
Centers' distributions of brokerage scores. The service areas surrounding these

three network centers differ vastly in their levels of social organization. Network

79

Center A both enjoys and actively develops a highly sophisticated level of
neighborhood organization, featuring not only high number of organized
neighborhoods, but a 30-year-old umbrella neighborhOOd organization that brings
them all together and creates a larger, unified power base for neighborhoods
across that part of town:

"...We live in a neighborhood that has an activist history, that huge
portions of it are wall-to-wall neighborhood organizations, that has
already a significant social infrastructure. And to the extent that we
have really focused on strengthening and deepening that social
infrastructure—Le, doing leadership development among emerging
neighborhood leaders, providing technical assistance to organized
neighborhoods and writing grants and addressing issues -- whether
drug houses or deteriorating housing or nuisance dogs or crime or
any of the myriad issues that urban neighborhoods struggle with —
that makes our work easier. You know, so—and that's a factor. I
think that the fact that we're standing on the shoulders of people
who in 1970 created the [umbrella neighborhood organization]—
you know, this 30-year old advocacy organization. And then
continued that through [neighborhood] summit activities and
through that period in the 80's and early 90's where we saw
proliferation of neighborhood watches. And [Network Center A]
stands on the shoulders of people and [Neighborhood A] initiatives
for the last several decades.” — Director A

As described by this Network Center’s director, this extant capacity has a
profound impact on how the Center can operate within the neighborhood as well
as on its behalf:

"It makes our work as brokers much easier, because we can build
on that social infrastructure. We can actually, we can launch a
project using that already existing infrastructure...When we do just
about anything here, one of the first things that we do is figure out
how the organized neighborhoods fit into this game plan. [Network
Center A] is governed by presidents of the nine organized
neighborhoods that make up its most immediate service area, one
at-large member and then the principal of [the nearby elementary
school]. So we are embedded so much in the neighborhood. But
again, what makes a lot of what we do possible, is that there is a
social infrastructure—that there is this other [neighborhood-level]

80

network of organizations, of entities that we coordinate with.” —
Director A

In contrast, the neighborhoods surrounding Network Center B are
disorganized and lacking in leadership. This explains their strong connection to
faith-based institutions and why their advisory board is more diverse. Although
there is a high degree of resident participation in Network Center-based
activities, the local neighborhood associations remain relatively passive with
regard to neighborhood-based work, and engagement in neighborhood-based
problem solving and planning typically is not coordinated through neighborhood-
based organizations:

"Neighborhood leadership in [Network Center B's] area is stagnant.

[The director] works really hard to bring new community people to

the table based on the issue. But the leadership there is pretty
stagnant.” — Citywide Coordinator

"I have gotten agencies and other people to get involved over there
[in one of the surrounding neighborhoods]. But that neighborhood is
so divided now, it's hard to work in there...right now it's rebuilding."
— Director 8

This lack of neighborhood capacity shapes Director B's activities as well,
leading this director to invest additional energies in raising neighborhood-level
awareness of issues and rounding up neighborhood stakeholders to bring to the
table. Interestingly, Director B expressed concerns that in the absence of efforts
to develop local neighborhood organizations, the strong presence and high level
of activity of the Network Center was perhaps having a negative impact on the
level of local neighborhood organization by enabling a “free rider" effect:

"...And I think that with [one of the local neighborhood associations],

since [Network Center B’s] been here, they've grown a little lax in

trying to go out and find things for their community. They’re saying,
‘Well, the Center's there, the Center will take care of it for us.'...l've

81

noticed that the attendance went down at [their] meetings...and I
held a neighborhood forum meeting a couple of months ago, and
sent out 800-something newsletters, and announced that we’re
having a neighborhood forum...And so I announced at that meeting
that there was a neighborhood organization meeting. So many
people were not aware in the community about neighborhood
organization meetings...and when I announced it, their attendance
rose the very next night, and then this month they had a very high
attendance again. And so I think that they’re hurting themselves by
relying on us to do everything.” - Director B

Network Center C experiences yet another set of Challenges with regard
to neighborhood social infrastructure. While the neighborhoods in that part of
town, like those surrounding Network Center A, contain a large number of
neighborhood organizations, these organizations are highly factionalized and
engaged in power struggles amongst themselves:

"It's been difficult...to consistently bring neighborhood leaders to

the decision making table at [Network Center C], and that's a

political thing...between leaders and each other. Between the
neighborhoods on [that side of town].” - Citywide Coordinator

"YOU know, they've got turf wars and 'Somebody's gonna do this,’
and 'I'm not going to do that if they're going to be there' and so
there's all this political crap—that's just the way it is. Life in the city.”
— Citywide Coordinator

This means that, in contrast to the passivity encountered by the director of
Network Center B, the director of Network Center C must sometimes grapple
with neighborhood leaders who may perhaps perceive the involvement of other
neighborhood-based entities as an intrusion on their "turf." As a result, working
with - and brokering on behalf of — neighborhood organizations in this part of
town might be particularly challenging for this Center; local schools may, by

comparison, provide much more accessible partnership opportunities.

82

There is an interesting contrast between what one might think about the
connectedness of neighborhood associations in Neighborhood A based on this
neighborhood’s strong social infrastructure and the connectedness of these
organizations as evidenced in Figure 2. An examination of the reported ties by
neighborhood and organization type indicates that the Neighborhood A
neighborhood associations that did participate had both a lower mean outdegree
(i.e., the number of organizations they reported having any ties to) and lower
mean nonzero tie strength (i.e., the average rating of reported ties) than
neighborhood associations in the other two neighborhood areas (see Appendix L
for analysis). This combined with the high degree of centralization in the
Neighborhood A network suggests that this pattern may not be a fluke but may
instead represent a genuine feature of the neighborhood’s social structure. That
is, one of the aspects of this neighborhood’s high capacity with regard to its
social infrastructure may lie in the way it has structured itself through weak ties
among neighborhood-based organizations connected through the three key hubs
of the Network Center, the local summit process, and the umbrella neighborhood
organization.

3.2.2 Brokerage Between Neighborhood and City-Level Organizations

Brokerage scores for the combined “gatekeeper/representative" brokerage
role were examined to address research question #2: To what extent do
neighborhood network centers play a bridging role in fostering connections
between neighborhood-level organizations and system-level organizations?

As these numbers indicate, all three of the network centers occupy

significant positions of brokerage as both infonnation-sharing and joint planning

83

intermediaries between neighborhood- and city-based organizations. Network
Centers A and B show standardized partial scores for information sharing of
41.20 and 12.85, respectively. The gatekeeper/representative score for Network
Center C for information sharing was even higher, at 54.26. Interestingly,
however, Network Center C has the lowest score for this type of brokerage for
joint planning and programming (6.94). In contrast, Network Center A has the
highest standardized brokerage score of this type at 25.76, followed by Network
Center B at 16.01.

Network Center C's substantial drop in brokerage capacity is Visually
evident, once again, in the examination of plots showing the distribution of
Network Centers’ brokerage capacity between types of neighborhood and city-
Ievel organizations (Figure 3). As with the previous figure, this figure graphically
Views the breakdown of direct and brokered ties by combination) of organization
type, with darker shading indicating direct ties between organizations of that type
and lighter shading indicating Network Center brokered ties. A key for the

organization combinations by number is shown on page 85.

Table 7. Key for combinations of organizations shown in Figure 3.

 

1

Neighborhood Associations by Agencies/Institutions

 

 

 

 

 

 

 

 

 

2 Neighborhood Associations by City/County Government Offices
3 Neighborhood Associations by Nonprofits

4 Schools by Agencies/Institutions

5 Schools by City/County Government Offices

6 Schools by Nonprofits

7 Faith Orgs by Agencies/Institutions

8 Faith Orgs by City/County Government Offices

9 Faith Orgs by Nonprofits

 

 

85

 

Figure 3. Patterns of direct and brokered ties between neighborhood and city-
level organizations by Network Center for Information Sharing (Info) and Joint
Planning and Programming (JPP).

86

 

As described above, this Network Center’s gatekeeper/representative
brokerage scores do retain significance levels for joint planning and
programming. However, it is apparent that although this Network Center is
involved in an extensive array of information sharing relationships, its brokerage
role for the more intensive joint planning relational ties is again highly
concentrated in relations involving local schools.

Network Center A and B show a somewhat more even pattern between
their information sharing and joint planning brokerage roles between
neighborhood and city-level organizations. Network Center A’s most dramatic
brokerage roles appear to be between neighborhood associations and both
agencies and local government. The brokerage role between neighborhood
associations and agencies appears particularly high and consistent between
information sharing and joint planning/programming ties, suggesting that bridging
between these stakeholders is a particularly active aspect of Network Center A’s
agenda. Network Center B, on the other hand, shows less-dramatic levels of
brokerage role “contribution" to the connectedness of organizations in that
neighborhood to city-wide organizations, but these brokerage roles are
distributed more evenly across all of the possible combinations of these
organization types. This suggests that Network Center B is both less focused on
neighborhood association involvement in particular, incorporates a broad base of
neighborhood-level organizations in its work, and distributes its bridging activities

widely across the available resources in the broader City-wide service system

88

These broad patterns were largely confirmed in the qualitative follow-Up
interviews with Network Center directors and the Citywide Network Center
Coordinator. When asked to give examples of the types of bridging activities it
performs between neighborhood- and city-level organizations, Director A stated:

"It's bringing people in this neighborhood—in organized
neighborhoods particularly—face to face with institutional
representatives, is I think a particular specialty.” - Director, Network
Center A

"I think we're real effective in bringing, face-to-face, neighborhood
organizations with the institutions that have major impact on their
lives—whether housing or health-related. I think our work with
organized neighborhoods is pretty robust. — Director Network
Center A

As the formal brokerage scores indicate, Network Center B is similar to
Network Center A in that it plays significant brokering roles for both information
sharing and joint planning and programming between neighborhood-level
organizations and partners at the City level. This finding was corroborated by the
examples and evidence provided by Director 8 in the follow-up interview.

"I think that building those relationships [bridging citywide &
neighborhood organizations] are very important and now that I’m
sitting here thinking about it, I do do that all the time. I did not
realize that. I mean, because, if I met with [organization members]
and they give me information, I then therefore am going to
[neighborhood organizations], you know, repeating what they’ve
offered, what can they do, and then I sit on the Mayor’s Task Force,
and they’re very interested in this community now, so they’re asking
me about all the different agencies and people in the community
that are willing to work on projects, and what can they do, and then
the Mayor’s saying, ‘Now what can I do to help?’ or they’re telling
other people to come here and help us." - Director B

Similarly, the examples of bridging activity proffered by Director C
confirmed its heavy involvement with and focus on working with schools at the

neighborhood level:

89

"...The schools they have a certain curriculum that they follow. They
have their procedures and programs, and these agencies that
you’re talking about, they have their programs. And so we're here
to help the agencies and the schools become more incorporated, to
work together in a greater sense. Where the information can be
shared with the schools’ participants and the agencies’ participants.
So certainly there are some missed opportunities through lack of
knowledge. And that’s what we’re here for, to bridge that gap."

3.2.3 Brokerage Among City-Level Organizations

Brokerage scores for “consultant/itinerant" brokerage role were examined
to address research question #3: To what extent do neighborhood network
centers play a bridging role in fostering connections among city-wide/system-
level organizations?

As Table 5 shows, neither Network Center A nor B occupied formal
brokerage roles of the consultant/itinerant variety to any significant degree. In
fact, for relationships based on joint planning and programming, Network Centers
A and B showed slight (although not statistically significant) tendencies away
from occupying a brokerage position between city-level organizations, suggesting
that city-level organizations with which these Network Centers have strong joint
planning/programming relations tend to have strong relationships among
themselves as well.

However, Network Center C does show significant positive scores for
consultant/itinerant brokerage. This holds true for both information sharing and,
albeit to a lesser degree, joint planning/programming ties when partial brokerage

scores are examined.

3.3 One Model, Three Manifestations

90

As described above in Section 2.3, qualitative interviews with the Network
Center Directors and Citywide Network Center coordinator were intended to
provide a general “face” validation for and contextualize the brokerage scores
identified through the social network analysis phase. Although the overall
Network Center model was never intended to be a "cookie cutter" approach, the
three cases presented here and their corresponding patterns of formal brokerage
solidly demonstrate that these Network Centers, while all brokers and
intermediaries in their own way, are also quite different from each other. This
overall comparison is summarized in Table 8 below.

Interestingly, interviews suggested that the three Network Centers
examined here all play different roles in their work as neighborhood-based
intermediaries and service providers. In the next three sections, I take each of
these Network Centers in turn and present data to explain these various roles

and how these Network Centers function in them.

91

Table 8. Comparison of key features of Network Centers and surrounding
neighborhood areas.

 

NETWORK CENTER
A

“The Convener”

NETWORK CENTER
B

“The Entrepreneurial
Advocate”

NETWORK CENTER
C
“The Interface"

 

Characteristics of the Neighborhood Service Area

 

Economics Moderate- to low- Very low-income Very low-income
income

Organization/ Very high level of Generally low level of Mixed levels of

Social neighborhood neighborhood neighborhood

Infrastructure organization and organization and organization and
leadership. leadership. leadership.
Neighborhood Neighborhood
organizations highly organizations not very
coordinated. well coordinated,

indications of conflict.

Characteristics of the Network Center

Integration with Seamless integration; Not applicable - Parallel process with

Neighborhood Network Center is Summit process not yet minimal integration

Summit outgrowth of Summit well-developed

Activities

Primary Neighborhood-based Neighborhood-based Neighborhood-based

Strategies! technical assistance family support center family support center.

Functions center (with some (with some technical lnforrnation and referral,

Location/Facility

Director

programs). Collect
data, learn
systematically about
issues

Central location on
main street in
neighborhood;
storefront space;
across the street from
elementary school.

Long-term
neighborhood resident;
30+ years of
experience in
community organizing
and development.

assistance). Creatively
address
comprehensive issues
and emergent needs of
individual
neighborhood families
and significant
problems facing the
neighborhood.

Heart of neighborhood;
Freestanding
neighborhood center
combined with Head
Start classrooms

Long-time
neighborhood resident,
20+ years of
experience in human
services and
community
development.

after-school and senior
programs, emergency
needs, etc.

In part of old institutional
campus surrounded by
hurricane fence; In
relatively remote part of
large neighborhood
service area; Shared
space with a few other
nonprofit organizations.

Long-time neighborhood
resident; new
neighborhood leader but
little prior human services
or community
organizing/development
expenence.

 

92

3.3.1 Network Center A: “The Convener”

Qualitative data suggests that Network Center A plays its bridging role
primarily by convening institutional and neighborhood. partners around
neighborhood-identified and data-driven issues of concern. In its role, Network
Center A's work can be described as that of a convener, by administering
organized efforts to gather information, convert it to knowledge, and convene
neighborhood and city-level partners to craft neighborhood-based strategies and
solutions.

"For neighborhood associations and the City, I feel like we

constantly bring in people from various departments of the city to

meet with various neighborhood associations around different
issues."

"...I think that in the various sort of spheres in which we operate—
health, housing, community building activities—that we're real
instrumental in pulling various partners together in order to kind of
assess the state of the neighborhood around a particular issue and
then develop and craft strategies that will address those."

This work takes multiple forms, including both specific projects and
broader issue-oriented task forces.

Specific Projects. One of the projects Network Center A was actively
engaged in at the time of the interview was convening partners and residents
around creating a master plan for urban renewal in a BOO-household area of the
larger neighborhood area known as the "flood plain" — named because of its prior
designation as such due to low elevation and proximity to a river running through
the city. This zoning designation has effectively prevented financial investment,
resulting in significant decline over time. By educating and convening local

neighborhood organizations, township officials, and members of the city planning

93

department and state housing agency, Network Center A began a process to
move this neighborhood area forward.
"And so, for instance, we had several meetings over the last year
working with [three neighborhood organizations] to bring them face-
to-face with the emergency planning people, talking about what's
going to happen if that area floods, and what's the plan. Bringing
them together with engineers from the public service department
from the city of Lansing to talk about the engineering Changes that
have taken place in the last 15 years that mitigate flooding...and
there have been many that people don't know about. Bringing them

together with City planners to talk about what's—about the particular
problems in securing home improvement loans in the flood plain."

Task forces. In addition, the director of this Network Center described
several task forces they had actively convened to deal with issues ranging from
child safety to suicide prevention to sustainable food systems. Each of these
task forces included a broad range of between 12-20 institutional partners and
local entities In addition to neighborhood representation, all brought together to
develop neighborhood-based solutions to these identified issues.

"We are the institution that drives it. That actually identifies the

potential partners and engages them and brings them together in a

facilitated process, that we take charge of, that we take

responsibility for."

"As I think about it, that's all of what we do is that kind of brokering."

Qualitative data highlighted several defining features of Network Center
A's work that seemed to have significant bearing on how it served in a brokering
role.

Summit-Center Integration. First, Network Center A is unique among its
Network Center peers by virtue of the fact that it, for practical purposes,

comprises two "sister" organizations: the Network Center itself, and an

institutionalized "Neighborhood Summit" process, Similar to the neighborhood-

94

based planning processes often held in early stages of comprehensive
community-based initiatives (e.g., Aspen Roundtable). The Neighborhood
Summits in Lansing are facilitated community Visioning processes and planning
vehicles, using periodic town-hall forums, volunteer committees, and other input
mechanisms to identify issues of neighborhood concern and recommend
strategies for improving the community and bettering its residents. These
Summits and their lead organizations have varying relationships with the Network
Centers. For instance, unlike the other Network Centers, which began life as
physical centers for neighborhood-based programming and service delivery,
Network Center A launched in 1999 as a direct outgrowth of the ongoing
Neighborhood Summit activities launched in that part of town three years prior,
and the director of Network Center A also directs Neighborhood A's Summit
planning processes; in effect, the Network Center and the Summit in that part of
the city are inseparable.

Interestingly, despite this apparent focus on strategic processes, Director
A emphasized the importance and impact that shifting from being solely a
Summit planning process without physical facilities to becoming a hybrid entity
that could enact and house Summit-inspired programs and activities had had on
their ability to effect change in the neighborhood:

"You know what it does? It gives it gravity. It gives it substance.

You can touch the darn wall, you know? I mean, I don't know how

much longer we could have kept inviting people to meetings to

come up with ideas that would change the neighborhood without

having a place to house those ideas. The Youth Service Corps for

instance came out of the 'Strengthening Families' task force of the

[Summit]. We kept saying, yeah, that'd be great to have a chore
service—a community-service-type after-school project for kids

95

where they would actually do chores in the neighborhood, connect
with the seniors, connect with the local non-profits. We couldn't
have done that without the space! We could not have done it. You
know, so it would have been an idea...forever.- And I think people
get tired of that."

Thus, one of the key features of Network Center A's intermediary role is its
extensive use of both formal and informal approaches for gathering information
about neighborhood needs, facilitating planning around those needs, building the
capacity of neighborhood-level organizations to participate in those planning
processes, and then using the Network Center facility and programmatic
resources as a vehicle for helping the community system to deliver the
responses to those needs.

Focus on Neighborhood Capaciy. In keeping with its dual role, Network
Center A retains a function and flavor that emphasizes general neighborhood
development first and foremost, with program and service delivery serving a
Supporting role for that process:

"One function of neighborhood centers, I think, is to provide a

neighbor-friendly point of access to services. And that's why our

relationships with these—with the health and human service

institutions are so critical. But what we're also supposed to be

about—at least that's how we see it on [this side of town]—is about

genuine capacity building and community development that really

goes beyond just being a conduit for services. And this really has

to be the source for ongoing assessment, neighborhood planning,

implementation of projects, crafting of partnerships with key

institutions." — Director A

Indeed, a primary focus of this Network Center is enhancing the capacity
of the local neighborhood infrastructure and providing the information and

facilitation needed to bring neighborhood organizations and residents together

with institutional partners in creative and productive ways:

96

"[Network Center A] is a technical support center for neighborhood
organizations and community members who want to work together
to move the neighborhood forward. So they have a lot of work that
they do that allows them to collect data, Ieam systematically about
issues facing people in the neighborhood, and then collectively with
people in the neighborhood bring the people—facilitate really
between neighborhood people and service providers for change." —
Citywide Coordinator

This strategy is tuned to the particular characteristics of this Network
Center's director. Director A has specific strengths that support this focus on
"genuine capacity building and community development," including a long history
in community organizing and development work in the community dating back to
the Model Cities Programs in Lyndon Johnson's "War on Poverty" initiatives.

As such, Network Center A invests less of its staff time in working one-on-
one with individual neighborhood residents and in the kinds of programming that
are staples of Network Centers B and C:

"[Network Center A] provides technical assistance and '

development work for a neighborhood. I mean, you can go in there

and get food on Friday aftemoons...and you can go in there and tell

them that you didn't have a baby sitter. But they're not going to help

you find a baby sitter, they're going to put you in touch with [a

community referral database]. They're not a family resource
center."- CD

Extensive Outreach, Engagement, and Data-Driven Plapning. Network
Center A supports its capacity-building and neighborhood development work
through efforts to maintain ongoing access to neighborhood residents and
generate information about their specific needs. For example, on a formal level,
the Network Center has engaged in extensive door-to-door outreach involving
face-to-face conversations with up to 2300 neighborhood residents and

continually surveys the neighborhood:

97

"...What we've done on a neighborhood level is actually improve the
information that institutional decision makers have...we make it
possible for them to fulfill their mission in many ways, by making
sure that residents know about services and programs. But
also...We're constantly surveying this neighborhood...We can digest
the information that we get from our [surveys], map it, analyze it,
summarize it, and present it to [institutional partners] in order to
better enable them to shape products. . .that people will actually
participate in.” - Director A

In addition to formal data collection methods such as door-to—door and
other surveys, this Network Center also invests much of its energy in creating
and maintaining "listening structures," defined as:

"Forums, neighborhood meetings, and door-to-door knocking. Any
opportunity that has you having a conversation with a neighbor. We ,
do these in a more formal way with our quarterly forums—and

those are usually very topic-specific...and they're pretty varied—

four very different topics over the course of the year. But people in
the neighborhood and other supporters/stakeholders know that

there are at least these four opportunities every year for 30-50

pGOpIe to come together to have this conversation about some
aspect of life in the neighborhood."

The intense face-to-face work by the Network Center giVes it the
knowledge and credibility needed to recommend innovative and Viable
approaches and programs. Network Center A takes this process further by using
its outreach activities to bring residents into the Center, generating a cycle of
access to residents and their ideas:

"It's a model for being a neighborhood center without walls. For
actually going out into the neighborhood, you know—having those
porch dialogues and kitchen-table dialogues with hard-to-reach
people, and hard-to-reach neighbors, in order to link them to
services, to give them information about services. But we have a
major major emphasis here on linking information to action—not
just handing people brochures...but literally pulling them into action
around those issues by pulling them, bringing them into the center
for [programs]...it's really an extremely vigorous outreach effort to
pull people in. And because we do that, and because we invariably
collect data as we're doing that, that information carries some

98

weight when we go to the health department and say, 'here's what
we need in terms of...money and here's what it's for...”

Developed Credibility and Trufist. Director A cited the credibility and trust
developed with institutional partners since the inception of the Network Center
and Summit process as a key contributor to their ability to effectively broker now.
As a result of conducting these sorts of activities and making effective
neighborhood-based solutions available to Citywide partners, this Network Center
has evolved relationships with institutional partners that have promoted direct
investment of programmatic funds into an increasingly broad array of
neighborhood-based strategies.

"We have a four-year relationship here with the health department
but it looks nothing today like it looked four years ago...the
relationship that we have today is much more direct, there's no
money filtered through [other fiduciary organizations]. It's huge, it
may hit $80,000 this coming year. And it funds a multitude of
programs and projects and in the process addresses a whole lot of
different health issues...And it really involves the health department
putting more resources in the neighborhood to essentially support
strategies that are neighborhood-crafted and neighborhood-driven.
And that's important. The health department is not deciding to do
this, we're deciding to do this - in cooperation with them, in
partnership with them...The health department is not crafting the
strategies, we're crafting the strategies."

This evolution was described as a developmental process of increasing
trust on the part of institutional leaders who experienced the return on investment
in programs and strategies implemented through the Network Center.

"What's happened in our relationship is that they, in the first couple
of years, learned to trust our judgment and our capacity to
deliver...You start small, you get small contracts, you promise to do
certain things, and then you do those things. And then you let them
know that you've done them. And then you take it to the next step
and say, okay, trust me a little further, know that we've delivered
this in the past, let us try this. And so I think that it's a—that's the
way you build a relationship of trust. We would not be able to—l

99

mean, clearly they would not be thinking about investing $80,000
out here if they didn't think that we were having significant impact
on the neighborhood."

"I think that if [institutions] can trust that you are a reliable partner,
they are eager to invest in you, to invest in neighborhood-based
work. You know, and in fact, I think that there is greater belief on
the part of those institutional decision makers that there is a
neighborhood piece of the solution. That in addressing huge social
issues related to—oh, poverty, or ill health, or deteriorating housing
or whatever—you know neighborhoods don't have all of the
answer. But they have an absolutely key part of the solution. And
so we try to be the neighborhood representatives. Or the piece that
rep-resents the neighborhood solution. Or at least the part that is
struggling to come up with those solutions that are neighbor-
driven."

However, Director A notes that in order for the Network Center to serve as
a truly effective intermediary, the process of building trust must continually be
maintained with both institutional partners and local neighborhood residents and
organizations:

"And then again, you know, walk that fine line between cultivating
that relationship with an institutional partners and relationships with
people and organizations—including neighborhood organizations
but also churches and schools—the social infrastructure—of the
neighborhood. And work on both fronts simultaneously to deepen
the relationship and expand the relationships. And then to make
sure that you are accessible to both. That you have the input
coming from both directions, that you're an ear for the health
department, for instance, or sparrow as they talk about what their
hopes are around this partnership, as well as having an ear—being
an ear to the neighborhood, for what their hopes and visions of
what a healthy neighborhood might be.

3.3.2 Network Center B: "The Entrepreneurlal Advocate"
In contrast to the general-purpose convener role played by Network
Center A, Network Center B's bridging role can be best described in terms of

entrepreneurship and advocacy.

100

The entrepreneurial aspect of Network Center B's role is exemplified in
Network Center B's identification of and response to the specific needs of an
emerging subpopulation in the neighborhood. Because of Network Center B's
intimate connection with the surrounding neighborhood, the director there was
able to receive 'early waming' of a pending influx of parolees into the
neighborhood, recognize that a proactive approach would be required, and craft
a coordinated response:

"...We had, I believe...34 new parolees, paroled into this

neighborhood [over a two month span of time]. That's a lot of

people. Now, what are we going to do to make them productive

citizens? We need to—one, get them educated. They need to know

about our GED program. How do we do that? We get the

information out in newsletters, but are they going to read them?

How many people can read in this neighborhood? The next best

thing to do is say, 'Hello, Patty at the Parole Board? This is

[Network Center Director B] and we want to help you re-enter these

parolees as becoming productive neighborhood participants, okay?‘

Let's get them involved in the GED program if they haven't gone

through anything in the Department of [Corrections]. And let's get

them involved with Michigan Works, so they can go over there and
learn job resumes, and how to get a job..."

As a result of this and other conversations with partners ranging from
employment agencies, the Prosecuting Attorney's office, the Police Department,
schools, and faith-based organizations, Network Center B secured interest and
buy-in to a project that would aim to transition new parolees into the
neighborhood and prevent them from 'slipping through the cracks' into likely
recidivism.

The approach taken by the Network Center director in these situations
akin to that of an entrepreneur, in that she identifies issues, crafts solutions, and

then uses her relationships and influence with organizations in the neighborhood

101

and across the service system to assemble a "team" to further develop, invest in,
and execute these responses.

"You just come up with ideas and try to....l guess get everybody to
sit at the table and talk to them and whatnot. And try to get
everybody to sit and make a commitment to a specific project.
You've got to get them at the table and them keep them at it..."

The importance of Director B's role is highlighted by the fact that this issue
ordinarily would not appear on the radar of most social service agencies, nor
would ordinarily be Viewed by the Parole Board itself as a neighborhood
phenomenon. This example suggests that, indeed, the opportunities to both
apply and coordinate different programs and services to target specific problems
and populations often far exceeds the capacity of individual agencies and service
providers to identify them:

"I work with the health department each and every single day—and
here I did not know they had a specific project aimed towards
males. So when I start talking about parolees re-entering and
talking about a population of young men 21 to 35 years of age, [I
find out from them] 'Oh, well we've got this health project we're
doing and they can get free health trainings and stuff.’ And it's like,
do you know how many men I know that don't have health
insurance, that could have been using this program, and could
have been screened for prostate cancer?...But because I didn't
have them thinking on that level, you know, of working with just
men...And [now we] bring up this [male parolee issue], and health,
and how can we make this person become a healthy, productive
community member, and not return to our prison system, and
everybody's wheels start thinking, and it's like, 'Oh!‘ and here's all
this information."

However, by virtue of Network Center's B position and role in the
neighborhood as the entity continually scanning the environment and
understanding what kinds of programs and partners might fit into a solution, this

awareness can be generated when needed.

102

Although Network Center B is similar to Center A in the sense that it is,
itself, actively selecting and drawing together partners, a feature of the
"entrepreneurial" role is that this often occurs around highly specific interventions
and concrete projects.

Advocacy. Network Center B also serves as an advocate by continually
keeping the needs of Neighborhood B on the agenda of city-level agencies and
service providers:

"[Agency people] need to come back and re-educate the
community, or they have to be reminded 'Don't forget about us,
we're still here. We're always going to be here. You need to come
back.‘ But when the agency is done with their work, or the grantors
they come in, 'Oh, I've got a grant, I want to do abstinence classes
with the kids, the teenagers.‘ They do it for ten weeks and then—
they're gone, you know, we don't see 'em no more, but they need to
keep on coming back. So my job is to keep on reminding those
agencies and organizations, 'Come back here, we need you still,
don't run away.”

While Network Center B does convene some task forces locally, it often
serves its key "advocacy" role by getting a seat at existing tables rather than
creating new ones:

"Well, like in the Mayor's task force. You know, we meet bi-weekly.
And they were very interested in [Network Center B] because of the
housing task force. and they had heard through many different
other agencies or their employees about how fast we had taken the
housing task force from October of last year, to June of this past
year, and how we have progressed—like painting houses, and
getting things fixed up, and getting surveys completed, and things
like that. And how much we were into the red-tag houses and the
homeless people, and the high rent, and home ownership and
whatnot. and they came and asked me to come to the mayor‘s task
force and talk about what we've been doing. And because we went
there, they chose our neighborhood as a place that they want to
invest into. And so now, that task force is built of code compliance
officers, Parks and Recreation — every figurehead that the city
employs is there at that table, saying, 'What can we do?‘ Board of
Water and Light is saying, 'Do you need new drain systems? Do

103

you need new lighting?‘ Parks and Rec are saying, 'How can we
help you improve your park? Do you want more parks in the
community?’ You know, the head president of the landlord
association sits on that board. He's saying, 'Okay, you're having
problems with these landlords? Let me know these people's names.
We will get to it.’ So it's bridged this big huge gap between the
neighborhood and me to have to go to each one of those
individuals - they're all sitting there in that room. And all I've gotta
say is what I need, and they're there answering..."

"I've been taught a long time ago by someone that if you're not
sitting at the table, you're going to miss out. You've got to always
stay at that table, so that's been helpful. And I guess I am bridging
that gap between the City and the neighborhoods then, because if I
didn't go and attend those meetings, they would be, 'They're not
interested, we'll go on to a different neighborhood I guess.”

Qualitative data suggested that key factors in serving this particular type of
bridging role included Network Center B's role as a hub of neighborhood life and
the trust and credibility developed with organizations and residents alike.

Network Center B as thriving hub. As a full-service family resource center
and arguably the focal point of neighborhood life in Neighborhood B, Director B
and the center staff are particularly close to the day-to-day happenings and
emergent needs in the surrounding neighborhood.

Network Center B is located in a small, freestanding community center
next to a local park, centrally located in an area with few other public spaces—in
essence, it is the center of the neighborhood it serves. Between the extensive
roster of youth and family programming offered at Network Center B and its
central role as a public gathering space and resource in its local neighborhood,
this center draws several hundred residents in every week:

"...and they all bring with them problems, issues, concerns, realities

of their life. And the director there has the ability or has the

responsibility really to assess what those needs are and then to put
peOple together to help solve a problem. So it might be that they

104

come in initially with a health problem, but the health problem
probably relates to lack of insurance, so she's able to connect them
with [low-income insurance programs], but then that's connected to
lack of resources, so she brings in FIA and the health department
and F IA will sit down with her and help her try to solve that problem.
It might have to do with child abuse and neglect, so FIA again is
involved. But it might be that the family simply doesn't have
adequate child care, and so she'll bring in the Child abuse and
prevention services, so that the family can have some of their
needs met within the neighborhood, but they wouldn't have
necessarily have known about [those resources] if they hadn't come
into the center. They can get parenting education at the center, if
the kids are old enough they could enroll in Head Start. If the kids
aren't old enough for Head Start there's a drop off day care center
that's free. They can get their GED there so—yOU know, she can
connect people with resources and have those agencies
understand that they're all working together to help a particular
family.” — Citywide Coordinator

In this sense and as stated by Director B, this Network Center is working
towards a vision of a "wraparound" system that can reach all local residents as
needed. In this vein, this Network Center strives to function primarily as a hybrid
between a neighborhood center and a comprehensive outlet for social service
delivery, carrying a wide range of comprehensive programs for residents of all
ages, information and referral services coupled with assistance in accessing and
negotiating community resources, and direct assistance/supports in the form of
basic needs closets and food distribution programs.

From this vantage point, this Network Center has easy and extensive
access to neighborhood residents and knowledge about the current and
emerging issues they face. This puts Network Center B in a position to both
advocate for particular resources and services to bring to residents through the
Center as well as craft comprehensive strategies that it can then seek investment

in from organizational partners.

105

Trust and credibility. Director B, like Director A, put a heavy emphasis on
the importance of the trust and credibility developed and maintained with both
organizational partners and local residents. In Network Center B's case, this
trust and credibility was described as growing out of both the Director's
rootedness in and knowledge of the neighborhood as well as a developmental
process of building a track record of good judgment in assessing needs and
picking winning strategies.

Director B has years of experience from both working within and being a
client Of the local social service system. Observations of and discussions with
this director over the years suggest that experience lends a certain savvy to her
work with these organizations, as well as a special understanding of how specific
and often fragmented services can intersect in the lives of whole persons. She
described it this way:

"...You've got to have heart. Your heart's gotta be there, I guess I

want to say. And I think the reason why [we're successful is] I and

the other people who work on my staff—we've experienced some of

the things that these people [low-income neighborhood residents]

have experienced, so we understand where they're coming from,

and so we can take it to the city level or the other higher levels, and

say, 'Look. They need it here, and they want it here, but they don't
know how to change.’ It just works, somehow or another."

Network Center B, like A, also describes the Network Center's
relationships with partners as evolving considerably over time. One of the noted
features of this change is the level of effort needed to convince partners of the
level of local need or the viability of Network Center-crafted programs and
strategies. As described by the Director:

"Basically [now I can] walk in there, and say, 'Look, I've got a GED
Class that I need money for to pay a teacher, I've got 66 people

106

here.‘ And they're like, 'Okay, how much do you need?‘ That's how
it is now, but before they wouldn't do that...They had to Ieam to
trust me, I guess—[trust] my judgment."

"... I wouldn't have that kind of response if I wasn't sitting in the
position I'm sitting in now. Where, if [the neighborhood association
president] would have called, she would have called a hundred
people and probably still be sitting there."

In summary, Network Center B has, over time, developed a reputation as
a reliable detector of neighborhood needs and an effective administrator of
programs that will be a good investment of service system resources. As a
result, this Network Center has evolved into a crucial advocate and intermediary
for residents in that part of town.

3.3.3 Network Center C: "The Interface"

As quantitative data suggests Network Center C is positioned to serve
extensively as an information bridge between organizations at both neighborhood
and city levels. The qualitative data supports these findings by suggesting that
Network Center C functions primarily as a local entity for direct programs and
"information and referral" services. As such, in a brokering capacity Network
Center C's role could be described as an interface among and between
neighborhood and city levels.

Operationally, Network Center C is akin to Network Center B in that it
serves as a family support center to funnel programs and services into the local
neighborhood to meet the needs of local residents. In this capacity, Network
Center B is most active in its after-school programming, information and referral
services, and basic needs provision:

"We do that by helping people with the material goods, non-

perishable items like food and Clothing and all things of that nature.

107

We also have programs based on the needs of this community
here—that we've had forms and surveys and these people in the
community have came in and made their problems known and
things that they'd like to see happening in the-community...and
we're working on those things and we have programs to meet those
needs. We develop [programs] as the needs arise. Say, for
instance, if there's a number of people in the community that want
to see a GED program here...well, the funding has been cut and
we don't have a GED program here. But we do refer them to places
where they can attend a GED program. So we do a lot of referrals.
People come in here; they may need help with their rent. We refer
them to the FIA department, the Salvation Army, or different other
agencies that help with those things."

"If a person comes in here and they may not have any food. Well,
we have a basic needs closet here, we will share with those people
what food we have until they can get some funds or whatever. But
we may not have what they need and we may have to refer them to
the [Church] who has a food bank...or wherever we need to refer
them to help them with whatever situation or problem they have is
what we do." — Director C

However, this Network Center does not appear to engage in either the
types of neighborhood-based planning and facilitation that drive Network Center
A's work or the level of advocacy and programmatic entrepreneurship that
characterize Network Center B. Instead, as an interface, Network Center C
interacts with and shares information with both City-level and neighborhood-level
organizations, but does not appear to take a leadership role in developing the

capacities of either of these levels of the system:

"There have been neighborhood associations that have meetings
here—that are allowed to have meetings here at [Network Center
C]—and I've sat in on those meetings, and they've had problems,
neighborhood association problems, organizational problems in
their community with drugs. And prostitution. And I have helped
them to make the necessary connections with the police
department to step up the patrol or the surveillance or whatever to
rid the community of those negative aspects."

108

"I'm Chair of [a committee] for the Mayor's Office and naturally, that
information that I bring back from those committees I share with the
people in the community. Because that's what that's for, we want
the people in the community to understand and know the rights that
they have, and the resources that are available to them through
these other organizations that they don't make contact with."

As these quotes and other examples illustrate, Director C appears to
enact its bridging role largely as an informational connector or "pass-through"
between residents and service system organizations. However, in confirmation
of the quantitative analyses presented above, it does not appear that this
Network Center has appreciably leveraged its inforrnation-sharing ties into the
deeper and more powerful joint planning and programming relationships and
resulting synergistic activities common to the other two Network Centers.

There appear to be several factors working against this Network Center's
ability to take on a significant role in this regard. These are: the capacity and
experience of the Network Center Director, the location and lack of neighbor-
friendliness of the facility occupied by the Center, and the disconnect between
the Network Center's activities and other neighborhood-based planning
processes in that part of the city.

Director Capacig/Experience. The director of Network Center C came to
neighborhood work relatively recently, after retiring from a prior career and
realizing the extent of social ills and deterioration in his own neighborhood. This
inspired Director C to take up neighborhood activism and become a respected
leader in the local neighborhood. However, this director still understandably

lacks the experience and capacity of either Director A or Director B.

109

This affects Network C's capacity to serve in a bridging role in two ways.
First, Network Center C has a greater tendency to implement one-dimensional
programming that lacks the creative packaging, comprehensive focus, or
facilitation of ongoing resident engagement with the Center that are featured in
the activities of Network Centers A and B:

"...The director of [Network Center C] doesn't have that level of
long- term human service agency involvement. So the approach to
solving problems, to working with people is different. And that
results in programs being implemented there which are less
imaginative. Okay, there the kids get out of school at 3 o'clock and
somebody oughta be there to help them with their homework. So
volunteers are brought in to help them with their homework after
school. And resources are identified in the community that can
provide some enrichment activities for those kids after they get their
homework done. So there's Boy Scouts, Girl Scouts, and at times
4H. There is a senior dining site there, so there's a free lunch
program twice a week there. The seniors come in and eat, and they
leave. Where if there was a different level of sophistica-
tion...training, education, experience, those seniors would be seen
as a resource. A dialogue would be occurring on a regular basis
about what they're dealing with, how the Network Center can help
them, what services they see that are needed that aren't there. And
that would happen with the families of the children as well. But
because of that different level of sophistication, it's Viewed
differently—'Okay, kids are here, we need tutors, let's go get
tutors.” - Citywide Coordinator

Second, Director C's comparative lack of experience with bringing people
and organizations together at either the neighborhood or city level means that, in
contrast to the other two network centers, Network Center C is rarely at the
center of joint planning efforts:

"[Facilitating planning] is not a skill [the director of the Network

Center C hasl—He is not a trained facilitator. He...doesn't know

about planning. He sees these kids, and they must need

something to do after school, and he'll call [the university] and get

some students over there to tutor them. But how to bring people in,

listen to what their concerns are, and jointly bring partners in to
create? For instance, employment is a huge problem over there.

110

People don't have adequate income. They don‘t have access to
employment services...So people need help with resume
development, interviewing skills, clothes to wear to work, how to
position yourself, how to find a job, a job club. We've had
community meetings over there for the last three years; this comes
up as a priority issue for the Network Center to deal with. YOU
know, I'm going to stand back from my role, and say, that's not my
job. My job is not to listen to the community voice [in the
Neighborhood C Area], and hear that they need help with
employment and go over there and do that—that's his job. But he
doesn't know how to do that."

In sum, data suggests that Director C's relative lack of experience is
related to a more static view of programs and services and a greater focus on
disseminating information about what programs/serviceslresources already exist
and a less proactive approach with regard to identifying what new possibilities
could and should be implemented.

Physical Center Facilig. In contrast to the centrally located and neighbor-
friendly facilities occupied by Network Centers A and B, the building occupied by
Network Center C is poorly positioned in the neighborhood. This center is
located in a building on an old institutional piece of land and is remote from much
of the neighborhood service area. As suggested by one informant, this creates
not only natural physical barriers to accessing the Center's programs and
services, but psychological ones as well:

"How much of a barrier is a cyclone fence to the [Network Center

C], when there's a guard house at the driveway...and you're new to

the neighborhood but you're in trouble, and you don't know if you
need a password?” — Citywide Coordinator

"I mean, this [Network Center B] is 390 kids in the summer....and
this [Network Center C] is probably 15 kids in the summer.
Network Center B is anybody come, anybody...Where, [Network
Center C] is very limited. The people who know about it, the people
who trust it, the people who've been going there..."

111

Thus, while estimates suggest that up to 100 area residents walk through
the doors of Network Center B on a daily basis, Network Center C suffers from
considerably lower awareness and participation amOng residents in its
surrounding neighborhoods.

Lfik of Summit Integration. Finally, whereas Network Center A operates
in concert with both its sister "Neighborhood Summit" organization and an
umbrella neighborhood organization, Network C operates in a local environment
that has similar comprehensively-focused organizations that do not cooperate
and in fact are sometimes in competition for resources. For example, the
organization that operates a parallel Neighborhood Summit and comprehensive
development function in Neighborhood C reportedly has little direct
communication and coordination with that Network Center. This rift was cited as
an additional disadvantage for that Center, as it caused corresponding divides in
resident participation and institutional funding between these neighborhood-
based organizations.

3.4 Summary

In summary, the formal brokerage score data indicates that, overall, the
Neighborhood Network Centers do occupy significant brokerage roles in the
community network; and they consistently show the strongest brokerage
potential as gatekeeperl representatives between organizations in their home
neighborhoods and citywide organizations, with two of the three network centers
also serving as coordinators among organizations within their own

neighborhoods. This suggests that while the Network Centers may serve,

112

through their brokerage roles, as an important vehicle for facilitating collaboration
in the local community service system, this role is not generic in focus. That is,
rather than serving as all-purpose intermediary agents in the system (as would
be suggested by occupying significant brokerage roles across all three types
measured in this study), the Network Centers bridge more specifically around

neighborhood-based issues.

113

CHAPTER 4
DISCUSSION

4.1 Overview

As this thesis has described, the present study was unable to fulfill its
original intent of capturing the influence of Network Centers in shifting the
structure of relationships in the larger community system. However, in the
process of exploring the network data collected, an infrequently used analytical
approach was identified that yielded a direct assessment of the Network Centers'
brokerage roles—that is, the capacity of the Network Centers to actively bridge or
broker among neighborhood-level organizations, between neighborhood- and
city-level organizations, and among city-level organizations. The patterns
revealed by these brokerage scores, viewed within the context of the individual
Network Centers' resources and styles of operation and the social infrastructure
of their surrounding neighborhoods, suggest that this measure may have
significant potential in application to understanding the roles and effectiveness of
community-based intermediaries.

In the following sections, l illustrate this potential first by presenting an
interpretation and integration of these overall patterns as they relate to the local
community system. Second, I discuss the connections between apparent
features of Network Center functioning as it relates to brokerage capacity and
related issues presented in the literature on collaboration, community based
initiatives, and service integration. Third, ltum my attention to both the
limitations inherent in this study and related cautions regarding application of

social network methodology in general, and brokerage analyses in particular, to

114

similar research in the future. Finally, and in light of all of the above, I discuss
directions for future research to build on this exploratory work to continue the
task of developing a constructive empirical and theOretical base for
understanding neighborhood-based intermediary organizations in their
complexity.
4.2 Interpretation of brokerage patterns

By design, Network Centers could be expected to have a broad array of
linkages at the neighborhood and city levels. Network Centers have a broad,
place-based mission; they are not tied to any one approach or programmatic
strategy. This comprehensive approach means that they are in a position that
allows them to see people, place, and the problems and potentials of these
comprehensively and holistically, and to naturally forge relationships among a
wide range of stakeholders. Given this, it is useful to explore in greater depth the
fact that the brokerage score patterns did not reflect a uniform assumption of
what capacity these neighborhood-based intermediaries might have to broker
relationships, but instead varied widely (1) across the three Network Centers
examined; and (2) between information sharing and joint planning and
programming types of ties.

4.2.1 Differences across network centers

One of the key features of this data is the difference in patterns of
brokerage capacity across the three network centers. While all three Network
Centers exhibited significant brokerage capacity to some degree, Network Center
A showed the strongest capacity for brokering between organizations at the

neighborhood level as well as between neighborhood-level organizations and the

115

broader system. Network Center B was also strong in this regard, but followed
behind Network Center A in measured brokering capacity. Network Center C
was inconsistent with the patterns of the other two, Showing very high brokerage
capacity for information sharing across all types of organizations, but much lower
capacity for joint planning and programming.

Importantly, examining the brokerage scores in light of the Network
Centers' unique neighborhood and operational contexts suggests elements that
contribute to the level and sophistication at which Network Centers can bridge ,
among organizations in the community system. One theme that emerges across
the case studies is groundedness of the Network Centers in their neighborhood
contexts. That is, the two Network Centers (A and B) that exhibited the greatest
physical, programmatic, and experiential integration with their neighborhoods
showed the greatest capacity to bridge in action-oriented ways (i.e., joint
planning and programming) both within their own neighborhoods and between
their neighborhoods and the larger system. These aspects are discussed further
below.

Physical integration. This factor can be summed up simply as: location
matters. The location and accessibility of a Network Center facility may impact
how readily it can be adopted by residents as a hub of neighborhood activity.
Network Centers A and B are both well positioned in this regard. As described
in Section 3.3 above, Network Center A is located in a storefront space on one of
the main thoroughfares in the neighborhood, making convenient, easy to find,

and comfortable to enter. Network Center B is a well-known space and the

116

primary activity hub for its neighborhood. In contrast, Network Center C suffers
from a more isolated location in its neighborhood area that is inconvenient for
many residents to access; in addition, the building this Network Center occupies
has a less welcoming exterior and interior appearance. As such, the experiences
of Network Center staff suggest that the facility housing the Network Center
affects its ability to house programs and services that will yield large-scale
participation on the part of neighborhood residents. This is particularly important,
as access to residents may directly facilitate the ability of Network Centers to
foster joint planning and programming relationships. It is reasonable to infer that
as hubs for high levels of many kinds of neighborhood activity, Network Centers
A and B are able to feed a cycle of involvement on the part of residents and
organizations alike: the high level of access to residents that both of these
Network Centers can boast means that they can offer city-level organizations
better on-the—ground information and higher participation rates in sponsored
programs, thus helping agencies and institutions in accomplishing their own
missions. This may be one factor in organizational interest in these Network
Centers, as they can serve as vehicles for outreach and for minimizing the cost
of service delivery. In turn, these Network Centers have been more successful in
securing the programming and planning relationships with partners that bring
additional resources into their neighborhoods, enabling additional programs and
services that further increase the value and Visibility of these Network Centers to

residents.

117

Qperientiajinteggtion. Another factor that likely has significant impact on
the relative capacity of the Network Centers to broker is the overall experience of
the Network Center directors in the kinds of activities related to a brokering role,
as well as their seeming legitimacy in that role. Both Directors A and B have
extensive experience in community work in their neighborhoods, and have had
the opportunity over the years to develop considerable skill and savvy in working
with residents and formulating local strategies. In contrast, Director C came to
his position as an emerging neighborhood leader who had developed positive
working relationships with some city-level organizations in the course of his early
forays into neighborhood activism, but did not yet have the experience or skill
needed to guide planning processes or develop innovative programmatic
strategies. These differences may also serve to explain much of the difference
between the Network Centers in joint planning and programming brokerage
scores. Further, the experience of the directors may be coupled with their
legitimacy (and by extension, the legitimacy of their corresponding Network
Centers as organizations) as perceived by both local neighborhood and City-wide
organizations.

Legitimacy within the social context has been described as a key factor in
successfully serving an intermediary role (e.g., Burt, 2000a). This legitimacy can
come through action (Chaskin & Peters, 2000), as demonstrated in a track record
of accomplishment. In addition, the perceived legitimacy of Network Center C
may not be entirely due to that director's relative inexperience, but may also be

impacted by the more divisive neighborhood politics in the area of the city served

118

by that Network Center. Chaskin (2001) suggests that "issues of trust and
legitimacy have much to do with the state of the existing organizational
environment. In organizationally dense contexts, or where there is a history of
contentious relations among organizations, the work of a broker organization is
likely to be made more difficult, and acceptance of its role may be problematic."
(p 156). This description is aptly applied to Neighborhood C's area, which has
many existing neighborhood organizations that either do not cooperate or are
divided into competing factions. As the brokerage data shows, Network Center C
has been able to serve a brokering role with schools much more so than with
neighborhood organizations in its region. It is possible that to some of these "old
guard" neighborhood leaders, Network Center C lacks legitimacy due to the
relative inexperience of its director; alternatively, legitimacy may be denied by
some of these organizations due to perceived competition for resources or for the
Clout of representing the neighborhood to outside interests.

Unfortunately, it is possible that some aspects of this situation may
contribute to a vicious cycle. In reviewing the factors that motivate organizations
to enter into interorganizational relationships, Oliver (1990) suggests that:

"Even in the absence of a government mandate, government

representatives may encourage organizations to work together in

order to consolidate the implementation of public policies or to

rationalize spheres of activity through more integrative and

cooperative social service delivery. Organizations that project the

appearance of rationalized activity and cooperation through joint

program activity often are able to mobilize more funding and

authority to provide services in a specific domain than are
autonomous organizations." (p. 256)

Thus, a certain amount of legitimacy in the eyes of local neighborhood-

level organizations might be gained through the funding and institutional support

119

that could stem from a demonstration of active capacity in joint planning and
integrating services. However, in order to successfully engage neighborhood-
level organizations in that process, Network Center C would have to successfully
distinguish itself as a legitimate convener for such processes apart from other
organizations engaged in those activities in the neighborhood.

Prpgrammatic integration. A third factor that may have an impact on the
brokering capacity of the Network Centers is their programmatic integration with
other related efforts in the neighborhood. As described by Chaskin (2003), "The
organization chosen or created [as a neighborhood-based intermediary] operates
in a context that is often already well populated with a range of associations,
organizations, and crafted coalitions that would also claim—in particular cases or
around particular issues—to speak for and act on behalf of the neighborhood and
its members" (p. 163).

b This situation is true of both Network Centers A and C. However, whereas
the neighborhood organizations in Neighborhood A are highly organized through
an umbrella neighborhood organization that Closely coordinates its work with
Network Center A and its affiliated Neighborhood Summit process (all largely
using the common plan developed through that process to guide their overall
work), Neighborhood C operates in an area where multiple active neighborhood
associations and comprehensive or multiple-issue neighborhood-oriented
nonprofits coexist, but do not as a whole coordinate their activities around a
shared agenda for the larger neighborhood. In addition, while the neighborhood-

level networks for both Network Centers A and C have comparable centralization

120

scores (e.g., 0.61 and 0.63 respectively for joint planning and programming),
Network Center A is one of the most central organizations in that network
(normalized degree centrality for joint planning/programming = 100, rank = 1),
whereas Network Center C is relatively less central to the planning and
programming network in its area (normalized degree centrality for joint
planning/programming = 61.54, rank = 4).

In contrast, Network Center B is the only organization in its neighborhood
that works comprehensively on behalf of the neighborhood. In addition, the
network in Neighborhood B is rather decentralized in comparison to
Neighborhoods A and B (e.g., centralization score of 0.33 for joint planning and
programming). While Network Center B is the most central organization in its
local neighborhood network (normalized degree centrality for joint planning/
programming = 77.78, rank = 1), it is, in essence, serving to coordinate in a
neighborhood that is considerably less structured than Neighborhood A or C.

Thus, Network Center A's strong capacity as a broker may be related to is
strong positioning in a network that is centralized and thus streamlined to
promote ease of intermediation between the neighborhood and outside entities.
Network Center B's scores indicated somewhat less capacity for brokering; while
Network Center B is a highly central player in its local network, the more diffuse
nature of this network may reduce the efficiency with which this Network Center
can broker ties. Finally, Network Center C's task of brokering may be
significantly more challenging given the fact that there are multiple organizations

engaged in cross-cutting partnership activities in its area; this may contribute to

121

diffusion of partnership interest and additional difficulty in attaining legitimacy as
an organization working on behalf of neighborhood residents.

In addition to these factors contributing to the relative integration and
groundedness of the Network Centers in their respective neighborhoods, there
are stylistic differences in the way these Network Centers operate in their
neighborhoods that may also contribute to differences in brokering capacity or
activity. Specifically, Network Center A and Network Center B present two
related but different models of neighborhood-based intermediary work. Network
Center A's work is very strongly focused on convening task forces and other
groups of organizational partners around researched and resident-defined
neighborhood issues. In contrast, much of Network Center B's work was
described in a way that seemed more "entrepreneurial" in nature—crafting
neighborhood-based responses to more focused needs and pulling in the right
combinations of partners to implement these particular strategies. This
combined with the heavier focus at Network Center B on direct program and
service delivery and individualized assistance to residents suggests that the
partnership and brokering activities engaged in by Network Center B may be
both more time-consuming (by Virtue of being more dispersed over many specific
responses and projects) and more intensively centered on Director B and the
amount of time and energy this directOr has available to craft these specific
project collaborations. The structure of Network Center A's work, in contrast,

may simply bring larger numbers of partners together.

122

4.2.2 Difference between Information Sharing and Joint Planning and
Programming

It is not surprising that overall patterns indicate lower levels of direct
relationships for joint planning and programming. As described by Fried et al
(1998), information sharing ties are the easiest to maintain and require lower
commitment level than relationships such as joint planning and programming that
Involve what Laumann et al (1978) call "interpenetration of organizational
boundaries". Indeed, the transition between more informal, inforrnation-oriented
networks to the more formalized linkages embodied in joint activity networks has
been described as an evolutionary process (Alter & Hage, 1993).

One of the clear patterns shown in the brokerage data was the large
difference in Network Center C's brokerage capacity between information sharing
and joint planning and programming roles. In addition, this Network Center
showed brokerage capacity for information sharing that was higher than that of
the more established (and, arguably, more effective in their own ways) Network
Centers A and B.

As noted above, Network Center C's director was described as having
less capacity (and perhaps legitimacy) to facilitate joint planning relationships
and craft programmatic strategies than the other Network Center directors.
Therefore, one possible interpretation for this Network Center's greatly reduced
brokerage scores for joint planning and programming is simply that this Network
Center lacked the capacity to fulfill that role. Perhaps in compensation for this,
this Network Center has sought out a prominent role as a broker of information in

its neighborhood, as evidenced by its high brokerage scores for information

123

sharing as well as its high degree centrality in its neighborhood’s information
sharing network (normalized degree centrality for information sharing = 100, rank
= 1). This Network Center may be able to foster su-ch a wide range of
information sharing relationships precisely because they require less effort to
maintain. In contrast, Network Centers A and B appear to be much more
focused in their relationships, and their brokerage roles for information sharing
are similar to their roles for joint planning. This suggests that their information
sharing activities are less broad in nature, but rather concentrated with those
organizations with which they engage in the more intensive work of planning and
creating programming on behalf of the neighborhood. It is therefore plausible
that as (and if) Network Center C gains legitimacy, confidence, and experience
and matures into a more active joint planning role, the wide range of its
information sharing activities might shrink to more Closely reflect the patterns
shown by Network Centers A and B.

Finally, it is important to note that the relative lack of brokering capacity for
joint planning and programming evidenced in scores for Network Center C does
not necessarily imply that this Network Center isn't yet a highly valuable asset in
its neighborhood. This Network Center provides services that are important to
residents in its surrounding neighborhoods. While it is true that this Network
Center serves fewer residents than Network Center 8 does, this contribution to
the lives of residents should not be minimized. For example, Network Center C
has been able to report improved outcomes for participants in its programs, such

as better grades for youth attending after-school tutoring activities. In addition,

124

since the inception of this Network Center, the local police department has
reported that crime rates have decreased in the neighborhood areas surrounding
it. Clearly, the presence of any neighborhood resoUrce such as this can have
significant direct benefits for residents in these low income and under-resourced
neighborhoods. However, the more dynamic range of programs and capacity
building activities engaged in by the other Network Centers (and perhaps
particularly Network Center B, which presents a similar "family resource center"
model to Network Center C) demonstrates some of the potential missed by not
incorporating more joint planning and programming work.

Unfortunately, given the complex mix of factors that may contribute to
Network Center C's reduced capacity in comparison to the other Network
Centers, it is difficult to know what path this Network Center might take to
increase its capacity to broker. Some possible steps might be:

° Foster more activity through a more accessible/friendly location and/or
promoting the Network Center's activities more effectively in the larger
neighborhood area;

0 Work to increase the legitimacy of the Network Center by forging
necessary relationships across neighborhood associations in the area
(albeit recognizing that local neighborhood politics may make this step
extremely difficult) and providing training to the Network Center
director to enhance the director's skill and confidence; and/or

' Work to increase the programmatic integration of the Network Center

with other efforts in the area, by forging specific alliances with the

125

Neighborhood Summit and other major service-providing organizations
operating at the neighborhood level.
4.3 Implications
Extending the analyses above, the findings of this study offer potential
implications and "food for thought" for scholars and practitioners engaged in work
with community development initiatives that strive to serve capacity-building roles
as relational intermediaries or brokers.

4.3.1 Importance of Trust/Credibility
First, the directors of both Network Centers A and B described the trust,

credibility, and respect they had developed with both City-level and
neighborhood-level organizations as being a key factor in their success as
bridging agents. This is consistent with ideas expressed in both community
development literature (e.g., Ferguson & Dickens, 1999) and social capital theory
(e.g.,Portes, 1998) that emphasize the essential role of trust in the development
of alliances and partnerships. Ferguson and Dickens (1999) describe four key
components of trust that affect willingness of organizations to participate in
alliances, namely: trust in motives, trust in competency, trust in dependability,
and trust in collegiality. These forms of trust are particularly evident in the case
studies of the Network Centers presented in this study.

The need for developing trust in motives is expressed with regard to the
relationships between Network Centers A and B and their respective
neighborhood constituencies. Network Center A described this trust building in
terms of remaining accessible and providing a listening ear to constituents.

Network Center B described this in terms of "heart," dedication, and remaining

126

clearly a part of the neighborhood. In contrast, trust in competency and
dependability came more into play with citywide agencies and institutions. With
organizational partners, directors described their inCrementalIy developed,
proven track record—whether for delivering desired outcomes, developing
innovative and effective neighborhood-based programming, or accurately
spotting neighborhood trends in the making—as the basis of their "stock" of trust
with major organizational players.

Thus, the findings of this study provide additional support for the
importance of trust as an element in social capital and provide an additional
reminder to practitioners and initiative designers that trust is a component in
capacity-building initiatives that cannot be ignored and—importantly—cannot be
rushed.

4.3.2 Process-Product Integration

Another element of the Network Center model that may be of interest to
scholars and practitioners is the way that community building "process" and
related delivery of "product" are combined in the activities of Network Center A
and B (and most particularly A).

One of the fundamental issues cited in managing comprehensive
initiatives for community change is the tension between process and product:

CCls are by nature long-terrn enterprises. Even if they eventually
succeed, it will take years, if not decades, to achieve the magnitude
of improvement to which these initiatives aspire. Moreover, creating
the conditions for sustaining such Change over time requires
investment in the capacity of individuals and institutions in the
neighborhood, and the returns on this kind of investment are often
not seen for years. But at the same time, CCl stakeholders feel a
sense of urgency about the need to change conditions for
neighborhood residents and want to see progress as soon as

127

possible. In addition, as many of the participants across the
discussion groups made clear, CCIs must have something to show
at a relatively early stage in order to gain and sustain support both
from outside and inside the neighborhood. Funders of an initiative,
for example, generally want early signs that their support is paying
off and should be continued. And, neighborhood residents also
need evidence that their investment-in time, energy, institutional
resources, or political support-is worthwhile. (Aspen Roundtable,

1 998)

However, neighborhood-based intermediary models such as the
Neighborhood Network Centers demonstrate one potential avenue for balancing
this tension. Because the partnerships brokered or facilitated by the Network
Centers typically have some component that is tied to short-term delivery of
programs and services through the Network Centers themselves, these
processes continually have "products" inherently built into them at some level. In
addition, Network Center A (which is perhaps the most CCl-like of the Network
Centers due to its pairing with the extensive planning activity of the Summit
process and the neighborhood input mechanisms it uses) demonstrates how
many of the products—that is, new services to benefit residents—can be
creatively leveraged to further the process of capacity building in the
neighborhood. Director A brought this up in our interview:

"...I think that people tend to separate too much process and

product, process and product. And I’ve sat through meetings with

Summit folks who'll talk about, ‘Well, let's focus on the engagement

and planning process,’ and ‘Well, that's kind of product-y, those

health navigators are product-y”. Never mind that it was the health

navigators that were continually engaging neighbors and planning

new initiatives and doing assessment, planning, implementation,

assessment, planning, implementation. It was the housing

ambassadors—a product—that was really the vehicle for ongoing
assessment...you know, so I think it’s a false dichotomy—this
process-product is just...it drives me a little bit nuts, you know?

And maybe it’s because we are kind of this conjoined deal here,
because we are both Summit and Center, you know? One with its

128

emphasis on process and one with its emphasis on service delivery
and product—but honest to God, I can’t imagine not having both of
those! I mean, if you were only a program organization, I mean if
you were only a kind of a hub for agency people to send a social
worker or a nurse, or some kind of clinical practitioner or whatever
to do their service delivery - so it would be a little bit more
convenient for people to get those services. Big deal. I mean, is
that really change? Is that really dynamic and does it really hold the
promise of changing peoples lives and the quality of life in their
neighborhood?"

Many scholars of multi-stakeholder processes have sought to define
various levels of collaborative relationships to delineate the varying levels of
intensity that this work can attain (e.g., see Bailey & Koney, 1996; Himmelman,
1996). For example, Mattessich and Monsey (1992) make a distinction between
collaboration as a formal process, coordination as an informal process, and
cooperation as project-specific. Given these definitions, it is clear from
interviews, observations, and documentation that much of the Network Centers'
roles are executed at the less-stringent coordination and cooperation levels of
multistakeholder work. An unfortunate side-effect of such "levels"-based
typologies is that they often seem to subtly portray "true" collaboration (typically
defined as a higher level of commitment including exchange of information and
resources and mutual investment in building partners' capacity) as difficult-to-
achieve but nonetheless ideal. However, I believe that rather than minimizing the
importance of the Network Centers' roles, this raises important and interesting
questions for those considering how to foster more effective interorganizational
linkages in communities. Specifically, examples such as Network Centers
present an alternative, loosely-structured model from which to mobilize

networked action that may be appropriate for many concerns. For example,

129

while mobilization of an interorganizational network can occur when individual
organizations spontaneously realize the need for joint action, Laumann et al
(1978) suggest that this is likely to be atypical; instead, they suggest that
"preexisting network ties may be used by those actors most acutely affected by a
particular issue to generate a broader consensus" (p. 472). This describes the
"ad hoc" approach to collaboration embodied in the Network Center approach:
By facilitating connections through multiple, smaller-scale and typically time-
limited venues (e.g., task forces, specific collaborative programming initiatives,
etc.), it appears that Network Centers may be able to maintain a broad base of
linkages that can be leveraged in new combinations as needed. This is
illustrated in the example discussed above where Director B refrarned a
particular issue—re-entry of parolees into the neighborhood—into an actionable
focal point for collaborative action by multiple neighborhood and city-level
stakeholders.

In a sense both this process-product dynamic and the looser ad-hoc
nature of collaborative partnerships through the Network Centers may be
indicative of a broader factor in their adoption in the community: lowering what
one might term the "cost of collaborative overhead". The collaboration literature
offers many studies showing that one key factor influencing member satisfaction
with and commitment to partnership efforts is the perceived ratio of benefits to
coSts of participation (e.g., Bitter, 1977; Butterfoss, Goodman, & Wandersman,
1993; Chinman, Anderson, lmm, Wandersman, & et al., 1996; Kegler, 1998;

McLeroy, Kegler, Steckler, Burdine, & Wisotzky, 1994)

130

Thus, it is important to acknowledge the important potential role of specific
broker organizations with designated, comprehensive, neighborhood-serving
roles in helping to maximize the benefits and minimize the costs for city-level
organizations in partnering with neighborhood-level organizations. To the extent
that such organizations can engage institutional partners in ways that generate
tangible payoff for their efforts and make maximal use of their invested time,
partner organizations may grow more and more comfortable in collaborating in
comprehensive processes. For the Network Centers, this takes the form of (1)
bringing relevant partners together around specific identified issues or goals
rather than engaging all partners in a large-scale process; and (2) providing
additional functions that make the work of other organizations and institutions
easier, such as serving as a venue for programs and services that will reach
residents. providing outreach to neighbors about resources and programs
available to them, and serving as a source for neighborhood-based analysis and
a local "radar" for community needs.

4.3.3 Sustainability

Another core issue that faces many community-based initiatives—
particularly the sizeable percentage of these which are seeded by external
foundation money—is sustainability of the initiative once the initial funds have
ended. Chaskin, in describing the transition or sustainability phase of the Ford
Foundation's Neighborhood and Family Initiative, says:

"To the extent that collaboratives were successfully able to play [a

broker] role, they did so because the initiative provided them with

an ongoing, dedicated stream of funding from the Ford Foundation,

access to the skills and expertise of cross-site technical assistance
providers, and varying levels of access to broader resources

131

through the community foundations. Because of their access to

funding, they have also been able to Channel funds to other

organizations, or to develop programs that they could then transfer

to others for implementation. And because of their affiliation with a

national initiative of the Ford Foundation and a local community

foundation, the collaboratives had a relatively high profile that

supported the interest and involvement, over time, of a range of

participants. Unfortunately, because of the ongoing centrality of

external stimulus and support, sustaining this role may prove

difficult...the very conditions that have allowed the collaboratives to

develop their role as brokering organizations are now

disappearing." p. 158

Indeed, it is unknown just how sustainable interorganizational
relationships based on formal coalition structures are in the absence of continued
funding (Crisp, Swerissen, & Duckett, 2000). However, one might hope, given
their dual function as intermediary and vehicle for neighborhood-based service
delivery, that the Network Centers themselves will be considered a permanent
investment in the community infrastructure rather than constituting a temporary
process. In many community initiatives, it is this type of infrastructure—the lead
agency or staffed coalition that can maintain the process of linking organizations
and furthering a neighborhood agenda—that is often considered the element
within an initiative that should be sustained (Cornerstone Consuting Group,
2002)

The fact that these Network Centers are now operating without the benefit
of initial foundation operational "seed" money but instead secure their funding
from a mix of institutional partners and outside grants for programmatic activities
suggests that the gradual process of development used in this community has

been at least somewhat successful in developing infrastructure that is not

dependent on a large-scale external initiative to maintain. However, it is

132

important to note that the Network Centers are still a fairly new model, and there
is still a ways to go before they can be considered fully sustainable and
integrated into the community system. For example, their funding—both their
primary operating budgets and the bulk of their programmatic grants—must be
secured from year to year; while heavily supported by local institutions, the
Network Centers are not a line-item in the budget of any local entity. As such,
their funding and their continued existence continue to be unstable, despite their
demonstrated and potential contributions to the community system. In addition, it
is unclear how deeply many of the relationships between the Network Centers
and their partners are grounded in the respective Network Center directors—
while these lead staff members show a high level of commitment to their work,
this raises the question of how deep the setback would be to the Network
Centers and to the community system itself if these key staff should leave for any
‘ reason. i
4.4 Limitations

As in all research, tradeoffs have been made in the present study's design
to balance the research questions being asked, the desired outcomes of the
study, the real-world limitations of the study context and resources available for
research, and the needs of the community participants. As a result, there are
several limitations to the design of this study that warrant explanation and
justification here.

4.4.1 Methodological limitations of the social network analysis

One of the key lessons of this study, apart from insights gained into

intermediary roles played by Network Centers, is simply a heightened

133

appreciation for the complexity introduced and caution required when using
social network analysis techniques. As described earlier in this document, the
original analysis plan was not a fruitful approach for understanding the role of the
Network Centers. While I was able to identify an alternative approach, this was
not without its own pitfalls, as described below.

First, given the way the bridging functions performed by the Network
Centers was assessed and analyzed, there is no way to assert specifically what
the actual level of bridging or brokering activity performed by each of these
Network Centers as the time of data collection actually was. Indeed, Gould and
Fernandez are explicit in their definition of the formal brokerage measure that this
measure captures capacity of a given organization to broker exchanges between
the disconnected organizations. Indeed, it would be unreasonable to suggest that
any broker could or should actualize every transaction it is capable of brokering.
This points to brokerage roles—when held by a true community-oriented entity
that is acting to its ability on behalf of the community as a whole - as one
measure of community capacity, but not necessarily of community activity or
mobilization.

However, the interviews conducted with the Network Center system staff
provided useful examples of this capacity in action. This triangulation between
the quantitative and qualitative data suggests that actualization of Network
Centers' brokerage capacity at minimum follows a logical correspondence to the
potential expressed in the patterns of Gould and Fernandez brokerage scores.

Thus, these quantifications of bridging/brokering role have potential to assist

134

communities implementing intermediary structures to assess the role and
function of these intermediaries in more concrete terms than those typically
provided by qualitative case studies alone. However, the extent of actual
brokering activity in any concrete way is an empirical task that this study does not
address. While refinement of measurement and follow up strategies is clearly
called for in order to strengthen this approach, the findings of the present study
are worth building upon for future applied and theoretical work in this important
and growing area of community change and service coordination strategy.

A second limitation is presented by the symmetrization approach used.
This approach was deemed feasible with both the information sharing and joint
planning measures: given the wording of the questions, I felt that it could
reasonably be inferred with both relation types that respondents reporting
moderate-to-strong ties are describing a tie with genuine mutuality, since ties at
this level correspond to a level of communication and activity that would suggest
organizations are actively engaged with each other to a substantial degree.
However, this inference did not hold up for the measure of resource sharing, a
case in which the reported tie is more likely to be truly directional in nature but in
which directionality was not captured by the measure itself. Brokerage measures
using resource sharing as assessed in this survey would therefore have been
highly problematic to interpret. For example, in a given brokered triad, the
organization appearing as a "broker" may very likely instead be simply a common
funder for—or common funding recipient from—the "brokered" organizations in

the triad. For this reason, the measure of resource sharing ties was dropped

135

from the analysis altogether. However, it could be argued that mutuality is
potentially unclear with information sharing as well. That is, while it was an
assumption for this particular analysis that sufficiently high ratings of information
sharing would likely imply some level of mutuality in the exchange, it is possible
for an organizational member to believe they share quite a bit of information with
another organization but that the exchange is not returned. Nevertheless, I
believe that this concern is mitigated—at least in the context of this study—by the
fact that the Network Centers have described functions that seem reasonably
likely to include two-way exchange of information rather than positioning
themselves as primarily either senders or receivers of information.

Another limitation of the social network analysis comes from the selection
of the survey sample. While an attempt was made to represent as many of the
most active organizations as possible, the simultaneous need to condense a
complex array of organizations in the community down to a manageable survey
size may yet have left out many other organizations that are either directly active
in local service integration or community development efforts (the primary criteria
for inclusion) or would otherwise have been informative to include and study. As
an example of the latter, it came to my attention after Phase I data collection had
ended that only one of the faith-based organizations included in the sample was
an African-American church, despite the fact that there are many such churches
in the city. While racial orientation of organizations was not an explicit factor for
inclusion in the sample or a question in this particular study (although prominent

organizations specifically serving the African-American community were included

136

in the sample), in retrospect this raises interesting questions such as: Why
weren't more African-American churches identified? Does this hint at
segregation of these churches from the overall community and neighborhood
systems? If so, what are the dynamics behind this, and how could this inform the
Network Centers' work?

4.4.2 Methodological limitations of the qualitative follow up interviews

A frequent, if not ubiquitous, feature of social network analysis—
particularly when analyses are descriptive in nature—is the need to ground data
interpretation in a solid understanding of the context of the social system under
study. This can be said as well of community-based research in general.
Therefore, it is important to recognize that the follow-up phase of this research
only included formal data collection from Network Center insiders in the form of
interviews. My experience observing these Network Centers in action over the
past several years leaves me little doubt of the veracity of their accounts of their
activities; however, it must still be acknowledged that these stakeholders view
their communities through a particular lens. Overall, this study could have been
strengthened with the collection of data to contextualize the Network Center
brokering capacities and activities from a broader range of sources.

4.4.3 Overall limitations.

This study is limited in its scope, in that it focuses on the set of
interorganizational relationships within one urban community, and it cannot be
assumed that the analyses generated by this research are transferable to other
contexts. However, I have endeavored to provide sufficient description around

the overall community setting and the individual Network Centers and their

137

corresponding neighborhoods to enable consumers of this research to draw their
own conclusions about the extent to which these findings might be applicable to
other organizations and contexts of interest. ‘

4.5 Directions for Future Research

As an exploratory study of a particular type of neighborhood-based
intermediary in one community, this study provides a stepping stone for the great
deal of research that is yet needed on these approaches to community capacity
building.

While the brokerage measures used in this study did not fully assess the
brokering activity of the Network Centers, they did reveal differential patterns of
brokering capacity that reflected differences in the ways the Network Centers
operated, the capacity of their directors, and the capacity of their constituent
neighborhoods. As such, this measure shows promise as a tool for examining
‘ the roles that other intermediaries and broker organizations play in their
communities, thereby building a basis for better theory development around how
these organizations operate and how they can best be structured and
implemented to serve their intended aims. As such, I propose three directions for
future research in this regard.

One clear direction is to extend this work to other communities and other
neighborhood-based intermediaries in their many forms, to build a base of
knowledge of how community structure and community characteristics relate to
the capacity of organizations to serve in brokering roles. This work could include

comparative research across both varying types of neighborhoods and

138

communities as well as varying types of neighborhood-based institutions. For
example, are there particular forms of neighborhood-based intermediaries that
make more effective broker organizations than others? In addition, to what
extent are other service-providing intermediaries also brokering relationships on
a broader level? Future research might also take a cross—community view by
examining to what degree the presence or absence of brokering agents in a
community system affects that community's overall capacities for learning and
action.

Another direction is to empirically attend to the question of the extent to
which capacity to broker (as identified by the Gould and Fernandez brokerage
measure) translates into actual brokering behavior, and under what conditions
this happens. If this study had been designed with use of the Gould and
Fernandez measure in mind, some of this work might have been accomplished
through the addition of targeted items in the survey. For example, participants
could have been asked more pointedly whether each relationship they cited had
benefited from the work of the Network Centers in any way.

A third direction is to explore how the dynamics of these relationships
between Network Centers and the community and between organizations at
varying levels of the community shift over time. This might shed some additional
light on the role of broker organizations in communities.

For example, one open question is the extent to which Network Centers
might maintain an important role among organizations once these organizations

have gained experience in working together. Some research has found that

139

previously unconnected organizations are more likely to enter an alliance if they
have common partners (Gulati, 1995). Is it possible, then, that as the Network
Centers continue to create the space for neighborhood-level organizations and
city-level organizations to work together that these organizations will find that
they can work together without this intermediary involved?

We do not presently have the data to answer this question. However, the
data that we do have raises interesting possibilities. For example, Neighborhood
A, the most organized neighborhood of the three studied, is also the most
centralized neighborhood, with its relationships largely forming around a central
"hub" of three dominant organizations: Network Center A, the neighborhood
Summit process, and the umbrella neighborhood organization. This is intriguing,
because the neighborhood associations in this neighborhood have a decades-
long history of working together—therefore, it can reasonably be assumed that
this social structure does not exist due to lack of access of organizations to each
other, but rather has evolved because neighborhood organizations in this
neighborhood find this to be a preferable way to organize their work.
Interestingly, this sort of centralized network structure was found by Provan and
Milward (1995) to relate to effectiveness of community systems in delivering
mental health services. Thus, it should be considered that the function of
intermediary organizations may not be simply to promote more direct
connections between organizations, but to maintain federations of organizations
and promote efficient operation of the overall network. However, there is likewise

a fine line between bridging structural holes and fostering those same structural

140

holes (Burt, 2000b). Further research is needed to explore and tease apart the
ways in which these intermediaries serve to enhance the social structure of
interorganizational relationships with a cautious eye towards understanding the
extent to which these organizations may act to make themselves necessary
partners out of their own organizational survival interest.

In addition to these questions, some lessons can be learned from the
challenges of this study for future research. It was expected that data from
this study might perform similarly to other interorganizational network studies
(e.g., Van de Ven & Walker, 1979) and show clearly defined subgroups that
would aid interpretation of the Network Centers' roles. As described previously in
this document, this was not the case. However, the lack of subgroup presence in
this data set and the use of brokerage measures does not imply that community
systems researchers should not test for subgroups—on the contrary, I am of the
opinion that such a test should be a preliminary diagnostic element for any
community systems network study; and subgroups, if found, should be examined
closely. Indeed, the lack of cohesive subgroups in this study's community
system is telling, as subgroups within this type of network could be seen as a
sign of confined opportunity. Laumann et al (1978) define an opportunity
structure as:

"...a subnetwork within which exchange relations tend to be

confined as a function of the resources involved, legal or

institutional constraints on permitted partners, geographical

proximity, functional similarity, or preexisting organizational

overlaps. Organizations in different opportunity structures may be

severely limited in their effective access to each other, even though
they have complementary needs and resources.” - p. 471

141

Thus, examinations of community capacity that draw on social network
analysis techniques should clearly attend to whether these divisions in
opportunity structures exist in particular communities and what they mean for the
functioning of those community systems. This represents yet another area for
empirical exploration.

Future research should also explore alternatives to the methods of
capturing change in the network attempted in this study. As discussed in Section
2.2.6 above, the change data for this network, as measured, did not provide
productive avenues for furthering the analysis. It may be in part that these
retrospective questions were simply too broad in time span and tended to rely on
individual participant recall of organizations on a group by group basis. As a
result, this measure seemed to have questionable reliability in practice, as
participants varied widely in their ability to recall the organizations in the prior
section. While 63 of the 70 participating organizations referenced a shift in at
least one relationship, the distribution of this reporting tended toward the
extremes, with many organizations referencing only a few changes and others
citing changes (particularly in reference to new or stronger relationships) for as
many as half of the organizations in the sample. Considerably fewer changes
were reported for relationships that had grown weaker, possibly as a result of
social desirability bias. While beyond the intended scope of this study, a genuine
longitudinal data collection may have been a better choice for capturing the

desired change dynamics.

4.6 Conclusion

142

Neighborhood-based intermediaries/"broker” organizations offer promise
as a means for building community capacity and better local infrastructure to
support productive interorganizational relationships. This study examined one
community’s implementation of this approach and revealed both that the actual
roles and capacities of these broker organizations can vary widely even within
one community. Further, formal brokerage measures were identified as a
potentially revealing analysis technique for exploring the dynamics of these
intermediary organizations and their roles and for developing more and better
theory about how these approaches can and do function within community
systems. As such, this study offers important potential contributions to our

knowledge of and practice with these types of community initiatives.

143

APPENDICES

144

APPENDIX A

Participants by Organizational Role

145

Table 9. Participants by organizational role

 

 

 

 

 

 

 

 

Mean years
Organization Type Role Type N with
organization
City-Level Organizations
Agencies! Institutions Administrator/executive (e.g., director or 9 12.2
director of major division)
Admin/Boundary spanner (e.g., 4
administrator/coordinator of “community
relations” or “community projects")
City/County Government Administrator/executive (e.g., director, 1 1 17.2
coordinator, supervisor)
Nonprofit Organizations Administrator/executive (e.g., director, 8 10.1
coordinator)
Admin/Boundary Spanner (e.g., 8
director/administrator of community-related
division)
Neighborhood-Level Organizations
Neighborhood President/vice president/coordinator 1 1 10.2
Associations
Schools Principal 10.7
Other administrator (e.g., Director of
Student Services) or boundary spanner
Faith-based Pastor/associate pastor 3 23.3
Organizations Other boundary spanner (e.g., coordinator 1
for community development programs)
Other
Network Centers Director 2.3

Other intermediaries

Director/Coordinator/Facilitator

146

APPENDIX B

Invitation to Participate - Phase I

147

[Date]

[Name]
[Organization]
[Address]
Dear [Name]:

Michigan State University, in collaboration with the Neighborhood Network
Centers and their partners across the city of Lansing, is conducting a study to
understand how agencies, organizations, and institutions at the city level form
relationships with each other and with neighborhood organizations, schools, and
other groups at the neighborhood level. This research is intended to help
organizations such as yours gain a better understanding of how different groups
work together, what factors impact the development of collaborative
relationships, and how these relationships can be enhanced for the benefit of
neighborhoods and organizations in Lansing.

As an initial phase of this research, we are conducting surveys with
representatives from agencies and organizations across the city and in the
neighborhoods that Network Centers serve. [Organization Name] was identified
for the important role it plays in the lives of neighborhood residents. We would
very much like to include your organization in this study.

This survey would be conducted by telephone with a researcher from
Michigan State University and is expected to last between 30 and 40 minutes. If
you choose to participate, your organization will be entered into a lottery to win
$100. For this survey, we will be asking questions about the extent to which your

organization has shared information, shared resources, and/or engaged in joint

planning and programming with other organizations in Lansing. For this reason,

148

we are interested in talking to the person in your organization who is likely to be
the most knowledgeable of the different connections between your organization
and others.

Someone from our research team will be calling you within the next two
weeks to determine your organization’s interest in participating in this study and,
if your organization is interested in participating, to decide who might be the most
appropriate person from your organization to interview.

Thank you for your consideration. If you should have any questions or

concerns, please feel free to call me at (517) 355-3825.

Sincerely,

Shelby Berkowitz

Michigan State University

149

APPENDIX C

Phase I Informed Consent

150

STATEMENT OF INFORMED CONSENT

Michigan State University, in collaboration with the Neighborhood Network
Centers and their partners across the city of Lansing, is conducting a study to
understand how agencies, organizations, and institutions at the city level form
relationships with each other and with organizations, schools, and other groups
at the neighborhood level. This research is intended to help organizations such
as yours gain a better understanding of how different groups work together, what
factors impact the development of collaborative relationships, and how these
relationships can be enhanced for the benefit of neighborhoods and
organizations in Lansing.

As part of this study, we would like to learn about your organization’s
collaborative relationships with other organizations in the city of Lansing. We
would like you to participate in an interview with staff from Michigan State
University. The interviewer will ask you to describe your organization, the degree
to which your organization collaborates with specific other organizations in the
community, and any significant changes that have occurred in these collaborative
relationships over the past few years. The interview will be conducted by
telephone and should take between 30 and 45 minutes to complete.

As a way of thanking you for your participation, if you choose to participate in this
interview you can be eligible to win $100 for your organization in a lottery.

Participation in this research is completely voluntary. You may choose not to
participate at all in this study. If you do choose to participate, you may refuse to
answer any question or end the interview at any time without consequences.

Participation in this research is confidential. Only members of the Michigan State
University research team will have access to the completed surveys or other
information obtained through the interview. All of the information will be treated
as confidential to the maximum extent permissible by law. Only common themes
and patterns will be described in any reports emerging out of this research — at
no time will your name or any information specifically identifying you or your
organization be included in any report of the results.

If you agree to participate, please sign below and return this form in the enclosed
envelope to:

MSU/Department of Psychology
Strengthening Connections Project
135 Snyder Hall
East Lansing, MI 48824-2945

151

 

Agreement to participate

l have read or have had read to me the above information on this research study.
I voluntarily agree to participate in this study. I understand that I am free to
withdraw from this study at any time and that my participation in this study will be
confidenfiaL

 

SIGNATURE:

 

 

NAME (PRINTED): DATE SIGNED:

 

 

If you have any questions about this study or your participation in this study,
please feel free to call the MSU project at (517) 355-3825.

If you have any questions about your rights as a participant in this study, you

may contact Ashir Kumar, MD, University Committee on Research Involving
Human Subjects, (517) 355-2180.

152

 

APPENDIX D

Phase I Pre-Interview lnforrnation

153

Michigan State University, in collaboration with the Neighborhood Network
Centers and their partners across the city of Lansing, is conducting a study to
understand how agencies, organizations, and institutions at the city level form
relationships with each other and with neighborhood organizations, schools, and
other groups at the neighborhood level. This research is intended to help
organizations such as yours gain a better understanding of how different groups
work together, what factors impact the development of collaborative
relationships, and how these relationships can be enhanced for the benefit of
neighborhoods and organizations in Lansing.

As an initial phase of this research, we are conducting surveys with
representatives from agencies and organizations across the city and in the
neighborhoods that Network Centers serve. This data will be used to construct
visual representations (network maps) of the relationships between organizations
in the community system. This will help us get a better understanding of how
different kinds of organizations “fit together” and also guide follow-up study.

We will ask you questions about the relationships your organization
currently has and has had in the past with other organizations and groups in the
city of Lansing. Different organizations work together in multiple ways and with
varying degrees of intensity. Some organizations and groups have little to no
contact with each other. Others have very active relationships where they share
information, are involved in joint projects, share resources such as money or
staff, or help each other in other ways.

During the interview, we will ask you about the extent to which your
organization shares information, shares resources, or engages in joint planning
or programming with a number of other organizations. These kinds of
relationships are summarized for you below:

 

Kinds of Relationships:
Information Sharing information about clients, services offered, future
Sharing plans, issues, etc. Information sharing can occur in a variety

of settings, including during advisory board or coalition
meetings; during one-on—one or small—group meetings; or on
the telephone.

 

 

 

Resource Sharing financial resources with another organization,

Sharing sharing staff time, sharing volunteers, or sharing space
arrangements.

Joint Planning Working together with another organization on programs,

and planning activities or strategies together, conducting activities

Programming (such as outreach) on behalf of another organization, or
helping another organization to write grants or do other
activities.

 

 

154

 

We will ask you to rate the extent to which your organization has these types of
relationships with each organization we name on a scale of zero to five:

 

Not at all

 

Minimal

 

A little

 

Somewhat

 

Quite a bit

 

 

 

(fi-bQJN—IO

To a very great extent

 

155

 

APPENDIX E

Phase I (Social Network) Survey

156

Michigan State University, in collaboration with the Neighborhood Network
Centers and their partners across the city of Lansing, is conducting a study to
understand how agencies, organizations, and institutions at the city level form
relationships with each other and with neighborhood organizations, schools, and
other groups at the neighborhood level. This research is intended to help
organizations such as yours gain a better understanding of how different groups
work together, what factors impact the development of collaborative
relationships, and how these relationships can be enhanced for the benefit of
neighborhoods and organizations in Lansing.

As an initial phase of this research, we are conducting surveys with
representatives from agencies and organizations across the city and in the
neighborhoods that Network Centers serve. This data will be used to construct
visual representations of the relationships between organizations in the
community system.

Your responses will be confidential — at no time will your organization’s
name be used in conjunction with the data you provide. I would also like to
remind you that you may choose not to answer any question and you may also
end your participation in this survey at any time.

The following questions will ask you about the relationships your
organization currently has and has had in the past with other organizations and
groups in the city of Lansing. We are interested in getting the most accurate
picture possible of the way organizations are networked with each other in

Lansing. Therefore, if there are any questions you don't know the answers to or

157

which you feel are best answered by someone else from your organization,
please let me know and I will be happy to contact that person separately. This
survey should take approximately 40 minutes to complete.

For the purposes of this survey, the organization you are responding for
is: j . Please
answer all questions on behalf of this organization.

Do you have any questions before we begin?

SECTION A:

 

First, we would like to get some general information about your organization.

 

1. What type of organization does [organization name] represent? [check all
that apply]

[:1 Neighborhood [:1 Human Services E] Elementary,
organization or association Agency (02) intermediate, or
(01) secondary school
(03)
[:1 Local government entity [:1 Private business or [:1 Faith Based
(04) commercial organization Organization (06)
(05) '
[:l Public institution (07) [:ICollege or university [I Coalition or
(08) Collaborative
Group (09)
C] Other (specify) (10):

 

2. What are the primary focus area(s) of your organization? [check all that
apply]

[I Health (01) [I Public Safety (02) D Social Service
(03)

I] Economic Development 1:] Community [:| Education (06)

(04) Development (05)

C] Other (specify) (07):

 

3. What is your position or role in this organization?

 

 

4. How many years have you been with this organization?

 

158

 

 

 

5. Is your organization involved with any local collaborative efforts or multi-
organizational advisory boards? (Read list and check all that apply):
Coalitions or collaborative efforts (please list):

[:1 Human Services Advisory Council (HSAC) (01)

Healthy Communities/Summit Processes

[:l NorthWest Lansing Healthy Communities/Summit Process (02)
[:1 East Side Healthy Communities! Summit Process (03)
E] South Side Healthy Communities/Summit Process (04)

Neighborhood Network Center efforts:

[:I Citywide Network Center Board (05)

[:l Baker/Donora Focus Center Advisory Board (06)
[:1 Wexford Community Schools Network Center (07)
E) Allen Neighborhood Center Advisory Board (08)
E North Network Center Advisory Board (09)

E] South Network Center Advisory Board (10)

 

159

 

SECTION B:

Different organizations work together in multiple ways and with varying
degrees of intensity. Some organizations and groups have little to no contact
with each other. Others have very active relationships where they share
information, are involved in joint projects, share resources such as money or
staff, or help each other in other ways.

First, I would like you to think about the ways that [your organization]
currently works with other organizations in the community. I am going to go
through a list of organizations with you. For each organization I name, I am
going to ask you about the extent to which your organization shares information,
shares resources, or engages in joint planning or programming with that
organization.

These types of relationship are listed on the “cheat sheet” we faxed you
along with the informed consent form that you signed. I can go over these
bfiefly.

Examples of information sharing include sharing information about
clients, about services offered, future plans, issues, etc. Information sharing can
occur in a variety of settings, including during advisory board or coalition
meetings; during one-on-one or small-group meetings; or on the telephone.

Examples of resource sharing include sharing financial resources with
another organization, sharing staff time, sharing volunteers, or sharing space
arrangements.

Examples of [oint planning and grggramming include working together
with another organization on programs, planning activities or strategies together,

conducting activities (such as outreach) on behalf of another organization, or
helping another organization to write grants or do other activities.

Do you have any questions so far?

I am going to go through the organization list now, and for each
organization I name, I am going to ask you about the types of relationships I just
described. I would like you to rate the extent to which your organization currently
has that type of relationship with that organization, on a scale of zero to five, with
0 meaning “not at all", 1 = minimal, 2 = a little, 3 = somewhat, 4 = quite a bit, and
5 = to a very great extent. These ratings are also shown on the “cheat sheet".

Do you have any questions before we begin?

160

 

 

L250 .8505. .95. mm. E0: :.
wLocomeLEEoo
Lo Emom @500 E95...

 

 

 

 

0: 00.. :35 520.. 5.020
coL.L.moo
95:0... 0530.. .2020
98.23 LcaEEoo 030.00
69.90 5200
9.80...
chzEEoo .025 cLLVEan.
.0050
0.0%. >ocom< Scoucmnouc.
>LLEm... $920.5. .0 0.05

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mmoLo 00m. :00:0E<

 

 

 

0.359.. 5.00... cwoLLmE< c00E<

 

 

 

 

023m 9.. a. :0 L20. wcoL.m~Lcm Lo $05 .480 99: :0 0.8 L: .99: cm .. con .2. 5.3

cm mL 99.. .L 305. 0:. .0. 8.0 :00 20> .2. 2 now: .L :05 L960; >..:8_..Lcm.m m. Lo 5985 >...c8.....cm.m cozom m9. £me; 30. .8... .05 .cLLL.L>>
2,0: m. coLLmNLcmmLo .05 5.; 25:25.9 98.85590 50> $59.3 :oLEwE o. 20> 9... 8.0 u L .m:o..m~.cmmLo 0mm... Lo 5000 20:25 on . mm 262
52... .mL. .920 m c :9... o L. L 8.5 0.8 0E8 2:. $9.. no :2 o> 03 .5 m m E... .

o... 5.; 20.8.9 m>mc .0: >9: Lo >0E 0305 Lo coLENLcmmLo 50> :0ch 3033.590 Lo mmLocmmm waoLLm> Lo 7... m 90.. . .56 o L. .m0Loc0m.<

 

 

 

 

 

 

 
 

 

 

 

 

 

 

 

 

mcLEEmLmoLn. 05.05 9:55
. w @5591 .52. 0058?. :oLfiELoEL
mmozo. 02 u x .. . «oEo:
mmx<M>> 103.2 u >> LL05 mL .023 $05. 0.00 mchwLEtstwcm 02 n 8 £505. Eon... H mm
«@020me . 3:0; 023 8:05:09... 3:93 .005 b0> m o. n m 0cm ..Ln m 2ch u v .LmEsoEOm
1032 u w ;.._..50> :0 0:00:50 9.0:. u m 6...: m u N ..0EL:LE u w .052 u o LEExm:
>>wz u z . . . . L . . . .m... coL.m~LcmmLO
L . . m . . . . o . o L m <

 

161

 

 

 

   

000L000
00000.01 00E3I 00.000. .0 >..0

 

“:50... 00000.9 0>00 .0 LE 00L.0NL00 L0 50 .00. 00.000 0

 

 

   

 

.03EE00 000

   

 

 

 

00520 00.05:... .0000

 

00500 00005.. £000.50...

 

2.5L; 00000.0. 0>00 .0 LE 00..0NL00 0 50 .00. 000L.0~L00 L0 00000 5.0.. .000. 0E00 .3000 x00 0. 00... 0..

 

 

00500 0 0... .50.).

 

02.0050“. 00LL0>>

 

0090030.... .0 ._0.000 L00.0:.0>

.25.

 

0260005. 00.00 0.3..

 

00.0005 000.03. 50300.;

 

0. < 00 0050 0300-5.

 

000.0 m 5.00... 3000 0..

00.00.5500

   

 

>004. 00..0>.0w

 

L00.0:.0> 5.000 0050”.

.500. semen.

 

(02> 000n— x00

 

 

8:2 800.230 580.25.

 

.9025 cm 02.2 0252. mo

 

08.000 02.3.5200 :0 .55.

00.00LEE00

 

 

9020
02.00080.

000000
00.50000.

00.050590
.0 05000.0 .50..

 

 

 

 

mwOZO. 02 .I. x
mmx<w>> 103.2 N >>

mwwzomhw
IODE n w
>>mz n 2

«0.00.0
LL00. 0. .055 0300..
use; 2.; 00500090
50> :0 0000.000 205
0. .>> . .

0.00 00.00.500.300 02 H mm .505. .000. u 00

30000 .005 >L0> 0 0. u m 000 .._0 0 0.50 n v ..00>>0E00
u m .0..... 0 u N .L0EL0L0L n F .0002 n o n.00.me

.0. 0 0 00 00..0~L00 L0 50 0000 .0006 .003 0 H

 

.0... 02.050090

 

 

 

 

u.

 

 

 

o . 0 . m

 

<

 

 

162

 

 

   

0000.00... .02 00.0.00m

 

 

0 .00 .000. 0E00 5000 x00 0. 00... 0

 

  

.<.z 002.00 ...00500
.302 000000000... 0000.000 .02

 

 

0.L0t< 000.5 3m.)—

 

.

5850 :02

 

 

 

 

00.0000m. 30.2 0000 E00 0.
. .9020: 0.06 00 .00..)—

 

 

II|IL
..EE:w 005 0.300

 

:EESW 02m «WNW

 

  

 

 

.0000000Ln. ..EE

:0 000.0000 .02 000 0025...... 00...:0EE00 5.00...

0250.0. 00..L0...EE00
>0..00... 00.000.. 002.0002

 

 

8.00.0 .8000
0.050505. >.0:00 E0000.

 

>00L0..<
00.500021 >.0:00 E0000.

 

 

500 0.0001 0300 E00 0.

 

.00E00m0n.
5.00... 20:00 E0000.

 

 

 

E 02.0000 .0 50 .00. 00.000 0 0:00 00. 5000 :0 x00 0. 00... 0.. .502 00.000 < 0:00

 

(E E0300 E0000.

 

 

2.0.L2. 00000.0. 0.00 .0

 

0.5.0 .850 0.000..

 

 

.00E.L0 0n. 00:00 0.0000

 

00.50001 .0 0000 00.0000

 

LL00200 0090000902 00.0000

 

00.00LEE00 0.000... 0.000..

 

 

 

00.EE0.00.n.
a 00.000... 0.2.

00.5.0
0050001

00.020
0000520.

 

 

 

 

 

 

0.00 00.00.502.000 02 u 00 r265. .00? n 00

9000.0 .020 >.0> 0 0. u m 000 ..L0 0 0.50 n v ..002.0E00
u m .0..... 0 u N ._0EL0.E n F .0002 u c 00003.

.0. 0 0 00 02.00.00 .0 50 0000 .0008 .002. 0...

 

 

.0... 00L.0~L000.O

 

 

 

 

 

  

00020.. 02 u x no...
1va<m.>> 103.2 u >> 00... 0.50.5 0305.
10020100 , 0303 003 000000090
I022 u 0 . 50> 00 000.00.000.05 .
2.02.2 0.... a L...

 

D

 

. 0 . m

 

<

 

163

 

.
_

..m I E0.00m

_ .8090, .0005 00.2.

5.00.0000 0. 0.00000 00. .0 0500 .0 .0._ 0 0>00 . 2.02 0.00000

 

.0500 0.02002
.8000 3.0:EE00 0.00.0.5

 

.0500 0.03.02 0500

 

.0500 0002002 ......02

 

 

.0500 0000.000 .02 00.2

 

 

.0500 0200”. 0500000000

0..2. 00000.0. 0>00 .0 .E 00..0~.00 .0 50 .00. 0.0.000 0.02002 .000. 0500 5000 0.00 0. 00... 0 _ 2.02 0.0500 0.02502

 

.<. 2 0.00.0... 0.00.0;

 

 

.0002 0000.000 .02 00.0.00>>

 

 

00.0>> 0000.000 .02 00.00.:ow

 

 

.<.Z 000.020....”— «OQ mo._n_

 

. .0 0000.000 .02 00.0>>\..0..on_

 

.<.z 80E

 

00..0.0000<
0000500902 .003... 0.0

 

00..0.0000<
3.0.5500 02.0.0002

 

 

00..o< 0. 00.5... 0.000 .02

 

00..0.0000< 30...>>\000>>=00v_l

 

.<. Z O0®Il

 

.<. 2 00000001

 

..< Z 5:0... mo...

 

....1_<u.

 

000000005

 

00.EE0.005
.0 00.0005 5.0...

 

9.0000
0050001

 

00.500
02.00005.

 

 

0.00 00.005.02.000 02 u 00 €2.05. .00? u 00
.5050 .00.0 00> 0 0. u m 000 ...0 0 0.50 n 0 ..002.0E00
n m .0..... 0 u N ..0E.0.E u r .0002 n o ”50.xm.

 

”0. 0 0 00 00..0N.00 .050 0000 500.0 .002. 0 .0

.0.0 00..0N.000.O

 

 

 

 

 

00020.. 02 n x 00.00..
Immxfia. ....00l2 u 2. ..o... 0. .002. .265.
1002050283 9.2. 503.503
1.035. n w . .00.. 0.0 0000.000 0.00.
2.sz" 2.. 0. 43.004.00.00...

. . ,m , . ,

 

o

_

0

m

 

 

<

 

.2669 2259.6 9: .8

8:999 ncm wmho ..xa new ...>>.. ...w.. ...Z= 32.52 9 wwmma >023 €050: UEQQEoo £9.95 QE Hm._.OZ mw>>m=>mm.r2:

n0 20:0ww

50> ..8 2.2625 3.30:8 33 a 9m: 33.. _ 262 am: So :0 m>mc m3 wco=m~_cmm.o 9: ho =m 39:. 3me

 

5:96 00 EmE o_m>mo 53>

 

mcwtmmm mczo>

 

E60 58.

 

£359...

 

moth< ho $308 30

 

220 2.5 can mxom

 

 

 

E COSNNEN 50 530 was“ mgocm m 5240

we: we QEOw waonm

85:2 .6 £308 om

 

xmm 2 0x: P_ 302 .mm_o:w < 530>

 

 

.. ...5? 96:29 99. E

Bozow EcwEom >>o=_>>

 

 

 

 

Emu8< 50:2“. 5:35

 

.wr Scam

 

 
 

_oozom QUE—2 =_ cotan—

 

625 222.2 26

 

 

_oocow EcoEEw 0 2.2.502

 

Begum buEwEQm fun. 8505.

 

xEmumo<
.8ch 233.". 8963.22

 

 

Begum. ScoEmm mac 4

 

.m.I £050>w

 

mEEEEan.
w mEccmE E6...

mermcw

 

858mm

9.55

 

8:958:—

 

 

E
mm¥<M>> 103.2 M >>
lag

1022 n w

>>mz n Z

«05a:

.85 m. 355 «>65.
28; oz; coszEme
50> co 252:8 PBS

  

Emu m:_mm_Etm>>mcm 02 n ma, ”..;ocx race. u we
Ccflxw 620 E? m 9 u m new #5 m 95¢ u v .umczoEOm
u m 62: m n N __mE_:_E n v 6:02 u o €3wa

 

”E o m ca 535% 8.50 mmouacflxwficz o._.

«m: cozmuchO

 

 

u_

 

 

D

 

_ o

_ m

 

 

<

 

165

”0:0: 05500030508500 0:05 000 00> :0. . 05:00 0:: :0: 0:050000 :0 0::05800 :050 >:0 0>0: 00> 00 .06.: 50>
00.85% 0.8: _ 0:0 3:99. 00> :08 38> 00> 00“. 0><I _ 020:0000 05 :0 0.2 0.2:: 4.0.5 .2000

A ..... >:>>:00_:0: 00:05 00:00:02 :0:; 05 ..0: 00> :00

...0:0E>:0 :0: :00

5.? 9:05:00: 0 0>0: 0000 5:00:35 50> :05 0:0.:0N.:00:0 00 00:0: 00> 0.x 0>0: :. ..... mo.
0:00:00

:05 0. .0:0E>:0 :.:00 :00 0: 0:0.0: 0: 0000 5:05:35 50> :05 0:0.:0N_:00:0 >:0 ::0.::0E ...02 00 00> ”7.x 0: E

 

w0<n_.w meS. Dmmz ...: m0<n. “.0 ¥0<m mm: 00:5 :0>0 :0:.002, 550:0 0020:0506: 000...: 0>0: >...>>

E 55.00 :. ..>>.. 5.? 00:0 0005 0:08: 0000 0:00> 2,0: 0 00 0: 0000 >05 :05 30: 00.003 0:0 :05
00.:0:0.:0.0: :050 >:0 :0 0.55 00> :00 _. . 50.00; 0:03 00:00.05 00> :05 0030:0500: >:0 000 :.:00 . ”m0. .0:00> :50:
:00: 05 :0>0 :0:.003 :30:0 0>0: :05 0050:0500: 00 00:0.::0E 00>

 

A>I>> 0:.0:0:0:00:0 :0: 000:0:
w0<aw mes. 0002 30> “.. 00.0.0 “.0 v.0<m 00... «>03 0.5 :. 000:0:0 0020:0500: 0005 0>0: >55

E 55.00 :_ :0: 5.? 00:0 0005 0:05: 0:00:050 >_00:00_0:00
00 0:0: 0.002, 00> :05 0020:0500: :050 >:0 0:05 0:0 .:0.. 05 :0>0 :00: 0:0.:_...._. .0:00> :50: :00: 05 :. :0_:0~_:00:0
50> :0: :00:0::0 550:0 0>0: :05 09:05:00: 00 00:05:08 00>

 

005:0 000.2
0002 00> n: 00E :0 ¥0<0 000 88.58.60: 39: 0005 :0 2058.050 05 090:8: :0 850:. 0:2; a 89::

000.50: 0>0: 0020:0500: :50: 0005 >55 :0000 :5 0 0:: =0: 00> :00

E 55.00 c. ..z: 5:? 00:0 000...: 0:05: 0:050N.:00:0 50> :0:
:50: 0:0 :05 00:0: 00> 0320:0000: :050 >:0 0:05 0:0 .:0.. 05 :0>0 :00: 0:.:.:....... .0:00> 30: :00: 05 :0>0 :0.:0~.:00:0
50> :0: 0020:0500: 30: 0:0 :05 50.05505 00 00:05:00: 00>

 

166

APPENDIX F

Reported Tie Change Data

167

Table 10. Degrees of reported changes in relationships by type and region of
reporting and cited organizations, aggregating new/stronger reported ties and
weaker/defunct reported ties.

 

 

 

New/Stronger Weaker/Defunct
Cited relationship changes Mean Nomaalfed Mean Nomflfied
lndegree lndegree“ lndegree lndegree"
City Level
From City-level 2.479 0.069 0.854 0.024
Organizations
From Neighborhood A 0.125 0.016 0.104 0.013
organizations
From Neighborhood B 0.792 0.088 0.417 0.046
organizations
From Neighborhood C 1.625 0.181 0.208 0.023
organizations
Neighborhood A
Neighborhood 1.333 0.036 0.333 0.009
organizations (from city)
Neighborhood 0.583 0.083 0.000 0.000
organizations (from
neighborhood)
Network Center (from 4.000 0.108 0.000 0.000
city)
Network Center (from 0.000 0.000 0.000 0.000
neighborhood)
Neighborhood B
Neighborhood 1.889 0.051 0.111 0.003
organizations (from city)
Neighborhood 0.778 0.097 0.444 0.056
organizations (from
neighborhood)
Network Center (from 4.000 0.108 2.000 0.054
05W)
Network Center (from 2.000 0.222 1.000 0.11 1
neighborhood)

 

 

 

*Mean indegree represents the mean number of times each organization in a category was cited
(i.e., respondent organizations reported changed ties with that organization).

“Normalized mean indegree divides mean indegree scores by the number of participating
organizations in a given reporting group (i.e., the number of organizations that could potentially
cite a changed tie), allowing comparisons across cells.

168

Table 10, Continued

 

 

 

New/Stronger Weaker/Defunct
Cited relationship changes Mean Nomjfied Mean Noaneaallnzed
lndegree | n degree“ . lndegree I n d e gree“
Neighborhood C

Neighborhood 2.462 0.067 0.538 0.015

organizations (from city)

Neighborhood 2.000 0.250 0.154 0.019

organizations (from

neighborhood)

Network Center (from 5.000 0.135 3.000 0.081

city)

Network Center (from 3.000 0.333 1 .000 0.11 1

neighborhood)

 

169

APPENDIX G

Transitive vs. Brokered Triads at VarYing Cutoffs

170

80% r

70%

60%

50%

40%

30%

20%

1 0%

0%

 

INFORMATION SHARING

 

 

 

 

2 1 Z 2 Z 3 Z 4 Z 5
Dichotomization cutpoint

V-‘i—TRTANS—ITTVE TRIPLES As '%

OF IJK PATHS

, J+BROKERED AS % OF IJK

PATHS

- + IJK AS % OF TOTAL TRIPLES

+TRANSITIVE TRIPLES AS %

OF TOTAL TRIPLES

i +BROKERED AS % OF TOTAL

2., TRLPEES -

Figure 4. Transitive vs. brokered triads at varying cutpoints for information

shanng

171

JOINT PLANNINGIPROGRAMMING
80% , ,.

70%

' -OV:TARATNSITIVE mfiéé'xs %
OF IJK PATHS

+ BROKERED AS % OF IJK
' PATHS

+IJK As % OF TOTAL TRIPLES

60%
50%

40%

+TRANSITIVE TRIPLES AS %
' OF TOTAL TRIPLES

+BROKEREO AS % OF TOTAL
20% 7 7 . _ . TRIPLES _

 

30% , h-

10%

 

 

0%

 

2 1 2 2 2 3 2 4 2 5
Dichotomization Cutpoint

Figure 5. Transitive vs. brokered triads for varying cutpoints - joint planning
and programming.

172

APPENDIX H

Phase II Interview ProtocOl

173

 

NETWORK CENTER: DATE:

 

 

1. What do you see as the role of the network center in the community?

 

2. [describe bridging] - To what extent do you think that the network center
serves in this type of role?
a. Can you give me some examples?
b. Are there particular types of organizations that you see this role
strongly with?
c. Are there particular areas in which you think this operates very
strmqu?

 

3. SPECIFIC AUTHENTICATION CHECK-INS:

a. Do you experience the network center as brokering in these ways?
How so? How not?

b. In what ways are you brokering between these types of orgs?
c. Where do you think there are missed opportunities?
d. What helps you to use your role to effectively bridge orgs?

e. What gets in the way of bridging?

 

4. Do you think having an organization like the network center in a bridging role
is important?

5. What have you seen happen as a result of your bridging efforts?

 

 

6. What kinds of lessons would you share about being in this sort of role?

7. What advice would you give to a new network center?

 

174

 

APPENDIX I

Calculation of Standardized Brokerage Scores

175

### function to calculate expected values and variances for formal brokerage
### scores per Gould & Fernandez (1989)
##1##
m
GFeprrokerage <- function(data.frame, d, groupcolumn=2){
data.frame[,groupcolumn]<-as.factor(data.frame[,groupcolumn])
K <- nlevels(data.frame[,groupcolumn])
i <- as.matrix(summary(data.frame[,groupcolumn]))
i <-cbind(i,i-1,i-2,i*(i-1),i*(i-1)*(i-2))
x<-apply(i,2,sum)
##calculate expected values for raw brokerage scores
Ew1 raw <- ((d"2)*(1-d))*(i[,1]-1)*(i[,1]-2)
Ew0raw <- ((d"2)*(1-d))*(x[4]-i[,4])
Eb10raw <- ((d"2)*(1-d))*((x[1]-i[,1])*i[,2])
Etraw <- ((d"2)*(1-d))*(x[1]-1)‘(x[1]-2)
##calculate estimated variances for raw brokerage scores
vanNt raw <- Ew1raw*(1-d"2*(1-d))+(2*i[,2]*i[,3]*(i[,1]-3))*(d"3*(1-d)"3)
varw0raw <- Ew0raw*(1-d"2*(1-d))+2*((x[5]-i[,5])*(d"3*(1-d)"3))
varb10raw <- Eb10raw*(1-d"2*(1-d))+2*(i[,2]*(x[1]-i[,1])/2+ (x[1]-
i[,1])*i[,2]/2) *(d"3*(1-d)"3)
vartraw <- Etraw*(1-d"2*(1—d))+2*(x[1]-1)*(x[1]-2)*(x[1]-3)*(d"3*(1-d)"3)
##calculate expected values for partial brokerage scores
Ew1 part <- Ew1 raw * ((1-(1-d"2)"(x[1]-2))/(d"2*(x[1]-2)))
Ew0part <- Ew0raw * ((1-(1-d"2)"(x[1]-2))/(d"2*(x[1]—2)))
Eb10part <- Eb10raw * ((1-(1-d"2)"(x[1]-2))/(d"2*(x[1]-2)))
Etpart <- Etraw* ((1-(1—d"2)"(x[1]-2))/(d"2*(x[1]-2)))
##calculate estimated variances for partial brokerage scores
varw1 part <- van~1 raw * ((1-(1-d"2)"(x[1]-2))/(d"2*(x[1]-2)))"2
varw0part <- varw0raw * ((1-(1-d"2)"(x[1]-2))/(d"2*(x[1]-2)))"2
varb10part <- varb10raw * ((1-(1-d"2)"(x[1]-2))/(d"2*(x[1]-2)))"2
vartpart <- vartraw*((1-(1—d"2)"(x[1]-2))/(d"2*(x[1]-2)))"2
### output preliminary calculations
print(c(”CalcuIations based on network density: ”,d))
g<-rbind(i,x)
colnames(g) <- c("n","n-1”,"n-2","n*(n-1)","n(n-1)(n-2)")
rownames(g) <- c(1:(nrow(g)—1),'"I'OTAL")
printtg)
rawvalues <-
rbind(Ew1raw,varw1raw,Ew0raw,varw0raw,Eb10raw,varb10raw,Etraw,vartraw)
print(rawvalues)
partvalues <- rbind(Ew1part, varw1part, Ew0part, varw0part, Eb10part,
varb1 0part,Etpart,vartpart)
print(partvalues)

176

##fill in dataframe with expected values, variances, and calculated

standardized brokerage scores

data.frame$ExpectedCoordRaw <-

Ew1raw[unclass(data.frame[,groupco|umn])]

data.frame$ExpectedConsRaw <-

Ew0raw[unclass(data.frame[,groupco|umn])]

)A.5

)A.5

data.frame$ExpectedGateRepRaw <-
Eb10raw[unclass(data.frame[,groupco|umn])]
data.frame$ExpectedTotalRaw <- Etraw
data.frame$VarianceCoordRaw <-
varw1 raw[unclass(data.frame[,groupco|umn])]
data.frame$VarianceConsRaw <-
varw0raw[unclass(data.frame[,groupco|umn])]
data.frameSVarianceGateRepRaw <-
varb10raw[unclass(data.frame[,groupco|umn])]
data.frame$VarianceTotalRaw <- vartraw
data .frame$StandardCoordRaw <- (data.frame$Coordinator -

data.frame$ExpectedCoordRaw)/(2*data.frameSVarianceCoordRaw)“.5
data.frame$StandardConsRaw <- (data.frame$Consultant -

data.frame$ExpectedConsRaw)/(2*data.frame$VarianceConsRaw)".5
data.frame$StandardGatekeepRaw <- (data.frame$Gatekeeper -
data.frame$ExpectedGateRepRaw)/(2*data.frame$VarianceGateRepRaw

data.frame$StandardRepRaw <- (data.frame$Representative -
data.frame$ExpectedGateRepRaw)/(2*data.frame$VarianceGateRepRaw

data.frame$StandardTotaIRaw <- (data.frame$Total -

data.frame$ExpectedTotalRaw)/(2*data.frame$VarianceTotalRaw)".5
data.frame$ExpectedCoordPart <-
Ew1part[unclass(data.frame[,groupco|umn])]
data.frame$ExpectedConsPart <-
Ew0part[unclass(data.frame[,groupco|umn])]
data.frame$ExpectedGateRepPart <-
Eb10part[unclass(data.frame[,groupco|umn])]
data.frame$ExpectedTotalPart <- Etpart
data.frame$VarianceCoordPart <-
varw1part[unclass(data.frame[,groupco|umn])]
data.frame$VarianceConsPart <-
vanNOpart[unclass(data.frame[,groupco|umn])]
data.frame$VarianceGateRepPart <-
varb10part[unclass(data.frame[,groupco|umn])]
data.frame$VarianceTotalPart <- vartpart
data.frame$StandardCoordPart <- (data.frame$CoordinatorP -

177

“.5

“.5

data.frame$ExpectedCoordPart)/(2*data.frame$VarianceCoordPart)“.5
data.frame$StandardConsPart <- (data.frame$ConsultantP -

data.frame$ExpectedConsPart)/(2*data.frame$VarianceConsPart)".5
data.frame$StandardGatekeepPart <- (data.frame$GatekeeperP -

data.frame$ExpectedGateRepPart)/(2*data.frame$VarianceGateRepPart)
data.frame$StandardRepPart <- (data.frame$RepresentativeP -
data.frame$ExpectedGateRepPart)/(2*data.frame$VarianceGateRepPart)
data.frame$StandardTotalPart <- (data.frame$TotalP -
data.frame$ExpectedTotalPart)/(2*data.frame$VarianceTotalPart)“.5
output.frame <<- data.frame

output.frame
write.table(output.frame,file="clipboard",sep="\t",col.names=TRUE)

}

178

APPENDIX J

Brokered Triad Classification Routine

179

#########§t%#

modtriad.classify.R

Shelby Berkowitz
berkowi4@msu.edu

note: last tested under R 1.8 — may not work
with newer versions

usage: btriads(inputmatrix, nodelabels)

GENERAL REQUIREMENTS:
Must have sna and gregmisc packages already installed on system.

These can be installed directly off the online CRAN archive from within the menu
on the windows R console. If you intend to run this function repeatedly, it would
be worthwhile to remove commands under #1 from function body and instead put
them in global Rprofile or personal .Rprofile files.

##fifififififififififi

Input parameters:

inputmatrix - adjacency matrix, binary values

Optional inputs:
nodelabels - n length vector of node labels. Default is the labels 1 to n.

Outputs:

matrix named newtriadinfo with columns 1-7:

(1) broker, (2) brokee1, (3) brokee2,

(4) broker's ties, (5) brokee1's ties, (6) brokee2's ties,
(7) triad classification

btriads <- function(inputmatrix, nodelabels=NULL){

btriads = NULL

# 1 necessary setup - load SNA & gregmisc libraries, set recursion limit
to max
library(sna)
library(gtools)
options(expressions=1 e5)
# 2 assess input parameters
# 2.1 size of matrix
N = nrow(inputmatrix)
# 2.2 check if nodelabels are valid
# 2.3 assign nodelabels if does not exist

if (is.nu|l(nodelabels)) {
nodelabels <- 1:N

180

}

else {

if (length(nodelabels) != N) {

print("Number of labels must equal number of nodes")
return(NULL)

I

# 3 Assemble array of valid vectors to evaluate (unique combinations of
3 nodes from total set)

nodesets <- combinations(N, 3, v=1:N, set=TRUE,
repeats.allowed=FALSE)

triadinfo <- matrix(0,ncol=7)

#4 - run triad.classify function on vectors within array:
#4.1 - set up actions for triad classification:

#triclasslist <- function(ijk) {
for (ind in 1:nrow(nodesets)) {
ijk <- nodesets[ind,]
trioode <- triad.classify(inputmatrix, g=1, tri=ijk)

#4.1 b - translate Davis & Leinhardt triad codes to numeric (using
pajek sequence numbers) for vector compatibility

#Use following code when interpreting brokerage on nondirected
#relationships (e.g., non-reciprocated ties not interpreted as
#only "sending" or ”receiving” & therefore 021 D (outbound-only
#ratings from broker) & 021 U (inbound-only ratings to broker)
#still included as brokered relationships:

#if(length(tricode) !=0){

ptricode <- switch(tricode, "003" = 1, "102" = 3, "201" = 11, "300" =
16)

#410 - if triad code indicates brokerage, figure out who is broker
if (!is.null(ptricode)) {
i<-ijk[1]
l<-iikl2]
k<-ijk[3]
ities <- sum(inputmatrix[i,j], inputmatrix[i,i], inputmatrix[i,k],
inputmatrix[k,i], na.rm=TRUE)
jties <- sum(inputmatrix[i,j], inputmatrix[i,i], inputmatrix[j,k],
inputmatrix[k,j], na.rm=TRUE)
kties <- sum(inputmatrix[i,k], inputmatrix[k,i], inputmatrix[j,k],
inputmatrix[k,j], na.rm=TRUE)

181

tieconfig <- c(ities,jties,kties)

currtriadnos <- c(i,j,k)

# reorder triads so broker 1st (order by decreasing # of ties,
then by decreasing node #) _

triorder <- order(tieconfig,currtriadnos,decreasing=TRUE)

currtriadinfo <-
c(currtriadnos[triorder],tieconfig[triorder],ptricode)

# add to output matrix:

triadinfo <- rbind(triadinfo,currtriadinfo)

}
# (ELSE) OO NOTHING
#}
l

#5 - clean up matrix
# 5.1 - ditch first row of zeros
trows = nrow(triadinfo)
cleantriadinfo <- triadinfo[2:trows,]

orows <- nrow(othertriadinfo)
cleanothertriadinfo <- othertriadinfo[2:orows,]

#cleantriadinfo <- triadinfo

# 5.2 - replace node numbers with node labels

cleantriadinfo[,1 :3] <- nodelabels[cleantriadinfo[,1:3]] -
cleanothertriadinfo[,1z3] <- nodelabels[cleanothertriadinfo[,1:3]]
newtriadinfo <<- cleantriadinfo

newothertriadinfo <<- cleanothertriadinfo
write(t(newtriadinfo),file='triadoutput.txt', ncol=7, append=FALSE)

# append attribute fields for brokered triad output for easier cross-
tabulation in Access/Excel

attrappend <- function(inputmatrix, attrfile){
orgids <- attrfile[,1]
orgnames <- attrfile[,2]
orginits <- attrfile[,3]
orgregion <- attrfile[,4]
regiondesc <- attrfile[,5]
orglevel <- attrfile[,6]
Ieveldesc <- attrfile[,7]
orgclass <- attrfile[,8]
classdesc <- attrfile[,9]

182

outputmatrix <- cbind(inputmatrix,

matrix(orgnames[inputmatrix[,1 :3]],ncol=3), matrix(orginits[inputmatrix[,1:3]],
ncol=3), matrix(orgregion[inputmatrix[,1:3]], ncol=3),
matrix(regiondesc[inputmatrix[,1 :3]], ncol=3), matrix(orglevelfinputmatrix[,1:3]],
ncol=3), matrix(leveldesc[inputmatrix[,1:3]], ncol=3),
matrix(orgclass[inputmatrix[,1:3]], ncol=3), matrix(classdesc[inputmatrix[,1:3]],
ncol=3))
write(t(outputmatrix),file='attrappendoutput.b<t', ncol=31,

append=FALSE)

mycolnames <-
c(paste("B",0:2,sep="_”),paste(”nties",0:2,sep="_"),"tritype",
paste(”orgname”,0:2,sep="_"), paste("orginits",0:2,sep="_"),
paste("region”,0:2,sep="_"), paste("regiondesc",0:2,sep="_"),
paste("level",0:2,sep="_"), paste("leveldesc",0:2,sep="_"),
paste("class",0:2,sep="_"), paste("classdesc",0:2,sep="_"))

write.table(outputmatrix, file = ”attrappendoutput.b<t", append
= FALSE, quote = TRUE, sep = ",", row.names = FALSE, col.names =
mycolnames)

}

183

APPENDIX K

Brokerage Score Tables

184

 

185

 

 

2R- 8.N- £3- 9.? RT R. T ~.<-_.E<n_
ER- 8R- med- m;- R.T R. T T<-OOmm<Io_mz
ER- 8N- and. Rd- RT R. T m.<-OOmm<IO_wz
8&- BR- 8.»- Rd- 8.? 33. m-<-Oowm<_._o_mz
New- 8R- 8“. vod- R. T R. T T<+EE
8R- 8R- owd. 8.»- 8. T 8. T -<.OOmm<IO_mz
8R.- SN- 8s- 88- 8.0. 33- T<-Ome<IO_mz
RR- 8. T m:- 98. R. T R. T 330018
mo. T ~3- 3? Ed. R. T R. T -<-._OO:Ow
8. T 3. T E E ato- Rd. «LT-.0018
Rd- Eo- ood «mm and web m-<-OOmm<IO_mz
mod- mmd. m. 3R 3.: 3.8 E 5.5sz xmozfimz
Ed- 03 8.8. «Tow. $2 P .A «£5.23

. Ozizéooma

52.225; 5.0..
3m- 85.- SdT 8.2- mm. T R. T m.<.OOmm<IO_wz
ta- 2...”- 8.2- 8.2- 9.. T 3. T ~.<+E<n_
m;- ST 8.2- 8.2- 3. T 3. T m-<-OOww<xo_mz
m;- 85. 8.2- 8.2- 3. T 3. T T<-OOmm<IO_mz
8...”- 35- 8.2- 5.2. 3. T 3. T T<-:E<n_
8d. 95. 8.? Rd. R. T mm. T -<-OOmm<xO_mz
8a. 2a- 3? 8.»- 3T 3T T<-._OO_._Om
men- 35- 88. NS- 93- :6. T<-Oowm<zo_mz
RR- SR- and. Rd- ato- de- m-jOOIOm
muo- fld @ka «Imb- ch 8. T ~.<.._OOIOm
Rs. T 2T Em 8:0. 3.. E m-<.OOmm<IO_mz
86 Ed om. :- moR 8.9 E 47%:sz #6252
23 8.0 8.9. $3 and. P. 88 $2223

Oz_m<:w

20.2582.

< 39.58552
.38 82903:: .823 send-625 .9th sausage: zOF<N_z<OmO
Amzo_»<N_z<OmO cEO oz< oOOImOmIOmz
._m>m._->EO Ozozs zmmEmmv m>F<pzwmmmmmm zhmwwnmmmummnoo
pzfionmzoo EwammmeO

 

dab m: can 80:89.99. 3 .wvtozzocnzm £53, mcozm~_cm9o noon-59:99.. .2 $68 $99.05 .3 298.

186

 

 

mmlm mm R R 8 m: E Rad R-o-OOmm<:O_mz
E meld ll? :8 Ed. .R. .m E TO-Oaz
E E Iledm IRS mu! Morm- d-O-OOmm<.-.O_mz
dud Rd IIIR. Ilem 3 R d 0-52.sz xmozcwz

Oz.m<...m
29259.2.
o 805.8652
5 R- ER- odd- Rdd- 2d- Ed- Tm-._OO:Om

8R- TR- 3&- Rs- RE- NE- $-52“.

EN. EN. ts- dz- RF. T 9. T 5.15“.
an R- 3R- Rd. and. Rd. 3.? m-m-TEE
dew- dew- add- RT Rd- .3 m-m-doozom
8. T 8. T E RT 8d- Ed. utm-doozow
dv. T . 8. T wk .Rdle. ole-d- fld R-m-OOmm<IO_mz
8d- B.? a. d:- E E Tm-OOmm<xo.mz
dd F 3d I32 IR. 3 E mm Tm-..OO:Om
9. d- 8d. I59 '89 Rd and m-mmezmo xmozfimz

02.23508.
62.225". 5.0..
and- 8d- Rd- mad- 8d- td- Tm...OOxom

Rd- Rd- 8d- Rdd- 8. T 8. T «.m-EE
3d- 8d- 3d. Add- mm. T 8. T Tmitfi.
8. T 3d. 8d. Ed- dd. T 3. T mam-1E3.
2. T R:- dmd- Rd- 8T 3. T R-m-..ooxom
8d- 9&- E and. mwld 3d m-m-dooxow
3. T R. T mmd E 2d of Tm-OOmw<:O_mz
Rd 3.. ale: F dIB. 2.? R-m-OOmm<:o.mz
3d- 8. T I39 Rd clam Re 3.3sz 5.0252
film- ddd I? I8. : add R: 8.6018

oz_m<_..w
20:3sz.2.
m ogxmomzomz

Etna usage: .3th dead-.635 Egan. 829035 zo_e<N.z<Omo

82925255 cEO oz< oOOImOmIOmz
..m>w..->EO 02023 zmmzfimm. m>F<hzmmmEmm z_hmm>mnmmmnm_mnoo
52.5828 Emammxmio

 

omsczcoo ...? 03m...

 

 

903 399.05 O>=_doq Go. v 3 2285ch .2865 d959— Omstonc: “202

 

 

 

d R- Rdd- dd 2. dd d T 3.. T 3.. T TO-OOdd<IO.mz
dd R- Rdd- dd 2. dd 2- 3.. T 3. T THO-TEE
d R- 8d. dd d- dd d- Fdd. Rdd- To-oodd<zo_mz
dd R- ddd- dd d- dd d- .3. T 3. T «692
8d- dR- ddd- dd- 3. T 3. T «0.501%
dd R- dR- Rd d- ddd. dd T R. T d-O-Oodd<zo_mz
Rd N. dud- R d- dd R- R T dd. T d-O--_OOIOd
3. R- ddd- dd T R d- dR T dd. T 9:228
dd d- 8.? did-Id- NAIR- dd. T dd. T TO--_OOIOd
d T dd. T E mmld Rd dFd d-O-Oaz
dIle E .dd dd RIR-m ddd 9%:sz #62sz
fi WIN Ildddd IIdF dd Wm ddd dooodd<xomz
. d F .dlvldd IdliR FIR. E R.0-0odd<zo_wz
|dd : E Ildddv IIRdR dd dFd TO-Oaz

02.224808;
202.225“. 5.9

ddd- dd- 8.3- dd.:- dd. T ddT TOiEE
ddd- Rd- dd.:- dET dd. T ddT TO-OOmd<IO.wz
ddd- dd. 2...:- dd.:- dd. T dd. T 3.82
ddd. dud- dd. F T 8.2. dd. T dd. T d-O--_OO:Od
ddd- ddd- 8.2- 8.2. dd. T dd. T R-O-..OOIOd
ddd- Rd- 8.2. deT ddd. Rd. T Todowd<xomz
Rd- dtd- ddd- Rd- Rd. T Rd. T d-0-0odd<zo_mz
ddd- ddd- dds- dd? Rd. T Rd. T 0.222%
dd. T R. T g dd. dd. T Rd. T TO-..OOIOd
ddd- ddd- «dd dd; .dd dwd- d-O-On.z

.928 82982:: dean. 82983:: .985. 829825

62922255 GEO dz< doozmodromz

did->:0 Ozo2<v 232,»de w>£§zudwmdmm 2_me>wnwmmnm._mnoo zo:<~_z<omo

dzfidddzoo BEER-20

 

82.28 .3 20.8

187

Table 12. City-level organizations with significant positive brokerage scores for

information sharing between neighborhood-level organizations
(consultant/itinerant brokerage), by neighborhood subnetwork.

 

 

 

Consultant Brokerage Neighborhood Subnetwork
Information Sharing A B C

Raw Partial Raw Partial Raw Partial
GOVT-ClTY-14 fl M Q Q Q Q
AGENCYINST-ClTY-12 ZJ_9 Q E Q Q E
GOVT-ClTY-10 2A fl & 3A1 & &_.5_1
AGENCYINST-ClW-S 1.68 g m M & 11:51
GOVT-CITY-9 $_2_§ & 1-30 1-78 & &
GOVT-CITY-7 Q9 6._1__2_ 0.06 -o.12 1.84 fl
Gow.c|w-5 -1.21 -1.26 -0.93 -0.61 4;; m
NPO—ClTY-13 2.__J__Q L32 0.56 0.42 -1.32 -133
AGENCYINST-CITY-3 -0.02 0.70 2g 5._20 -1.02 -0.77
NPO-CITY-11 -1.21 -1.16 1.30 2.03 -1.32 -1.33

 

Note: Underlined figures are significant at the .05 level.

Table 13. City-level organizations with significant brokerage scores for joint

planning and programming between neighborhood-level organizations
(consultant/itinerant brokerage), by neighborhood subnetwork.

 

 

 

Consultant Brokerage Neighborhood Subnetwork
Joint Planning A B C

Raw Partial Raw Partial Raw Partial
GOVT-ClTY-14 6.41. 5.102 & iii 2.012 &
covr-cmr-10 3A AA m g _3.5_Q. fl
AGENCYINST—CITY-5 1.79 2._§_:1- ii 645 .4-_2§ 5._B_§
AGENCYlNST-CITY-12 Q Lg 1-43 1-08 Q &
GOVT.C|TY.5 -0.41 -0.64 -0.67 .041 g_0_g fl
NPO—ClTY-13 3._2§ 5:9; 1.43 1.08 -0.41 .043
AGENCYINST-ClTY-3 0.51 1.07 1.43 3.0; -0.10 0.28
GOVT-ClTY-12 1.06 1.70 -1.20 -1.20 1.93 2.___5___1
NPO-ClTY-11 -1.33 -1.33 1.96 2.;4 -0.88 -1.06

 

Note: Underlined figures are significant at the .05 level.

188

Table 14. City-level organizations with significant brokerage scores for

information sharing between city-level and neighborhood-level organizations
(gatekeeper/representative brokerage), by neighborhood subnetwork.

 

Gatekeeper/ Representative

Brokerage

lnfon'nation Sharing

Neighborhood Subnetwork

 

 

Raw Partial Raw Partial Raw Partial
GOVT-ClTY-14 fl flit & M :15—33 m
AGENCYINST-ClTY-5 31__-1_1 _55._gz_ dad-$2. L53. BAA) M
cow-curve iii—.02 55._12 fl M Q11 3A_-J_5
AGENCYINST-ClTY-12 @ M 2_5-__58 M M &
GOVT-ClTY-10 111—.11 fl & iii—59. M £53;
covr-cmr-7 @ fl 9;!) Q Q m:
GOVT-ClTY-S :L_5_2 m 5A fl 24$ 10.__&6
AGENCYINST—ClTY-3 Lia 244;: 12% L51 0.57 g
NPO-ClTY-13 & M g 9_.52_ M 1.17
NPO-ClTY-14 -1.79 -1.88 & M L431: fl
NPO-ClTY-G 4;; g 1.04 m ;3._9§ fl
AGENCYINST—ClTY-1O 1.49 1.31 -1.72 -1.43 @ &
AGENCYINST-ClTY-1 -2_.19 -2.95 g 10._25 -6_.4_1 1mg
NPO—ClTY-ll :.5._.6_S_9 51% Q Q . ;§._5_5 ;7._0£_3
AGENCYINST-ClTY-Z g m M M 1% in)
NPO-ClTY-12 fl -9. 66 #6 33—95 E Q
GOVT-ClTY-12 101 :3._92 & g -0.01 1.65
GOVT-ClTY-B g Q 1.39 1.80 -103 -1.47
AGENCYINST-ClTY-4 i8; i7_8 1-91 fl .1619 w
AGENCYINST-ClTY-8 m 0.01 -7.60 w ifl fl

 

Note: Undertined figures are significant at the .05 level.

189

Table 15. City-level organizations with significant brokerage scores for joint

planning and programming between city-level and neighborhood-level
organizations (gatekeeper/representative brokerage), by neighborhood

subnetwork.

 

Gatekeeper/ Representative

Neighborhood Subnetwork

 

 

Brokerage

Joint Planning Raw Partial Raw Partial Raw Partial
GOVT-ClTY-14 M L18 2_L_Q3 @ w M
AGENCYINST—ClTY-S iii—£1. M AA__§_6 Q fill; fl
AGENCYINST-ClTY—12 4.4-LIE L55 m L51 m 112-1.4.
GOVT-ClTY-10 ifl; E M 4.6__A2 BQ_.ZZ 4140;
GOVT-ClTY-S gag 1_&__&8 111—£2. @ iii-4.2 @
AGENCYINST-ClTY-3 12A 251—51 @ m m &
NPO-ClTY-13 M 512—433 a w 1-93 SLEZ
GOVT-ClTY-9 & 0.64 g g g Q
GOVT-ClTY-12 M M 1-39 43-83 E M
AGENCYINST—ClTY—10 4A g -1.25 -1.88 ;9_2_9 Lag;
GOVT-CITY-8 m _10._25 all m -O.88 ._2._3_7
NPo-cmrd g 4.__J=1_ 0.83 __2.32_ -_3J_0 ;5_.Q_6
NPO-ClTY-1 1 £12 -_6._7§ Q L51 £92 ;4;._8_7
NPO-ClTY-14 2.33.6 :3._1§ Mg Lil 1-31 0.53
NPO—CITY-12 M M 1.39 003 ' L51; Q
AGENCYINST-ClTY-8 m fl 1&6 M3 £1 -3.38
AGENCYINST—ClTY-1 if; M 1.35: m -5. 73 -632

 

Note: Underlined figures are significant at the .05 level.

190

 

APPENDIX L

Reported Ties and Tie Strengths by Region/Organization Type

191

 

 

Table 16. Reported ties and ties strengths by organization type and region.

 

 

 

Information Sharing . Joint
Planning/Programmim;__
Mean Mean of Mean Mean of
outdegree average outdegree average
nonzero tie nonzero tie
strength strength
Citywide
Agency/lnstitutio 54.46 3.05 48.77 2.94
n
City/County 60.55 3.13 53.73 2.75
Government
Nonprofit 46.33 2.84 34.33 2.75
Organization
Neighborhood A
Neighborhood 31.33 2.54 23.67 2.73
Association
School 54.00 2.97 48.50 2.77
Faith 14.00 2.07 15.00 2.00
Network 67.50 3.38 59.50 3.23
Center/Summit
Neighborhood B
Neighborhood 53.00 3.94 32.50 3.89
Association .
School 51.67 3.22 44.67 3.01
Faith 24.67 2.43 22.33 1.88
Network 42.00 3.90 38.00 4.08
Center/Summit

 

Note: Mean outdegree represents the mean number of ties reported across organizations; mean
of average nonzero tie strength represents the average strength of nonzero ties across
organizations.

192

Table 16, Continued

 

 

 

Information Sharing Joint
PlanrfiLq/ProgramminL
Mean Mean of Mean Mean of
outdegree average outdegree average
nonzero tie nonzero tie
strength strength
Neighborhood C
Neighborhood 62.67 3.45 54.67 3.22
Association
School 50.00 2.62 51 .00 2.63
Neighborhood 43.33 3.17 40.33 3.23
Association
Faith 0 0 0 0
Network 55.00 3.23 46.50 2.28
Center/Summit

 

193

APPENDIX M

Network Maps

194

ll

The figures on the following pages show network maps of the ties (weight
>= 3) in the Network Centers’ subnetworks. Figures 4 through 9 show maps of
the ties between neighborhood-level organizations only, with node size
proportional to each organization’s score for coordinator brokerage. Figures 10
through 15 show maps of the ties each Network Center’s entire subnetwork,
including both neighborhood and city-level organizations, with node size
proportional to each organization’s score for gatekeeper/representative

brokerage. Organization types are indicated by node shape as shown below:

 

NETWORK CENTER,

SUMMIT, OR OTHER

INTERMEDIARY

NEIGHBORHOOD AGENCY/

ASSOCIATION |N3T|TUT|ON

SCHOOL CITY/COUNTY
GOVERNMENT

FAITH-BASED NON-PROF IT

ORGANIZATION ORGANIZATION

 

195

 

4

\ /
‘ -- //

Figure 6. Neighborhood-Level Network Of Neighborhood A - lnforrnation
Sharing. Node size is proportional to coordinator brokerage score.

 

196

 

 

 

 

 

{/4
__ WW \
/ V“ /

Figure 7. Neighborhood-Level Network Of Neighborhood A - Joint
Planning/Programming. Node size is proportional to coordinator brokerage
score.

     
 
 

 

 

 

 

 

 

 

197

     

NETWORK CENTER B

 

Figure 8. Neighborhood-Level Network of Neighborhood B -' Information
Sharing. Node size is proportional to coordinator brokerage score.

198

Ah.»
\Y’
\45

NETWORK CENTER B

       

 

 

   

 

Figure 9. Neighborhood-Level
Network of Neighborhood B — Joint Planning and Programming. Node size is
proportional to coordinator brokerage score.

199

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 10. Neighborhood-Level Network of Neighborhood C - Information
Sharing. Node size is proportional to coordinator brokerage score.

200

 

 

 

 

 

     
 

1 ETWORK CENTER C

/

 

Figure 11. Neighborhood-Level Network of Neighborhood C — Joint Planning
and Programming. Node size is proportional to coordinator brokerage score.

201

 

‘
-‘
l V ‘
.9 -
}

- ‘
‘ ~ 0 {lg ‘lva.“‘
lo 9: \‘ {’7/4 “" , \

““,///‘ 34\\\\ #9. r
«Nit-4.3-
\“V, K

  

 
         
 

Figure 12. Full Subnetwork for Network Center A (Neighborhood + Citywide) -
Information Sharing. Node size is proportional to gatekeeper/representative
brokerage score.

202

. Q 8‘. ..
s\"4

- -
'A‘IV‘I‘

   

J-

 
  
  

\

\«

V
‘v
.377
"

   
    

‘
,. ‘..
”Sis——

\lvtpii‘QO/' ' A
“/
, ~‘\\\

I_\

1
0,

    
  

II

Figure 13. Full Subnetwork for Network Center A (Neighborhood + Citywide) —
Joint Planning and Programming. Node size is proportional to
gatekeeper/representative brokerage score.

203

     

    

 
 
    

I-,:~4&":‘ ~ .
,- fé§\\!§r 0
1.9.5.9 \ ‘

0
Figure 14. Full Subnetwork for Network Center B (Neighborhood + Citywide) —
Information Sharing. Node size is proportional to g * ' ' . *-
brokerage score.

I' v|-" '

204

   
     
 
  

  
      

   
  

7 A“ //f‘\
'4 K \5" Z-jle/fll‘

S: \‘é ”
\ V \‘

     
  
    
     

 
       
     
  
 
 
  
 
  
 

 
  

“14's 0415.7"ng
o‘gfl‘VA‘bAv‘Q'". .7
-‘ t «v w“ *1 \., .1
my; 49A\‘§~11'l§‘.3§%\aul
1.73).“; $913.“ ‘1 5"‘\'\\'
,- 111%51-925?‘ <91?- '\ \
---9d\w «sis-98
’/ \ .';<—9 "'0' \‘~

.\\‘y¢\‘\\ WV

Figure 15. Full Subnetwork for Network Center B (Neighborhood + Citywide) —
Joint Planning and Programming. Node size is proportional to
- ‘ ' ---~'---- ‘ ".- brokerage score.

a r "'r"

    

 

 

 
       
       

   
 

             
      
   

205

  

fi \ \

'II-‘M.

  

Figure 16. Full Subnetwork for Network Center C (Neighborhood + Citywide) —
Information Sharing. Node size is proportional to a ‘ ' r-—-/.d,....._....t..t:.-
brokerage score.

206

    
 
 
  
    

   
      
 
       

' 1",1
\W ,

Alli!"
’/ VIII/(1’
y {iglik‘t’IlZ/lé%

\1 u
‘lll "‘;"91i':

] .
1‘:1="'"//.7//‘
MK

   
 

- "T7244

‘l

I:

”ll; .

  

s

  

Figure 17. Full Subnetwork for Network Center C (Neighborhood + Citywide)
- Joint Planning and Programming. Node size is proportional to

a r-Hllltupu---. ....- - DU .3 score.

207

APPENDIX N

NEIGHBORHOOD CENSUS DATA

208

Table 17. Economic status of Network Center primary service areas (per 2000
Census Data)

 

Network Center Percentage below 100% of Percentage below 200% of

Area poverty line ~ federal poverty line
(i.e., “low-income”)
A 20.5% 43.8%
B 26.7% 56.8%
C 32.5% 55.2%

 

 

209

 

APPENDIX 0

BASIC NETWORK MEASURES

210

Table 18. Basic Network Measures for Sub-Networks - Information Sharing Ties
by Network Center

 

 

 

Network ' Network Network
Center A Center B Center C
Density measures
(Whole Sub-Network)
Overall sub-network density .47 0.51 0.52
Block densities
Within city level 0.52 0.52 0.52
Within neighborhood 0.47 0.56 0.55
level
Between city and 0.31 0.40 0.45
neighborhood levels
Centrality measures
(Neighborhood level only)
Degree centralities for
sub-network
Min 2 2 3
Max 12 7 13
Mean 5.69 5.00 7.14
SD 3.29 1.79 3.14
Centralization 0.62 0.28 0.53
Network Center Scores
Degree
Raw 12 7 13
Normalized 100 77.78 100
Rank 1 1 1

 

211

Table 19. Basic Network Measures for Sub-Networks—Joint Planning &

Programming Ties by Network Center

 

Network I

 

 

Network Network
Center A Center B Center C
Density measures
(Whole Sub-Network)
Overall sub-network density .39 0.42 .43
Block densities
Within city level 0.42 0.42 0.42
Within neighborhood 0.49 0.51 0.46
level
Between city and 0.25 0.30 0.38
neighborhood levels
Centrality measures
(Neighborhood level only)
Degree centralities for
sub-network
Min 2 2 2
Max 12 7 13
Mean 5.846 4.6 6
SD 3.207 1.744 3.024
Centralization 0.61 0.33 0.63
Network Center Scores
Degree Centrality
Raw 12 7 8
Normalized 100 77.778 61 .538
Rank 1 1 4

212

APPENDIX P

Sample Qualitative Analysis Summary Tables

 

213

 

Table 20. Sample qualitative analysis summary table

: Network Center roles

 

 

 

NETWORK CENTER A NETWORK CENTER B NETWORK CENTER C
CONVENER ADVOCATE INTERFACE
' “Constantly“ bring in ° Advocate for ' Helped neighborhood
people from city neighborhood - important associations make
departments to meet w/ w/ agency cycles/attention connections w/ police
neighborhood span: "they do [program] department to step up
associations for ten weeks and then— patrol/surveillance

CONVENER - PROJECTS

' Food project—
neighborhood people,
local CSA, university
departments, other NPOs
- NC identifies potential
partners and engages,
facilitates, takes
responsibility

' Flood plain convening—
MSHDA, 3 neighborhood
orgs, developing
neighborhood orgs,
township, city planning,
umbrella -> master plan
for area—first time these
issues being discussed
across officials, and
pulling in neighborhood
leaders

CONVENER: TASK FORCES

' Task force examples—
suicide prevention 15
orgs; housing task force;
child safety—dozen inst
partners

' Form task forces, convene
partners and develop
neighborhood-specific
strategies

 

they're gone...my job is to
keep reminding... 'come
back here'"

Advocate at the table;
Mayor's task force ”they
had heard about how fast
we had taken the housing
task force...they asked me
to come...and because we
went there, they chose our
neighborhood as a place
that they want to invest
into"

BRIDGE

Spreading information—
e.g., meet with MSU
Extension. then go to
neighborhood
organizations "repeating
what they‘ve offered, what
they can do"/
representing
neighborhood at Mayor's
Task Force—they are
interested in the
neighborhood now.

 

l & R intermediary—
agencies call to ask if
network center can
connect clients w/ services

Describes work w/
organizations
separately...not bringing
together in combination
highlights information
sharing, does not discuss
planning/programming
Focus on direct links—
outside contacts and org
programs to community
members—e.g., brought in
service learning center to
do chores for 2 seniors,
brings back info from city
committee to share w/

people in community

 

214

 

Table 21. Sample qualitative analysis summary table: “what makes it work"

 

 

 

 

THEMES [A r8 . TC
CREDIBILITY/T RUST
From 9'0“”d'"9 ' intense face-to-face ' '...a lot of these city ' NA
rn _ work w/ people know that I
neighborhood, neighborhood -—> live in the community
strong local credibility in that I work in...l've
knowledge 0f discussing needs and faced many of the
Issues recommending challenges that the
approaches and other people face, I
programs (e.g., w/ guess I‘ve gained the
health dept)l respect of those
credibility over time— people because they
success -> know I've lived it...“
increased org
commitment
' outreach and data
collection —>
credibility/Influence
w/ inst.
From SUOCGSS ' Successful projects - ' focused, concrete ' NA
> credibility w/ collaboration—brings
institutions '...given partners together to

what we've been able execute and it gets
to do here, we can go done -> credibility

about revising...“ 'eamed about

' organic development: success/speed of
”start small...promise development of
certain things...do project. . .took interest

those...|et them
know...take to next
step...'trust me a little

 

 

further"
'"Sj- expenence ' “they, in the first ' change over time— ' NA
W/ “1‘19."ert couple of years, now can ask orgsfinst
[capacity 0f learned to trust our for resources,
network center judgment and our programs based on

capacity to deliver” assessment of need

RESULT ->leverage ”they had to Ieam to

to craft strategies trust me...trust my

judgment“

 

 

 

 

215

 

Table 21, continued.

 

 

 

 

 

 

 

 

THEMES A B C
3““ sides/both ' "we need to maintain ' "you've got to build ' NA
directions credibility in both trust on both
(neighborhood directions'lfu nder, sides'ltrust with
and CW) inst partners/"people neighborhood and
in this neighborhood ”be the person that is
need to know they going to go to the
can trust us...trust our other side (city)
effectiveness...trust
our commitment to
deliver“
' "walk that fine
line...work on both
fronts simultaneously
to deepen the
relationship'lear for
both
GROUNDEDNESS IN NEIGHBORHOOD
Connection to ' connection w/ ' dir A comment on B: ' N/A
place neighborhood “gives 'They're
us Iife...gives us rooted...strong
passion and connection to place
commitment and there...and I think
interest”/"we're so that's why it will
place-based here that survive”
we couldn't function e Rootedness ..>
anywhere else and credibility w/
Still serve this neighborhood
nelghbomood' residents "...I didn’t
have to move back
here. I could have
moved anywhere. But
I don't ever want
people to think, ‘Oh
[she] thinks she’s
something'...‘ doesn‘t
put self on level
above res.
Posrtioned for o o integrative view— 0 N IA

integrative view

 

 

sees opportunities to
collaborate where
partner orgs don't
have awareness

Grounded in
neighborhood—e.g.,
parolee issue—has
sense of whole
context, why parolees
returning to
neighborhood, etc.

 

 

216

 

Table 21, continued.

 

THEMES

[A

is

lo

 

PROCESS-PRODUCT SYNERGY

 

 

Good facility -)
strong
programmatic
base -) fuels
work

 

programmatic
“bottom" to
planning/developmen
t activities: "gives us
gravity“ "substance" -
-> product to match
process
facility—didn't
originally want to
become a center, but
realized needed a
‘place" in order to
grow—"housing staff,
holding meetings,
bringing people
together for forums
and activities...it
needs a space
together...to move
out from, come back
to...."

neighborhood center
creates basis for
partnership work—
”because we're
staffed and funded to
do it—to create the
listening
structures...create
opportunities to come
together...craft the
strategies”

Facility—means to
"wins” and tangibles
for community—“I
don’t know how much
longer we could have
kept inviting people to
meetings to come up
with ideas that would
change the
neighborhood without
having a place to
house those ideas."
”It would have been
an idea...forever. And
I think people get
tired of that.” - ”...do
something that
people can touch..."

 

Facility -) hub of
neighborhood life,
very high traffic, easy
to reach residents

 

Weak facility -)
weaker participation
in programs

Lack of foot traffic
tied to natural
barriers and
facility/surrounded by
fence/difficult to get
there/institutional

 

217

 

Table 21, continued

 

 

 

THEMES A B
Summit-Center ' deliberate/intentional ' Minor integration - CD: rift problematic:
Integration joining of summit/ summit not strong at ”you can't duplicate

 

center/ neighborhood
infrastructure:
”...because we see
what happens when
there are battles b/w
those entities in
different parts of
town"

 

time of data collection

 

effort—you can't ask
people to come to the
table twice to talk
about the same
issue“

CD: other NCs have
pieces of summit
work—C "left out in
the cold—no
communication”

Lack of summit-
center integration
problematic—pits two
processes against
each other. CD: ”it
makes for
competition for
money, for resources,
for people...“ summit
I'pulls on" dir C but
doesn't give back -
JP is summit's
strength

NC-A commentary on
summit/center rift in
C: '...they both
demonstrate what
happens when you
separate process and
product” - summit
assessment/planning,
center hub for
services.

 

218

 

 

 

APPENDIX Q

Qualitative Analysis Of Change Data

219

Table 22. Core themes in content analysis of open-ended survey items.

Core Themes By Meta-theme And Number of Examples
Valence Of Relationship Change Mentions

 

 

METATHEME: Exposure/Availability

 

Newl Stronger Board/ Committee 10 Participation on a board leading to
Membership increased contact and collaboration
Common Participation! 4 Exposure through common participation in
Project Involvement a program or project
Proximity 4 Co-Iocation of services, housing programs

in an organization's facility; Working with
nearby organizations

 

 

Increased Contact/ 3 Increased communication/responsiveness
Communication of other org; More frequent meetings,
better contact
New Programming 3 Upcoming programming; New
programs/contracts
Weaker/ Proximity 7 Relocation away from close proximity; No
Defunct longer doing programs at a particular
location
Common Participation/ 6 No longer part of a project; Specific project
Project Involvement ended or didn’t work out
METATHEME: Institutional/Resource Factors
Newl Stronger Funding Relationship 5 Took on fiduciary role for org; Funded or

received funding from org; Got funding for
new shared initiative, collaborative grants

Funding Changes 2 New grant/funding - allows $ to
collaborate, makes projects possible

Institutional Pressure/ 2 Initiative of parent (higher-level) institution;

Initiative Increased focus from high-level

collaborative body
Mandated 2 Grant/funding mandates collaboration;
Collaboration Statutory relationships

Weaker/ Funding Changes 9 Grant/funding for project(s) ran out; Orgs
Defunct partnered during grant but left aftenrvards
Contracting 3 Changes in how services are contracted;

Use another org/agency

220

 

Table 22, Continued

 

METATHEME: Other Change Reasons

New/ Stronger History/Project 5 Had good project/collaboration
Experience experiences, led to stronger relationship;
Gained track record, led to more projects;
Relationships maturing over course of
many projects

 

Weaker/ Conflict! 2 Confrontation with org ended relationship;
Defunct Misunderstanding Attitudes of participants - org has
misunderstanding
Decay Of Weak 2 Relationship wasn’t strong/substantial —
Relationship faded/dropped off
Fear of Collaboration 1 Finances/turf - fear successful partnership

will lead to merger/consolidation and
resource loss
METATHEME: ORGANIZATIONAL CAPACITY TO COLLABORATE

Newl Stronger Funding Changes 2 Org has become more organized and
funded
New funded staff role helps to build
relationships

 

 

 

 

 

Staff Resources 1 Have new staff — expansion of issues/work
Weaker/ Funding Changes 3 Current budget down to nothing
Defunct Not enough $ to continue program
Other org retrenching b/c of funding
Lack Of Resources— 4 People don't have time for you anymore
Time Too busy
Program too much work
Org Decline 3 Other org has become weak, inactive (nb:
all three referenced same org)
Staff Resources 3 Not enough staff to do collaborative
program
Change of staffing availability
METATHEME: Relationships embedded in individuals
New/ Stronger Personal/Outside 3 Staff/admin brought old connections to
Relationships new org position; Personal connections
carried over to org connections
Staff/Management 2 New staff/admin member brought new
Changes focus on neighborhood; New leader more
community-focused
Weaker/ Staff/Management 13 New management/leadership at other org
Defunct Changes - changed focus or not interested; Hard to

keep up with other org staff changes;
Director/staff turnover - lost relationship

221

Table 22, Continued
METATHEME: STRATEGIC REASONS

 

New! Stronger Common 6 Other org developed aspect that coincides
Focus/Interests! Goals with interests/work; ID have same
issues/goals
Proactive-Strategy 6 Org developing work in issue area prompts
Driven seeking out collaboration; Strategic

direction suggests new partnerships to try;
New opportunities surfaced
Change Of Needs 3 Response to new program priorities to
match ID'd needs; Org facing trends that
require specific new connections

Weaker! Common 9 Change of org focus/goaIs/agenda — loss
Defunct Focus/Interests! Goals of common interest; Other org prioritizing
different project; Focus not lining up; Less
in common
Change Of Needs 4 Need of community has changed —-less

impetus to work w/ specific orgs; Changed
outreach method that wasn’t working

Proactive-Strategy 3 Change of basic org strategies/approach
Driven
Incompatibility of 2 Org population focus no longer matches
Approach need; Other org too top—down
Pick/Choose 2 Enough on plate — pick/choose most

valuable opportunities
METATHEME: INTEREST INNALUING OF COLLABORATION

New! Stronger Org Initiated! Outreach 11 Active outreach and relationship building
by org; More effort to be involved; Drive for
exposure/credibility; Made partnering a
priority - working across dimensions,
reaching more

 

Other Org Initiated 4 Other org initiated - contacted w/ ideas;
We've now been invited
Other Org Interest in 4 Other orgs recognizing value of collab
Collaboration! more; Other org changing orientation to be
Partnerships more helpful, partnership oriented; Having

someone @ other org who wants to work
w/ us; Other org has increased cooperation

w! neighborhood
Weaker! Org Initiated/Outreach 1 Just “haven't kept the ball up" w! reaching
Defunct out
Other Org Interest in 4 Other org not working in community/not at
Collaboration! table; Other org pays lip service to collab
Partnerships but it doesn't play out; We approached
them but unsuccessful in engaging their
afienfion

222

Table 22, continued

 

METATHEME: NETWORK CENTER LINKS

 

New! Stronger Project Participation 3
Board Membership 2
Leadership of NC 2
director
Opening of Center 1

—5

Recovery from early
organizational turmoil

Agency availability at 1

Network Center
Weaker! Board Membership 1

Defunct
Opening of Center 1
Early organizational 1
turmoil

Agency availability at 1

Network Center

223

Formed through NC project [ref to NCs:
A,B,D
Sitting on NC board leading to
relationships [ref to NC: B (x2)]
Directly cite leadership; Dir identifies need,
common goals — creativity, commitment
[ref to NCs: B,D]
[ref to NC: A]

[ref to NC: C]
[ref to NC: C]

Had relationship when both on NC board
but not now [ref to NC: C]
Working more with center instead of
directly wl neighborhood org [ref to NC: B]

early problems w! board led to staff
turnover. important orgs leaving board [ref
to NC: C]

Orgs no longer coming into Center, don‘t
connect with as much [ref to NC: B]

REFERENCES

Ackoff, R. R. (1974). Redesigning the future. New York: Wiley.

Aiken, M., & Hage, J. (1968). Organizational interdependence and
intraorganizational structure. American Sociological Review, 35, 650-665.

Aldrich, H., & Herker, D. (1977). Boundary spanning roles and organization
structure. The Academy of Management Review, 2, 217-230.

Alter, C., & Hage, J. (1993). Organizations wanking together. Thousand Oaks,
CA: SAGE.

Aspen Roundtable. (1998). Voices from the field: Learning from the early work of
comprehensive community initiatives. Washington, DC: The Aspen
Institute.

Bailey, 0., & Koney, K. M. (1996). Interorganizational community-based
collaboratives: A strategic response to shape the social work agenda.
Social Work, 41(6), 602-611.

Bartunek, J. M., & Louis, M. R. (1996). Insider/Outsider Team Research: SAGE
Publications.

Bazzoli, G. J., Casey, E., Alexander, J. A., Conrad, D. A., Shortell, S. M., Sofaer,
S., et al. (2003). Collaborative initiatives: Where the rubber meets the road
in community partnerships. Medical Care Research and Review, 60(4),
638-948.

Bitter, G. B. (1977). From conflict to coalition: Disability groups, parents,
professionals. Volta Review, 79(5), 317-326.

Blank, M. J., Brand, B., Deich, S., Kazis, R., Politz, B., & Trippe, S. (Undated
report). Local intermediary organizations: Connecting the dots for
children, youth, and families.

Bolino, M. C., Bloodgood, J. M., & Turnley, W. H. (2002). Organizational
citizenship behavior and the creation of social capital in organizations.
Academy of Management Review, 27(4), 505-522.

Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). Ucinet for Vlfindows:
Software for Social Network Analysis (Version 6). Harvard, MA: Analytic
Technologies.

Briggs, X. (I. S. (1997). Social Capital and the Cities: Advice to Change Agents.
National Civic Review, 86(2), 11 1-116.

224

Briggs, X. d. S. (1998). Brown kids in white suburbs: Housing mobility and the
many faces of social capital. Housing Policy Debate, 9(1).

Bruner, C. (2000). Social services system reform and poor neighborhoods: What
we know and what we need to find out. Des Moines, IA: National Center
for Service Integration. '

Burt, R. S. (1992). Structural holes: The social structure of competition.
Cambridge: Harvard University Press.

Burt, R. S. (2000a). The network structure of social capital. Research in
Organizational Behavior, 22, 345-423.

Burt, R. S. (2000b). Structural Holes versus Network Closure as Social Capital.
In N. Lin, K. Cook & R. S. Burt (Eds), Social capital: Theory and research.
Chicago, IL: Aldine.

Butterfoss, F. D., Goodman, R. M., & Wandersman, A. (1993). Community
coalitions for prevention and health promotion: Factors predicting

satisfaction, participation, and planning. Health Education Research, 8(3),
315-330.

Center for the Study of Social Policy. (1996). Systems Change At The
Neighborhood Level. Retrieved 12l1, 2000, from
http:/Mww.cssp.orglkd09.htm

Chaskin, R. J. (2001). Organizational infrastructure and community capacity: The
role of broker organizations. In S. W. Hartwell & R. K. Schutt (Eds), The
Organizational Response to Social Problems (pp. 143-166). Oxford:
Elsevier Science.

Chaskin, R. J. (2003). Fostering neighborhood democracy: Legitimacy and
accountability within loosely coupled systems. Nonprofit and Voluntary
Sector Quarterly, 32(2), 161-189.

Chaskin, R. J., & Abunimah, A. (1999). A view from the city: Local government
perspectives on neighborhood-based governance in community-building
initiatives. Journal of Urban Affairs, 21 (1 ), 57-78.

Chaskin, R. J., Brown, R, Venkatesh, S., & Vidal, A. (2001). Building Community
Capacity. New York: Walter de Gruyter.

Chaskin, R. J., & Peters, C. (2000). Decision Making and Action at the
Neighborhood Level: An Exploration of Mechanisms and Processes.
Chicago: Chapin Hall Center for Children.

225

Chavis, D. M. (1995). Building community capacity to prevent violence through
coalitions and partnerships. Journal of Health Care for the Poor and
Underserved, 6(2), 234-245.

Chavis, D. M., Florin, P., & Felix, M. R. J. (1993). Nurturing grassroots initiatives
for community development: The role of enabling systems. In T. Mizrahi &
J. Morrison (Eds), Community organization and social administration:
Advances, trends, and emerging principles. Binghamton, NY: The
Haworth Press.

Chinman, M. J., Anderson, C. M., Imm, P. S., Wandersman, A., & et al. (1996).
The perceptions of costs and benefits of high active versus low active
groups in community coalitions at different stages in coalition
development. Journal of Community Psychology, 24(3), 263-274.

Coleman, J. S. (1990). Foundations of social theory. Cambridge, MA: Belknap
Press of Harvard University Press.

Committee for Economic Development. (1995). Rebuilding Inner-City
Communities: A New Approach to the Nation '3 Urban Crisis. New York:
Committee for Economic Development.

Cordero-Guzman, H. R. (2001). Interorganizational networks among community-
based organizations. Unpublished manuscript.

Cornerstone Consulting Group (2002). End games: The challenge of
sustainability. Annie E. Casey Foundation. '

Crisp, B. R., Swerissen, H., & Duckett, S. J. (2000). Four approaches to capacity
building in health: Consequences for measurement and accountability.
Health Promotion International, 15(2), 99-107.

Davis, J. A., & Leinhardt, S. (1972). The Structure of Positive Interpersonal
Relations in Small Groups. In J. Berger (Ed.), Sociological Theories in
Progress (Vol. 2, pp. 218-251). Boston: Houghton Miffiin.

Dryfoos, J. G. (1994). Full service schools: Revolution in health and social
services to families. San Francisco: Jossey-Bass.

Dunst, C. J., & Trivette, C. M. (2001). Benefits associated with family resource
center practices. Asheville, NC: Winterberry Press.

Emirbayer, M. & Goodwin, J. (1994). Network analysis, culture, and the problems
of agency. American Joumal of Sociology. 99, 1411-54.

Ferguson, R. F., & Dickens, W. T. (1999). Urban problems and community
development. Washington DC: Brookings Institution Press.

226

Fernandez, R. & Gould, R. (1994). A dilemma of state power in the national
health policy domain. American Journal of Sociology. 99, 1455 - 1491.

Finn, J. L., & Checkoway, B. (1998). Young people as competent community
builders: A challenge to social work. Social Work, 43(4), 335-345.

Foster-Fishman, P. G., Salem, D. A., Allen, N. A., & Fahrbach, K. (2001).
Facilitating interorganizational collaboration: The contributions of

interorganizational alliances. American Jouma! of Community Psychology,
29(6), 875-905.

Frank, K. A. (1995). Identifying Cohesive Subgroups. Social Networks, 17, 27-56.

Frank, K. A. (1996). Mapping interactions within and between cohesive
subgroups. Social Networks, 18, 93-1 19.

Frank, K. A., & Yasumoto, J. (1996). Embedding subgroups in a sociogram:
Linking theory and image. Connections, 19(1), 43-57.

Fried, B. J., Johnsen, M. C., Starrett, B. E., Calloway, M. O., & Morrissey, J. P.
(1998). An empirical assessment of rural community support networks for

individuals with severe mental disorders. Community Mental Health
Journal, 34(1), 39-56.

Fukuyama, F. (1995). Social capital and the global economy. Foreign Affairs,
74(5), 89-103.

Gabbay, S. M., 81 Leenders, R. T. A. J. (2002). Social capital of organizations:
from social structure to the management of corporate social capital
(Working Paper). Groningen: University of Groningen, Research Institute
SOM (Systems, Organisations and Management).

Gilchrist, A. (2000). The Well-Connected Community: Networking to the 'Edge of
Chaos'. Community Development Journal, 35(3), 264-275.

Gittell, R. J., & Vidal, A. (1998). Community organizing: Building social capital as
a development strategy. Thousand Oaks, CA: SAGE Publications, Inc.

Goodman, R. M., Speers, M. A., McLeroy, K., Fawcett, S. B., Kegler, M., Parker,
E., et al. (1998). Identifying and defining the dimensions of community
capacity to provide a basis for measurement. Health Education and
Behavior, 25(3), 258-278.

Goodman, R. M., Steckler, A., Hoover, S., & Schwartz, R. (1993). A critique of
contemporary community health promotion approaches: Maine - a multiple
case study. American Jouma! of Health Promotion, 7(3), 208-220.

227

Gould, R. V., & Fernandez, R. M. (1989). Structures of mediation: A formal
approach to brokerage in transaction networks. Sociological Methodology,
19, 89-126.

Granovetter, M. (1973). The strength of weak ties hypothesis. American Jouma!
of Sociology, 78(6), 1360-1380. -

Gray, B. (1985). Conditions facilitating interorganizational collaboration. Human
Relations, 38(10), 91 1-936.

Gulati, R. (1995). Social Structure and Alliance Formation Patterns: A
Longitudinal Analysis. Administrative Science Quarterly, 40, 619-652.

Hendrickson, J. M., 8. Omer, D. (1995). School-based comprehensive services.
In P. Adams & K. Nelson (Eds), Reinventing human services:
Community- and family-centered practice (pp. 145-162). New York: Aldine
de Gruyter.

Himmelman, A. T. (1996). On the theory and practice of transformational
collaboration: From social service to social justice. In C. Huxham (Ed.),
Creating collaborative advantage (pp. 19-43). Thousand Oaks, CA: SA GE.

Holcomb, P. A., Seefeldt, K. S., Trutko, J., Barnow, B. S., & Nightingale, D. S.
(1993). One stop shopping service integration: Major dimensions, key
characteristics and impediments to implementation. Washington, DC: The
Urban Institute.

Holland, P. W., & Leinhardt, S. (1970). A method for detecting structure in
sociometric data. American Joumal of Sociology, 70, 492-513.

Holland, P. W., & Leinhardt, S. (1978). An omnibus test for social structure using
triads. Sociological Methods and Research, 7, 227-256.

Kagan, S. L., & Neville, P. (1994). Integrating services for children and families:
Understanding the past to shape the future. New Haven, CT: Yale
University Press.

Katz, B. (1998). Give community institutions a fighting chance. In E. J. Dionne Jr.
(Ed.), Community Works: The Revival of Civil Society in America (pp. 64-
71). Washington, DC: Brookings Institution Press.

Kegler, M. C. (1998). Factors that contribute to effective community health
promotion coalitions: A study of 10 Project ASSIST coalitions in North
Carolina. Health Education and Behavior, 25(3), 338-353.

Knoke, D. (1990). Political networks: The stmctural perspective. New York:
Cambridge University Press.

228

Lappé, F. M., & Du Bois, P. M. (1997). Building social capital without looking
backward. National Civic Review, 86(2), 119-128.

Laumann, E. O., Galaskiewicz, J., & Marsden, P. V. (1978). Community structure
as interorganizational linkages. Ann. Rev. Social, 4, 455-484.

Lawless, M. W., & Moore, R. A. (1989). Interorganizaitonal systems in public
service delivery: A new application of the dynamic network framework.
Human Relations, 42( 12), 1167-1184.

Lin, N. (1999). Building a network theory of social capital. CONNECTIONS,
22(1), 28-51.

Liou, Y. T., & Stroh, R. C. (1998). Community development intermediary systems
in the United States: Origins, evolution, and functions. Housing Policy
Debate, 9(3), 575-594.

Maloney, W., Smith, G., & Stoker, G. (2000). Social Capital and Urban
Governance: Adding a More Contextualized 'Top-down' Perspective.
Political Studies, 48, 802-820.

Mandell, M. P. (1984). Application of network analysis to the implementation of a
complex project. Human Relations, 37(6), 659-679.

Marsden, P. V. (1982) Brokerage behavior in restricted exchange networks. In
P. V. Marsden and N. Lin (eds. ) Social Structure and Network Analysis.
Beverly Hills: Sage Publications.

Mattessich, P. W., & Monsey, B. R. (1992). Collaboration: What makes it work.
St. Paul, MN: Amherst I-l. Wilder Foundation.

Mayer, S. E. (1994). Building community capacity: The potential of community
foundations. Minneapolis, MN: Rainbow Research, Inc.

McKnight, J., & Kretzmann, J. (1984). Community Organization in the 80's:
Toward a Post-Alinsky Agenda. Social Policy, 14(3), 15-17.

McLeroy, K. R., Kegler, M., Steckler, A., Burdine, J. M., & Wisotzky, M. (1994).
Community coalitions for health promotion: Summary and further
reflections. Health Education Research, 9(1), 1-11.

Miles, M. B., & Hubennan, A. M. (1994). Qualitative data analysis: An expanded
sourcebook (2nd ed.). Thousand Oaks, CA: SAGE Publications.

Milward, H. B. (1994). Nonprofit contracting and the hollow state. Public
Administration Review, 54(1), 73-77.

229

Milward, H. B., & Provan, K. G. (2000). Governing the hollow state. Joumal of
Public Administration Research and Theory, 10, 359-379.

Mirabella, R. M. (2001). A symposium introduction: Filling the hollow state. Public
Performance and Management Review, 25(1).

Morrison, J. 0., Howard, J., Johnson, C., Navarro, F. J., Plachetka, B., & Bell, T.
(1997). Strengthening neighborhoods by developing community networks.
Social Work, 42(5), 527-534.

Morrissey, J. P. (1992). An Interorganizational Network Approach to Evaluating
Children's Mental Health Service Systems. New Directions For Program
Evaluation(54), 85-95.

Morrissey, J. P., Johnsen, M. C., & Calloway, M. O. (1997). Evaluating
Performance and Change in Mental Health Systems Serving Children and
Youth: An Interorganizational Network Approach. The Journal of Mental
Health Administration, 24(1), 4-22.

Morrissey, J. P., Johnsen, M. C., Calloway, M. O., 8. Ullman, M. (1996). Service
System Performance and Integration: A Baseline Profile of the ACCESS
Demonstration Sites (pp. 16).

Murray, M. (2000). Social capital formation and healthy communities: Insights
from the Colorado Healthy Communities Initiative. Community
Development Journal, 35(2), 99-108.

Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital, and the
organizational advantage. Academy of Management Review, 23(2), 242-
266.

Narayan, D. (1999). Bonds and Bridges: Social Capital And Poverty. World
Bank.

Oldenburg, R. (1989). The great good place: Cafes, cofi'ee shops, community
centers, beauty pariors, general stores, bars, hangouts, and how they get
you through the day. New York: Paragon House.

Oliver, C. (1990). Determinants of Interorganizational Relationships: Integration
and Future Directions. Academy of Management Review, 15(2), 241-265.

Perkins, D. D., Hughey, J., & Speer, P. W. (2003). Community psychology
perspectives on social capital theory and community development
practice. Journal of the Community Development Society, 33(1), 33-52.

Plant, R. W., & King, P. A. (1995). The family resource center: A community-
based system of family support services. In B. Ryan, G. Adams, T.

230

Gullotta, R. Weissberg & R. Hampton (Eds), The family-school
connection: Theory, research, and practice. Thousand Oaks, CA: Sage.

Portes, A. (1998). Social capital: Its origins and applications in modern sociology.
Annual Review of Sociology, 22, 1-24.

Provan, K. G., & Milward, H. B. (1995). A preliminary theory of interorganizational
network effectiveness: A comparative study of four community mental
health systems. Administrative Science Quarterly, 40(1), 1-33.

Provan, K. G., & Milward, H. B. (2001). Do Networks Really Work? A Framework
for Evaluating Public-Sector Organizational Networks. Public
Administration Review, 61(4), 414-423.

Provan, K. G., Veazie, M. A., Teufel-Shone, N. l., & Huddleston, C. (2004).
Network analysis as a tool for assessing and building community capacity
for provision of chronic disease services. Health Promotion Practice, 5(2),
174-181.

Putnam, R. D. (1999). The Prosperous Community: Social Capital and Public
Life. The American Prospect(13), 9.

Putnam, R., Leonardi, R., & Nanetti, R. (1993). Making democracy work: Civic
traditions in modem Italy. Princeton: Princeton University Press.

Rivard, J. C., Johnsen, M. C., Morrissey, J. P., & Starrett, B. E. (1999). The
dynamics of interagency collaboration: How linkages develop for child
welfare and juvenile justice sectors in a system of care demonstration.
Journal of Social Service Research, 25(3), 61-82.

Roemer, R., Kramer, C., & F rink, J. (1975). Planning Urban Health Sen/ices:
From Jungle to System. New York: Springer.

Romualdi, V., & Sandoval, J. (1997). Community-Based Service Integration:
Family Resource Center Initiatives. In R. J. lllback, C. T. Cobb & H. M.
Joseph (Eds), Integrated Services for Children and Families:
Opportunities for Psychological Practice (pp. 53-73). Washington, DC:
American Psychological Association.

Rosenheck, R. A., Lam, J., Morrissey, J. P., Calloway, M. O., Stolar, M.,
Randolph, F., et al. (2002). Service systems integration and outcomes for

mentally ill homeless persons in the ACCESS program. Psychiatric
Services, 53(8), 958-966.

Rubin, H. J., & Rubin, I. (1986). Community Organizing and Development.
Columbus, Ohio: Merrill Publishing Company.

231

Smith, M, Littlejohns, L. B., & Thompson, D. (2001). Shaking out the cobwebs:
Insights into community capacity and its relation to health outcomes.
Community Development Joumal, 36(1), 3041.

Smith, S. R., & Lipsky, M. (1993). Nonprofits for hire: The welfare state in the age
of contracting. Cambridge, MA: Harvard University Press.

Stoecker Randy. 2004. The mystery of the missing social capital and the ghost
of social structure: Why community development can't win. In R.
Silverman (ed.) Community-Based Organizations in Contemporary Urban
Society: The Intersection of Social Capital and Local Context. Detroit:
Wayne State University Press.

Stone, R. (1994). Comprehensive community-building strategies: Issues and
opportunities for Ieaming. Chicago: Chapin Hall Center for Children.

Turk, H. (1973). Comparative urban structure from an interorganizational
perspective. Administrative Science Quarterly, 18, 37-55.

Turk, H. (1977). Organizations in Modern Life. San Francisco: Jossey-Bass.

Van de Ven, A. H., & Walker, G. (1979). Coordination patterns within an
interorganizational network. Human Relations, 32(1), 19-36.

Vicary, J. R., Doebler, M. K., Bridger, J. C., Gurgevich, E. A., & Deike, R. C.
(1996). A community systems approach to substance abuse prevention in
a rural setting. The Journal of Primary Prevention, 16, 303-318.

Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and
applications. Cambridge, UK: Cambridge University Press.

Wynn, J. R. (2000). The role of local intermediary organizations in the youth
development field: Edna McConnell Clark Foundation.

232

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

IljlllljjjlljjjjjlIjjjjl