What Is Our Purpose Here? Network Relationships and Goal Congruence in a Goal-Directed Network

Abstract

“Whole” goal-directed networks are inherently full of conflict and tensions, but a certain level of agreement about network-level goals is important for goal-directed service delivery networks. This work examines goal congruency in a large, heterogeneous child and youth health and well-being network. By drawing on the dimensions of social capital to categorize types of network relationships, the tie portfolios that are associated with perceived goal congruence between network and organizational member goals are examined. Using crisp-set qualitative comparative analysis (cs-QCA), the results indicate that the perception of goal congruence is present when organizations have a high number of linking ties or a strong tie to network management staff combined with a high number of bridging ties. Implications are discussed for network governance and management, focusing especially on the types of relationships among network members that may need to be facilitated to support goal congruence, and thus, effective network functioning.

Full text

https://doi.org/10.1177/0275074019879506
American Review of Public Administration
1 –17
© The Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0275074019879506
journals.sagepub.com/home/arp
Article
As research on interorganizational networks has proliferated
in public administration over the past few decades, O’Toole
(2015) notes that progress has been made with regard to pub-
lic administration scholars “treating networks seriously” but
that more research and theory is necessary to inform prac-
tice. One such area is the considerable literature on the man-
agement practices involved in supporting network action.
O’Toole acknowledges not only the accumulated evidence
regarding the need for managers to develop relationships, but
the shortage of research examining structures and processes
as well. Network structure is often conceptualized as the
relationships among actors (Provan & Lemaire, 2012), but
which relationships and thus, what overall structure is neces-
sary to support networks is where more research is essential
to inform network management practice.
The basis of the network approach is that relationships are
the levers used to get things done; therefore, recognizing
which levers are linked to which outcomes is important to
understanding how to make networks work. This is espe-
cially important for the goal-directed network context, which
are intentionally formed entities comprised of three or more
independent agencies working together to achieve a collec-
tive goal (Provan & Kenis, 2008). These goal-directed net-
works are typically enacted to address a specific “wicked”
problem (O’Toole, 2015; Provan, Fish, & Sydow, 2007), a
problem which necessitates collective action among multiple
players across messy boundaries of action (Kettl, 2008).
However, despite the nomenclature, when a goal-directed
network pulls together multiple sectors, governmental per-
spectives, and covers a vast geographical area, reaching a
minimum level of consensus about the goals of the network
may prove difficult, if not impossible (Vangen & Huxham,
2012).
In such large, heterogeneous networks, agreeing about the
broader goals of the network as a whole, when numerous
organizations across different service domains and across a
large geographical area are involved, would likely be difficult
even in a setting with low levels of conflict or competition.
This consensus or agreement may not be the key issue, how-
ever, since agreement among network members alone does
not ensure contribution to the network because of the volun-
tary demands of networks, whether the network is mandated
or emergent (self-reference). In that case, how then does a
large, heterogeneous network attain member commitment to
network goals, and thus willingness to contribute?
Even in the case of a network where purpose is formu-
lated and defined early on, the need to constantly ensure
commitment or buy-in to that purpose is part of network
functioning. Goal consensus is the extent of agreement
879506ARPXXX10.1177/0275074019879506The American Review of Public AdministrationLemaire
research-article2019
1Virginia Tech, Blacksburg, USA
Corresponding Author:
Robin H. Lemaire, Center for Public Administration & Policy, School of
Public and International Affairs, Virginia Tech, 104 Draper Road, SW
(0520), Blacksburg, VA 24061, USA.
Email: rlemaire@vt.edu
What Is Our Purpose Here? Network
Relationships and Goal Congruence
in a Goal-Directed Network
Robin H. Lemaire1
Abstract
“Whole” goal-directed networks are inherently full of conflict and tensions, but a certain level of agreement about network-
level goals is important for goal-directed service delivery networks. This work examines goal congruency in a large,
heterogeneous child and youth health and well-being network. By drawing on the dimensions of social capital to categorize
types of network relationships, the tie portfolios that are associated with perceived goal congruence between network and
organizational member goals are examined. Using crisp-set qualitative comparative analysis (cs-QCA), the results indicate
that the perception of goal congruence is present when organizations have a high number of linking ties or a strong tie to
network management staff combined with a high number of bridging ties. Implications are discussed for network governance
and management, focusing especially on the types of relationships among network members that may need to be facilitated
to support goal congruence, and thus, effective network functioning.
Keywords
networks, goal congruence, social network analysis, qualitative comparative analysis (QCA)

2 American Review of Public Administration 00(0)
among network members about the goals of the network,
which is especially an issue with governance and defining
network purpose and goals. Once that direction is set, the
greater need in regard to enhancing commitment is in con-
vincing other network members that the goals of the network
align with their own mandates (Provan & Kenis, 2008). Goal
congruence is important in justifying the participation of
individual organizations in a network; therefore, the process
of constantly ensuring network-wide commitment is about
convincing all network members that the goals of the net-
work are congruent with their own mandates.
Therefore, this analysis will be an examination of the
structure of network relationships that are associated with
congruence of organizational and network purpose and the
role of the governance structure in that association. A mature,
formal goal-directed network, the Southern Alberta Child
and Youth Health Network (SACYHN), will be examined.
SACYHN was an example of a large, heterogeneous whole
network since it defined health broadly and included any ser-
vice sector related to child and youth health and well-being
and covered the vast geographical area of the southern por-
tion of the Alberta Province. Drawing on forms of social
capital, crisp-set qualitative comparative analysis (cs-QCA)
will be used to examine the different types of relationship
involvement across SACYHN and which portfolios of rela-
tionships are associated with congruence of network and
organizational goals.
Goal Congruency
The idea that a network can have a purpose in and of itself
that goes beyond the purpose of serving the needs of member
organizations is one of the fundamental ideas behind the
whole network approach and the study of goal-directed net-
works (Provan et al., 2007). In general, it is thought that
when there is goal consensus, network members will be more
committed to the network and more likely to work with one
another (Provan & Kenis, 2008). The argument proposed in
this article, however, is that enhancing commitment to net-
work goals is not a matter of goal consensus, but rather, goal
congruence.
Once the direction or goal of the network is set, either by
all or a subset of members, the greater need in regard to
enhancing commitment is in convincing all network mem-
bers that the goals of the network align with their own man-
dates. In the context of most goal-directed service delivery
networks, the network is dependent on the member organiza-
tions for the work of the network. In public sector networks,
these member organizations are constrained by their own
mandates. Therefore, carrying out the work of the network
by member organizations is harder to justify if that network
work is not perceived to align with the mandates of those
individual member organizations (McGuire & Agranoff,
2011). This is the reason why goal congruency among net-
work member goals and network goals is as important to
consider as goal consensus at the network level.
Goal congruence, at least in regard to the implementation
of policy reforms, has been defined as “the extent of agree-
ment between the official or formal policy goals of political
officials and the operative goals of the organizations or net-
works charged with delivering that policy” (Meyers,
Riccucci, & Lurie, 2001, p. 604). But even at the operative
level, there is a distinction between the formal goals as
defined by the collective, the network, and the goals of the
individual organizational members. Thus, here I define goal
congruency as the extent of agreement between the formal
network goals and the operative goals of the individual orga-
nizations on which the network is reliant.
Enhancing network-wide commitment, one of the central
purposes of network management (Milward & Provan,
2006), therefore, is arguably about enhancing the percep-
tions of congruence of purpose of individual network mem-
bers and with the purpose of the network as a whole. Provan
and Kenis (2008) also stated that it is the role of network
management to work with members to enhance commitment
to network goals. The general idea is that there is a close
association between network involvement and goal consen-
sus, whether directly through governance involvement or
indirectly through involvement via working with the man-
agement staff. Making and keeping commitment to the net-
work is also one of the main descriptors of mobilizing in the
Agranoff and McGuire network management framework
(McGuire, 2002) and a distinctive aspect of orchestrational
work (Bartelings, Goedee, Raab, & Bijl, 2017). Therefore,
much of the network management literature revolves around
the argument that network managers are integral to network
commitment and that argument is the starting basis of this
article. Stated as a hypothesis:
Hypothesis 1 (H1): Strong ties to network management
will be a sufficient and necessary condition for network
members to perceive the network’s purpose to be congru-
ent with their own purpose.
Does goal congruency, however, occur only through net-
work management or are there other forms of involvement
through which goal congruency can be maintained? Basically,
are strong ties to network management required for goal con-
gruency (i.e., necessary in the above hypothesis) and/or are
ties to network managers enough for goal congruency (i.e.,
sufficiency in the above hypothesis). Especially in a large
heterogeneous network, relying on network managers to
ensure goal congruency with all network members may be
unmanageable. Although network management is likely
important for maintaining goal congruency, it is also likely
that goal congruency can be achieved through relationships
with other network participants, and not necessarily solely
through ties to network management.
Specifically, the argument here is that ties to network
managers will not necessarily be the relationships through
which greater goal congruence will be attained. Network
managers have no doubt a role to play in achieving goal

Lemaire 3
congruency, but that role may be manifested in other ways
rather than through the direct influence of a dyadic tie. For
instance, Saz-Carranza and Ospina (2011) proposed the
importance of bridging work by network administrative
organization (NAO) staff in managing the unity/diversity
tension in networks. According to Saz-Carranza and Ospina,
the objective of bridging work is to “link diverse members to
generate unity for joint action in the whole network” (p.
357). What bridging work entails is mediating member inter-
action, and the specific ways in which Saz-Carranza and
Ospina note that bridging work is performed are by “creating
space for dialogue and interaction, recognizing member
involvement, mediating among members, and disseminating
information across member organizations” (p. 350). The
bridging work of network management is, therefore, one of
the means by which the relationships across diverse groups
are developed, but it is the resulting bridging relationships
that in turn lead to greater goal congruency. I argue that the
tie portfolio of network members, especially when connected
with other network members across different domains or
with members responsible for shaping network goals, may
also be key to goal congruency.
Forms of Social Capital
To categorize the different types of ties comprising a net-
work member’s portfolio, the dimensions of social capital
will be used. Social capital is a concept first put forward by
the sociologists James Coleman and Pierre Bourdieu and
popularized by Robert Putnam (1993, 2000) in his case for
the disappearance of social capital in American society.
Although there are many different ways to conceptualize
what social capital is, one way to conceptualize it in relation
to network structure is as the outcomes of network relations,
or the resources that flow through network structures (Szreter
& Woolcock, 2004).
The use of the dimensions of social capital here is not
intended to delve into the concept of social capital, but only
to borrow the dimensions of social capital for categorizing
the types of network ties that organizations maintain. Putnam
(2000) distinguished between bonding and bridging social
capital: bonding being the relationships between individuals
who see themselves as similar and bridging being the rela-
tionships among individuals who are not similar. In recent
years, another dimension of capital has also been proposed,
linking capital, as a way to specify relationships across for-
mal institutions or power differentials.1 Specifically, linking
social capital is defined as “the norms of respect and net-
works of trusting relationships between people who are
interacting across explicit, formal or institutionalized power
or authority gradients in society” (Szreter & Woolcock,
2004, p. 6). In the study of health and welfare, linking social
capital has been shown to impact welfare, especially the
absence of linking capital in poor communities (Krishna,
2002; Narayan, 2000).
As Provan and Kenis (2008) note, networks involve a
range of interactions; therefore, in this article, bonding, bridg-
ing, and linking will be used to categorize the different types
of relationships maintained by members of a network.
Bonding ties are categorized as those ties among organiza-
tions in similar service domains and bridging as the ties across
service domains. Then as governance, according to Provan
and Kenis, is the use of institutions and structures of authority
and linking capital derives from the relationships across insti-
tutionalized power or authority structures, linking interac-
tions can be categorized as those linking network members
with the network governance structure. Categorizing the
types of relationships in a network based on bonding, bridg-
ing, and linking allows for the isolation of which types of
relationships are associated with greater goal congruence.
Table 1 provides an overview of the various types of ties and
details about their operationalization in this study.
One way to ensure goal congruency may be by connect-
ing organizations with similar goals. This argument is a logi-
cal starting point because of homophily forces (McPherson,
Smith-Lovin, & Cook, 2001) but could also be counter-logi-
cal when one considers organizational competitive forces
(Bunger, 2013). If organizations are connected with other
organizations in the same domain, perceiving goal congru-
ency between their own goals and the network goals may be
more likely because of the similarity of their mandates with
network partners. However, if competitive forces are strong
in the context of the network, then network members may
question how the network goal of working together with
their competitors is aligned with their own organizational
goals. Thus, the relationship between goal congruency and
bonding ties is more complicated and may depend on context
and ties to the network management as intermediary.
Hypothesis 2 (H2): A high number of bonding ties,
assessed as ties to others in similar domains, may be a suf-
ficient but not necessary condition for network members
to perceive the network’s purpose to be congruent with
their own purpose. However, this relationship between
bonding ties and goal congruency may depend on context
and/or be sufficient only in combination with a strong tie
to network management.
In fact, selecting network members with similar goals is
often contrary to the rationale behind most public, goal-
directed networks. Often times, the rationale behind a net-
work is to build the connections across service domains to
create a more seamless service system. Thus, an alternative
approach to goal congruence is learning from others in dif-
ferent domains, learning about the overlap and boundaries of
domains, and where the goal congruence lies. Contrary to the
pessimism of his contemporaries, Georg Simmel argued that
social cohesion could still be achieved in the changing mod-
ern society by membership in diverse groups because span-
ning groups prevent differentiation (Kanter & Khurana,

4 American Review of Public Administration 00(0)
2009). For instance, one of the characteristics of the
International Typographical Union (ITU) that may have pre-
vented it from succumbing to the fate of Michels’s (1915)
iron law of oligarchy was the dense interpersonal ties among
members across occupations (Lipset, Trow, & Coleman,
1956; Tolbert & Hiatt, 2009). Therefore, a high number of
bridging ties provide a mechanism for goal congruence. By
connecting with those in different domains, the perspective
of network members is expanded regarding what the greater
network goals are. Rather than leading to greater loyalty to
their own domains, this allows for recognizing how their
own goals provide an essential piece in the greater aims of
the network. Thus, bridging ties may be a sufficient condi-
tion for goal congruency, though not necessary since bridg-
ing ties are not the only means by which goal congruency
may be attained. Stated as a hypothesis:
Hypothesis 3 (H3): A high number of bridging ties,
assessed as ties to others in different domains, will be a
sufficient but not necessary condition for network mem-
bers to perceive the network’s purpose to be congruent
with their own purpose.
The link between bridging ties and goal congruency is
based on the influence that other network members in different
domains will have on a network member’s perspective of net-
work goals. An additional means through which goal congru-
ency may be attained is through linking ties. Linking ties
provide a more direct means than bridging ties through which
goal congruency can be attained. If network members have
ties with those setting the network goals, these ties provide
channels through which influence regarding network priorities
can flow. In classic organization theory, influence around
organizational goals was perceived as unidirectional. The top-
level managers of the organization indoctrinate “those at the
lower levels with general purpose” even though those at the
bottom of the organization, “the ‘ultimate’ contributors,” were
to inform the top of the actual conditions to which the top is
insulated (Barnard, 1938, p. 233). In Barnard’s view, which in
turned influenced others like Herbert Simon, those at the lower
level could inform those setting the purpose, but the purpose
was set at the top and communicated downward. This unidi-
rectional flow would likely contribute to the inevitability of
oligarchy; whereas in the ITU, leaders were prevented from
having complete agenda-setting control through the right of
members to use referenda to bring issues to a vote (Lipset
et al., 1956; Tolbert & Hiatt, 2009).
In a network setting where only a subset of network mem-
bers is involved in the whole network governance structure, if
network members have ties to the subset of members making
Table 1. Types of Ties and Operationalization.
Types of tie Definition Operationalization Measure Calculation and ranges
Network
management
Relationship with network
management
Strong tie to the
NAO staff
Multiplexity of confirmed
tie to NAO staff
Multiplexity based on number of
confirmed activity ties out of 6
possible ties, actual range was 0 to 6
with a mean of 3.89
Bonding Relationships between
those who see
themselves as similar
(Putnam, 2000)
Ties to
organizations
within the same
service sector
Egonet density, ratio of
ties to organizations in
the same service sector
as ego out of total
possible number of ties
to organizations in the
same service sector
Number of activities included in density
calculation, meaning total number of
6 possible activity ties considered for
each dyad. Possible density range is 0,
no tie of any type to an organization
in the same sector, to 100%, all 6
possible ties to all organizations in the
same sector. Actual range is 0% to
45%, with a mean of 16%
Bridging Relationships among
individuals who do not
see themselves as similar
(Putnam, 2000)
Ties to
organizations in
different service
sectors
Same egonet density as
above, but using ties to
organizations not in the
same service sector
Same as bonding; actual range was 0%
to 47% with a mean of 14%.
Linking Relationships across formal
institutions or power
differentials (Szreter &
Woolcock, 2004)
Ties to the
governance
structure
Same egonet density, but
using ties to members of
the Steering Committee
Same as bonding and bridging, but
separate range and mean values
were calculated for the group of
organizations not serving on the
Steering Committee and those
that were members of the Steering
Committee. Actual range for non–
Steering Committee members was
6% to 30% with a mean of 13%; range
of 18% to 63% with a mean of 37%
for Steering Committee members
Note. NAO = network administrative organization.

Lemaire 5
up the governance structure, then these access channels pro-
vide bidirectional means of influence. The channel is bidirec-
tional because they offer the means by which other network
members can influence the goals set by the governance mem-
bers or by which they can be influenced by those members of
the governance structure to agree with those goals. Therefore,
many ties to those shaping the direction of the network, or
linking ties to the governance structure, provide another type
of network interaction through which goal congruence can be
attained. Specifically, a high number of linking ties in a net-
work member’s portfolio of ties will also be a sufficient con-
dition for goal congruence, though again not necessary as
other conditions provide paths to congruency. Stated as a
hypothesis:
Hypothesis 4 (H4): A high number of linking ties,
assessed as ties to the network’s governance structure,
will be a sufficient but not necessary condition for net-
work members to perceive the network’s purpose to be
congruent with their own purpose.
Research Methods
Research Setting
The investigation of the hypotheses will be based on an anal-
ysis of the SACYHN, which was founded in 2001 to facili-
tate more decentralized services for children and youth and
to address the problem of fragmentation in the delivery of
health services for children. The mission of SACYHN was to
use the collective resources and expertise of participant orga-
nizations to advance high-quality, coordinated programs and
services for children, youth, and families. Health for this net-
work was defined broadly as health and well-being; thus, an
inter-sectoral perspective was important, and one of the goals
of the network was to build respect and collaboration across
organizations in multiple child-serving sectors. These
included both public and nonprofit organizations in physical
health, mental health, education, social services, and justice.
In addition, importance was also placed on coordinating ser-
vices across geographical regions to offer specialized health
services to children and youth throughout the Southern
Alberta region.
The governance model for SACYHN resembled the NAO
model proposed by Provan and Kenis (2008). The network
was governed by an entity created for the sole purpose of gov-
erning the network. This entity was comprised of a steering
committee and staff paid to support the work of the network.
Responsibility for the setting of policy and planning decisions
was placed on the steering committee, which consisted of a
subset of network members. The staff had responsibility for
the actual operations of SACYHN, which involved managing
and coordinating network efforts. The SACYHN staff con-
sisted of a full-time director and several full-time staff mem-
bers. The staff was officially employed by one government
agency, the Calgary Health Authority, which also contributed
about two thirds of SACYHN’s funding. The remaining fund-
ing came from the financial resources committed by partici-
pating organizations.
Since Southern Alberta is a large region, covering the
entire lower half of a province that spans over 250,000 square
miles, the SACYHN network was divided into four smaller
regions. These regions were comprised of three rural and one
urban (Calgary) region, and the network members in each
region were formally organized into subnetworks. The
responsibility of these subnetworks was to represent the
needs of that region and to serve primarily in an advisory
capacity to the SACYHN Steering Committee.
SACYHN is an example of leveraged government (Kettl,
2008). Membership was comprised of mostly public organi-
zations voluntarily working together to better address child
and youth health and well-being from a multi-organizational
approach. Although most of the members were traditional
public organizations, their mandates covered different
domains, either because of service sector or geographical
domain. Thus, SACYHN also is an example of a large het-
erogeneous whole network where alignment of network
goals with individual member mandates was important to its
success.
Data Collection
Data were collected on SACYHN between September 2008
and March 2009. An effort was made to collect the data
before any potential system-wide changes were imple-
mented, changes that were alluded to by a recent change in
political priorities. A complete reorganization of the health
system began in January 2009 and SACYHN was formally
disbanded in 2010. Thus, the data captured a mature network
before the start of its demise due to changing environmental
conditions. The results of the study indicated high internal
legitimacy overall, suggesting SACYHN provides an appro-
priate case to examine the established structure of network
relationships and the resulting variance in goal congruence.
The data collection included an organizational question-
naire and key informant interviews. For data collection pur-
poses, SACYHN was bounded by the formal structure; any
organization with representation on the Steering Committee,
a regional subnetwork, or a working group was asked to
respond to the questionnaire. The initial number of organiza-
tions contacted was 53, and the total number of respondents
surveyed was 137 (multiple individuals were asked to
respond on behalf of larger organizations). The actual orga-
nizational response rate was 88% (42/48—five organizations
did not have a respondent identified or no longer existed),
while the individual response rate was 76%. The question-
naire was one adapted from network research by Provan and
colleagues (cf. Provan, Huang, & Milward, 2009; Provan &
Milward, 1995) and included three main components: orga-
nizational demographics, questions regarding organizational

6 American Review of Public Administration 00(0)
ties (i.e., network relationships), and perspectives on the
impact of SACYHN.
The first main section of the questionnaire involved spe-
cific questions about the organization, such as its general
staff and budget size and the proportion of its resources that
were generally devoted specifically to children and youth
services. Respondents were also asked to break down the
percent of their resources, totaling to 100%, which were
spent on specific children and youth areas (i.e., mental
health, physical health, education, and recreational).
The second main section consisted of the network relation-
ship questions. Respondents were provided with a matrix list-
ing all 53 organizational members of SACYHN plus one
“organization” added in to represent the NAO staff and were
asked to identify which of six types of activity links (if any)
their organization had with the other 53 organizations over
the past year. More specifically, respondents were asked to
“Please go through the list and indicate which ones your orga-
nization has been involved with for the provision of services,
programs, or activities related to child and youth clients over
the past 12 months for each of the types of relationships listed.
Please only consider those relationships that have value to
your organization and are more than incidental and infrequent
(i.e., not just emails, large meetings, occasional phone calls,
etc.).” The types of activity links chosen were those deemed
by network staff, with input from the Steering Committee, as
the most important activities to accomplishing the work of
SACYHN: strategic planning, shared resources, service
delivery, education, research/evaluation, and information
sharing. Definitions were provided for each type of activity
on the questionnaire. In the last column of the matrix, respon-
dents were also asked to rate the overall quality of the rela-
tionship for each existing relationship as well as some
additional questions about their most critical relationships
and the organizations they admired most.
The final section of the questionnaire targeted perfor-
mance. The main performance indicators consisted of two
multi-item questions. Both questions included items target-
ing various dimensions of performance, such as quantity of
services, quality of services, and capacity of organizations
and the system. The items in the questions were selected
based on use in previous studies (Provan et al., 2009; Provan
& Milward, 1995) as well as discussions with the SACYHN
staff about which items would be most appropriate to the
case of SACYHN. The first multi-item question asked about
the impact of the network on the respondents’ organization
and included 18 items and an “other” item as well.
Respondents rated each item based on a 5-point Likert-type
scale ranging from 1 = “SACYHN has had a significant
negative impact on my organization” to 5 = “SACYHN has
had a significant positive impact on my organization,” with
the midpoint 3 = neutral impact. The second multi-item per-
formance question required respondents to rate the impact of
the network on the overall child and youth system. This
question included 20 items and an “other” item as well and
was rated on a 5-point Likert-type scale similar to the previ-
ous question. Following the multi-item questions targeting
performance was a question about the challenges of working
within the network and a final performance question request-
ing respondents to indicate the extent to which their organi-
zation identified with SACYHN. Respondents were asked to
respond to this item using a 5-point Likert-type scale with 1=
“My organization does not identify at all with SACYHN” to
5 = “My organization identifies itself as an integral part of
SACYHN.”
In the last stage of data collection, key informant inter-
views were conducted with SACYHN staff and key individu-
als in the system. The list of individuals who were contacted
for an interview was developed using strategic sampling
(Kumar, Stern, & Anderson, 1993). That is, individuals repre-
senting different sectors, different organizational levels, and
different levels of involvement in the network (i.e., core/
periphery) were identified, to get a representative but
informed sample of opinions across the full network. The
interviews were conducted by three different individuals, so a
list of questions was used as a guide. The goal of the inter-
views was to gain a better contextual understanding of the
network, its operation, and the interviewee organization’s role
in the network, so the interview protocol was used only as a
guide. Some of the interviews involved multiple interviewees
resulting in a total of 25 individuals who were interviewed
during 16 interviews. Interviews lasted from 30 to 75 min.
Aggregation and confirmation. As the unit of analysis for this
study is at the organizational level, the individual responses
to the questionnaire on behalf of organizations were aggre-
gated for each organization.2 Of the 42 total organizational
responses, there were multiple respondents for 20 of the
organizations. Aggregation was necessary for the network
data and for the performance data, but not for the descriptive
data on the organizations since only one respondent for each
organization was selected to respond to the descriptive
questions.
Different processes were used for aggregating the differ-
ent sets of data to best capture the reason for variation among
respondents. For the performance data, individual responses
were averaged by organization. These responses were aver-
aged to gauge a shared attitude among those in the same
organization, along the lines used to assess shared attitudes
such as trust in inter-organizational research (Zaheer,
McEvily, & Perrone, 1998). The average measure intra-class
correlation coefficients were computed to justify the aggre-
gation of the responses to the organizational level.3
For the network tie activity data, if there was more than one
respondent for an organization, relationships with other orga-
nizations were recorded as long as one respondent indicated a
relationship. For instance, if Respondents 1 and 2 represented
Organization A, but Respondent 1 indicated a tie with
Organization B and Respondent 2 did not, it would still be
recorded that Organization A indicated it had a tie with

Lemaire 7
Organization B. The rationale for using this approach is that
not all individuals would have been aware of the working rela-
tionships their organization had with all other organizations.
The various respondents represented different program areas,
management levels, or network involvement and thus, may
have maintained relationships with different organizations.
Once the data were aggregated from the individual level
to the organization level, the tie activity relationships were
then confirmed. The confirmation process for network data
is used to improve the reliability of the reported data, due to
the demands on respondent’s memory (Marsden, 1990). The
minimization confirmation process (Wasserman & Faust,
1994) was used, meaning the lower of the values indicated
by both organizations in a dyad for each specific type of tie
was recorded. For instance, if Organization A indicated a
planning tie with Organization B, but Organization B did not
indicate a planning tie with Organization A, then no tie was
recorded between the dyad. Thus, relationships were
recorded as confirmed as long as both organizations indi-
cated there was a relationship for a particular activity.
The confirmation process does result in a loss of data.
By confirming the relationships, the densities of each of the
networks of activity were reduced, the change in density
ranging from 14% to 46%. This is fairly typical (Wasserman
& Faust, 1994) but is important in improving reliability
since the data are based on perceptions of network ties. In
addition, what is also lost is the reciprocity or lack of reci-
procity of relationships, which can indicate where there
may be unequal relationships. However, the focus of this
study is on the overall structure of the network, and a more
conservative approach to the data was worth the trade-off.
The confirmation process using the minimization process
underestimates where there are links and increases the reli-
ability that the confirmed relationships do in fact exist. For
this study, reliably assessing the overall structure is of pri-
mary importance.
Analysis
To investigate the hypotheses, cs-QCA was employed. QCA
is a different approach from traditional variable-based statis-
tics and instead uses a set-theoretic approach (Ragin, 1987,
2000). It is based on Boolean algebra, which allows for an
assessment of the combinations of necessary and sufficient
conditions that lead to an outcome of interest. It draws upon
the fundamentals of case-study research and examines con-
ditions both within cases and across cases. QCA uses “truth
tables,” showing all theoretically possible combinations of
conditions that relate to the outcome and the number of cases
present for each combination. Too many conditions included
in the solution can “individualize” the cases making it diffi-
cult to find a pattern across cases (Ragin, 2008). Thus, the
conditions included should be chosen with care and directly
relate to the necessary and sufficient relationships proposed
by hypotheses.
The application of QCA to the study of networks is not
new (see Stevenson & Greenberg, 2000), and QCA has been
argued to offer several opportunities to advance the field of
interorganizational networks specifically (Raab, Lemaire, &
Provan, 2013). The method was employed for this study for
various reasons. Not only is QCA useful as an approach to
studying small to intermediate N data sets, but the approach
also allows for causal conditions to be highly correlated, as is
often the case with network structure variables. More impor-
tantly, however, QCA is valuable because of its ability to
handle causal complexity and equifinality (Fiss, 2011).
Different configurations of network members’ connections
could well result in goal congruence, and rather than looking
for whether or not any one structural feature, such as ties to
network managers, might explain goal congruency, the goal
is to find all the structural configurations that lead to the
same outcome. Also, it is possible there are so-called “Inus”
conditions, or conditions that are insufficient by themselves,
but are necessary parts of the overall “causal recipe” needed
to explain outcomes (Ragin, 2008). What is of interest here is
not the net effect of any one aspect of network structure, but
which configurations of structures are associated with goal
congruency.
The use of QCA in our field has become more common in
recent years, but important to note are the critiques of the
method. Often times, it is used when researchers have a small
N and traditional statistical approaches are not an option;
however, there is danger in using QCA in those instances.
There are a number of issues with QCA, including issues due
to limited diversity (Seawright, 2014), sensitivity to mea-
surement error (Hug, 2013) and model and parameter speci-
fication (Krogslund, Choi, & Poertner, 2015), and the
occurrence of false positives due to results produced by
chance (Braumoeller, 2015). These are important concerns to
note, but also important to recognize is that QCA is a funda-
mentally different approach from the statistical approach.
Thiem, Baumgartner, and Bol (2016) explain the differences
between the two and clarify the misconceptions between the
two regarding algebraic systems, hypotheses classes, and the
concept of causal complexity. QCA was chosen for this study
because the approach fits with the aim of this study which is
implicational association rather than covariate association
(see Thiem et al. for a discussion of this difference). What is
being examined in this study is not whether types of ties
cause goal congruence or even whether the increase of one
increases the other. Instead, the argument is that certain types
of ties may be conduits for achieving goal congruency and
thus to examine whether there is an association between the
presence of certain tie portfolios and the outcome of goal
congruency.
The basic QCA technique is for the analysis of binary
conditions, called crisp-set QCA (cs-QCA), in contrast to
fuzzy-set QCA (fs-QCA) which was designed to address the
way cases vary in the degree to which they belong to sets
(Ragin, 2000). I used a crisp-set because my research

8 American Review of Public Administration 00(0)
question is aimed at understanding congruence versus not
congruence and because the construct of congruence does
not lend itself well to interpretation of gradations of congru-
ency.4 This means that for the cs-QCA analysis, the outcome
and conditions are either present or absent and are not cali-
brated to represent a gradient scale.
A total of 38 organizations were included in the cs-QCA
analysis. Four of the network members were eliminated as
cases because calculating bonding ties for these organiza-
tions involved too many assumptions.5 These network mem-
bers were different from any other network members and
thus, determining which ties would be considered as bonding
versus bridging ties would have been an arbitrary decision.
The eliminated organizations were either a Treaty 7 organi-
zation, a Federal organization, or a regional organization
located outside the Southern Alberta region.
Measures
Two outcome measures and six causal conditions were
included in the analysis. Two ways of operationalizing goal
congruence were examined as outcomes: high level of iden-
tification with the network [hiident] and perceptions of high
impact on organizations [hiorgimp]. The six causal condi-
tions consisted of four types of network ties and two organi-
zational attribute conditions. The four network conditions
were strong ties to network staff [stiesf], high bonding ties
[hibond], high bridging ties [hibridg], and high linking ties
[hilink]. These four network conditions were included to
examine whether the hypothesized conditions, high in bridg-
ing and linking ties, are sufficient or whether the two other
type of common ties also need to be present or are them-
selves the ties that are sufficient. The two attribute conditions
were new members [new] and non-health and social service
members [nonhss], and the reasons for their inclusion are
provided further below.
Outcome sets. Two different measures were used to opera-
tionalize the outcome of goal congruency. The first is the
level of identification with SACYHN. This variable was
measured by the item on the organizational questionnaire
asking respondents to rate, using a 5-point Likert-type scale,
the extent to which their organization identifies with
SACYHN. This item captures how congruent network
members perceived their own organization’s goals with that
of the network’s; however, this measure is based on only
one question.
As multi-item scales are often a better alternative for
measuring constructs (Venkatraman & Grant, 1986), a sec-
ond outcome was also used to validate the results based on a
different way of operationalizing goal congruency. The
multi-item scale that assessed perceptions of the impact
SACYHN had had on member organizations was used as the
basis of the second outcome. The number of scale items was
reduced using factor analysis and two factors were derived,
“improving services” and “increasing capacity.” These two
factors were highly correlated (r = .84) and thus, were com-
bined into a single scale (10 items, α = .81) to use as an
indicator of goal congruence. The link between perceived
network impact on member organizations and goal congru-
ence may not be entirely obvious; however, if an organiza-
tion perceives the mission of the network to be similar to its
own core mission, then it would be more likely to rate the
impact of SACYHN on its own activities highly. Otherwise,
if a respondent does not perceive a strong link between the
mission of the network and their own organization, they
may perceive the network is having an impact, such as on
the service system which was the other performance scale
included on the questionnaire, but they would not be likely
to perceive the network as having a substantial impact on
their own organization’s activities. This operationalization
of goal congruency is not a direct measure, but rather a way
of indirectly measuring perceptions of goal congruence
which provides a multi-item scale to validate the direct one-
item measure I am also using.
To create sets of conditions based on these two measures
of goal congruence, both variables were compared with the
overall means of each variable. The means correlated to the
tilt to positive of the original scale of the respective measure,
and using the actual mean rather than the scale point allows
for considering individual responses in comparison to the
midpoint in the data. Therefore, if the level of identification
with SACYHN perceived by organizational respondents was
greater than 3.3, the overall mean, then this organization was
included in the set of organizations with higher perceived
goal congruency with SACYHN based on identification
(hiident, n = 16, code = 1).6 For the other measure, per-
ceived organizational impact, if an organization’s mean was
greater than 3.65, the overall mean for perceived organiza-
tional impact,7 that organization was included in the set of
organizations with higher perceived goal congruency with
SACYHN based on organizational impact (hiorgimp, n =
26, code = 1).8 The descriptive statistics for all measures are
provided in Table 2.
Causal conditions. The first causal condition, strong ties
to network staff [stiesf] was measured by tie multiplexity
between network members and the SACYHN staff. Multi-
plexity has been argued as a way of capturing tie strength
because multiplex ties are more robust (Provan & Milward,
2001). Ties ranged from 0 (no tie) to 6 (a tie based on all
six possible activities), and multiplex ties are those that are
based on several activities. A crisp-set condition was created
by coding as “present” those ties that were equal to or greater
than 4 to capture those relationships with network managers
which were especially strong.9 Thus, if a network member
had a relationship with the SACYHN staff based on four or
more activities, then this was coded as 1 (n = 23).
The second causal condition was high bonding ties
[hibond]. Bonding was operationalized in the context of

Lemaire 9
SACYHN as those ties across regions but within the same
service sector. One impact of the SACYHN network dis-
cussed in the interviews was the relationships it helped build
across regions, connecting professionals with their counter-
parts in other regions. For instance, several interviewees dis-
cussed how medical professionals in one region could now
pick up the phone and call their counterpart in another region
whenever they needed advice. Although located in different
regions, the similarity due to professional domain conforms
to the conceptualization of bonding as ties between those
who see themselves as similar (Putnam, 1993).
The categorization of service sector was based on the
responses to a question in the questionnaire indicating the
proportion of its resources that were generally devoted to
specific services in the child and youth area. To create a crisp
set of those organizations high in bonding ties, the egonet
density of network member’s bonding ties was calculated.
Specifically, bonding density was calculated as the number
of ties a network member had with network members across
regions, but in the same service sector, out of the total num-
ber possible. The multiplexity of the ties was also taken into
account; thus, the densities were calculated using the total
number of types of activities reported and confirmed divided
by the total number of relationships possible. The density
values for each organization were then compared with the
overall mean score based on all network members’ bonding
density scores. If the density of a network member’s bonding
ties was greater than the overall mean, 0.16, then this was
coded as 1, indicating the network member’s egonet was
higher than average in bonding ties (n = 18). If the density of
a network member’s bonding ties was less than the overall
mean, then this network member’s score was coded as 0.
The third network tie condition was high bridging ties
[hibridg]. Bridging ties were operationalized here as the ties
across service sectors, which reflects ties to different
domains. The egonet density of a network member’s ties to
those organizations in service sectors other than its own was
calculated, using the same process as for bonding ties, and
compared with the overall mean density for this variable. If
the density of a network member’s ties was greater than the
mean density score of all the network member’s bridging
densities, 0.14, then this network member was considered
to have a higher than average number of bridging ties and
thus in the set of members high in bridging ties (n = 16,
code = 1).
The final network causal condition was high linking ties
[hilink]. Linking ties were operationalized as an organiza-
tion’s ties to Steering Committee members, which are
intended to distinguish the ties to the formalized governance
entity from ties to the network management staff (who
reported to the committee but were not members of the com-
mittee). This condition was created in much the same manner
as the previous two conditions, except for one major differ-
ence. The egonet density of network member’s ties to
Steering Committee members was still calculated, and this
density was compared with the overall mean to create the set
of those organizations high in linking ties. However, because
some of the network members serve on the Steering
Committee, and thus are quite likely to have strong ties to the
other Steering Committee members, these organizations
either needed to be eliminated from the QCA analysis or
handled in a different manner. Rather than eliminate all of
these members from the QCA analysis, the causal condition
accounted for their greater propensity to have more linking
ties, which was accomplished by the use of two mean values
for the construction of the crisp set. The egonet density of all
the network members’ linking ties were calculated, but then
two means were calculated: a mean density score for just the
Steering Committee members and then one mean score for
the remaining organizations. Then, if a Steering Committee
member’s linking density score was greater than the Steering
Committee mean, 0.37, it was considered as high in linking
Table 2. Descriptive Statistics for All Measures.
Outcome/condition Name M SD Minimum Maximum Crisp set = 1 Crisp set = 0
High network identification hiident 3.37 0.74 2 5 16 22
Perceptions of high network
impact on member
organizations
hiorgimp 3.65 0.44 2.43 4.5 26 12
Strong tie to SACYHN staff stiesf 3.89 2.14 0 6 23 15
High bonding tie density hibond 0.16 0.12 0 0.45 16 22
High bridging tie density hibridg 0.14 0.14 0 0.47 18 20
High linking tie density hilink 19 19
Non–Steering
Committee members
0.13 0.08 0.06 0.30 7 10
Steering Committee
members
0.37 0.04 0.18 0.63 12 9
New network members new 5.86 1.72 0.75 7.00 4 (new
organizations)
34
Non-health and social service nonhss 0 (mode) 0 1 14 24

10 American Review of Public Administration 00(0)
ties in comparison with other Steering Committee members
and thus coded as in the set of those organizations high in
linking ties (n = 12, code = 1). The non-Steering Committee
member organizations’ linking densities were compared with
the overall mean excluding the Steering Committee mem-
bers and if an organization’s linking density score was higher
than this mean, 0.13, it was considered high in linking ties in
comparison with other non-Steering Committee members
and thus also coded as in the set of those organizations high
in linking ties (n = 7, code = 1). The total set for the pres-
ence of high linking ties (code = 1), including Steering
Committee members and non–Steering Committee mem-
bers, consisted of 19 organizations.
Two organizational attributes were also included as causal
conditions. Including contextual conditions in QCA helps to
eliminate contradictory configurations in the analysis (Ragin,
2008). Some preliminary descriptive analysis suggested the
need to include two organizational attributes to determine
how the causal configurations differ for different sets of
organizations. The two conditions are new members [new]
and non-health or social services [nonhss]. Most of the net-
work members had been involved in the network since incep-
tion of the network, or close to it, approximately 6 to 7 years
(assessed in the first section of the organizational question-
naire). Only a few organizations had recently joined the net-
work and because of their more recent involvement, it would
be unlikely they would have built up the same bonding,
bridging, and linking ties as the other network members.
Therefore, any organization that had been involved in the
network for less than 3 years was coded as being in the set of
new members (n = 4, code = 1).
A condition was also included for the network members
not in either the health or social services domain. From the
preliminary and basic descriptive analysis, it was apparent
that the organizations in the sectors other than health or
social services (education, justice, and community services)
did not play as prominent a structural role in the network as
did the organizations in the health and social service sectors.
In addition, the overall means for perceptions of impact were
lower for these sectors than for health and social services.
Therefore, since the network was much more oriented toward
the health and social service sectors, perceptions of goal con-
gruence by those in other sectors could be more difficult to
attain. Thus, the network members were divided into two
broad categories: non-health and social services (n = 14,
coded as 1) and all other organizations (n = 24, coded as 0).
This condition then allows for causal configurations to
emerge for these network members specifically. For instance,
it may be that high linking ties is all that is needed for net-
work members in the health and social services sector to per-
ceive goal congruence. However, high linking ties may not
be enough for network members in other sectors, but instead
a strong tie to the SACYHN staff may be necessary for build-
ing their commitment to the network. Or it could be that
bonding ties are more subjective to competition forces for
those in these other sectors than those in health and social
services and thus, as proposed in H2, bonding ties would not
be the same conduit for goal congruence for this group as for
the other group.
Results
Tests for any necessary conditions were conducted before the
analysis as recommended by Ragin (2000), using the “neces-
sary conditions” crosstabs procedure included in the FSQCA
3.0 software (Ragin & Davey, 2016). No necessary condi-
tions were found, though strong ties to SACYHN staff were
close (81% consistency identification outcome and 77% for
organizational impact). Analysis was then performed on truth
tables, which show the theoretically possible causal combina-
tions, the number of cases fitting each combination and the
consistency of each combination. A consistency cutoff value
must be selected to determine which recipes are consistent
enough to be associated with the presence of the outcome.
The consistency scores selected were 0.78 and 0.89 for high
identification with SACYHN and high perceived organiza-
tional impact, respectively. These cutoff values were chosen
following the recommendation by Ragin (2008) to use a large
gap between consistency values for the selection of the cutoff
threshold. The next highest consistency values were 0.5 and
0.66, respectively (see the truth tables provided in the
Appendices A and B). The parsimonious solutions were
selected and are provided for each solution and presented in
Tables 2 and 3. Debate exist over which solutions are best to
present, so the parsimonious solutions were selected based on
the argument that parsimony is more tightly connected con-
ceptually with causality (Baumgartner, 2015).
The first solution, in Table 3, provides the results for the
set of organizations that highly identified with SACYHN. To
evaluate a QCA solution, the consistency of the solution
must be high enough to indicate that when the solution reci-
pes are present, the outcome is present most of the time as
well. The final solution had moderately high consistency,
0.86, meaning that when the solution recipes are present, the
outcome is also present 86% of the time. More specifically,
14 organizations fit at least one of the recipes and of those
14, 12 also highly identified with SACYHN.
The second consideration is coverage, and the solution in
Table 3 had moderately high coverage at 0.75. This means
that this solution explains 75% of the cases where the out-
come was present. More specifically, of the 16 organizations
that highly identified with SACYHN, 12 of the 16 are
explained by at least one of the solution recipes. Although
this coverage is not especially high, indicating that some of
the cases are not being explained, a high majority of the
cases are explained and more importantly, explained with
moderately high consistency.
The results indicate that the hypotheses are partially sup-
ported. The first recipe explains only one case, a new health
and social service organization, and indicates that despite not

Lemaire 11
having any of the other conditions present, this new organi-
zation highly identified with SACYHN. The second recipe
represents most of the organizations that highly identified
with SACYHN, 10 of the 12, and indicates that these health
and social service organizations had a strong tie to SACYHN
staff and high bridging ties. The last recipe (Recipe 3)
explains two organizations that highly identified with
SACYHN, and these two organizations had a strong tie to
SACYHN staff combined with both high bonding ties and an
absence of high linking ties.
The hypotheses are partially supported in that a strong tie
to SACYHN staff (H1), high bonding ties (H2), and high
bridging ties (H3) are associated with goal congruency when
measured by identification; however, high linking ties (H4)
are not. For a minority of the organizations high bonding ties
were sufficient (H2), but for the majority of the organizations
that highly identified with SACYHN, these organizations
were in the health and human service sector and had a strong
tie to SACYHN staff combined with high bridging ties.
Therefore, rather than bonding ties being the condition
needed to be combined with a tie to network management as
proposed in H1, it is bridging ties where that combination is
required for sufficiency. The strongest results indicate that a
strong tie to SACYHN staff and bridging ties are Inus condi-
tions and to attain sufficiency, they must be combined
together. This combination that emerged is important and can
be summarized as goal congruency being associated with a
strong tie to network management plus high bridging ties.
Although the results from the analysis of identification
with SACYHN partially support the hypotheses, the results
explain only 12 of the 38 organizations. This outcome set
was also based on only a one-item question; therefore, the
results presented in Table 4 provide additional leverage to
the analysis. The results in Table 4 are for the high perceived
organizational impact outcome set.
Overall, the solution presented in Table 4 had high consis-
tency (0.95), meaning that the outcome is present 95% of the
time that these recipes are also present. More specifically, 21
organizations fit the recipes included in the solution and of
those 21 organizations, 20 also had perceptions of high orga-
nizational impact. The coverage for the solution was again
moderately high, 0.77, which again means that 77% of the
cases where the outcome was present are explained by the
solution. More specifically, of the 26 organizations that rated
organizational impact high, 20 of those are explained by at
least one of the recipes of the solution.
The first recipe of this solution indicates that five organi-
zations that perceived high organizational impact had high
linking ties and not high bonding ties. The second recipe
explains one new health and human service organization that
lacked any of the other conditions. The third recipe explains
the highest number of cases, 14, with the combination of a
strong tie to SACYHN staff plus a high number of bridging
ties. The final two recipes represent the same two organiza-
tions and indicate that bonding ties alone were sufficient for
these organizations.
The results signify that again the hypotheses are partially
supported and are similar to the results for high identification
with one major exception. The exception is that for this out-
come, high linking ties emerged as a sufficient condition
(H4). Otherwise, the results are the same in that high bond-
ing ties are sufficient for two organizations, but more cases
are explained by the combination of a strong tie to SACYHN
staff and high bridging ties.
Taken all together, the results indicate that bonding ties
may indeed be a path to goal congruency; but at least in the
case of SACYHN, bonding ties are not the type of tie associ-
ated with goal congruency for the majority of the network
members. Instead, the recipes that represent most of the
organizations perceiving goal congruency were those that
either had high linking ties or a strong tie to network man-
agement along with high bridging ties. These results suggest
that ties to the governance structure is one path to perceiving
goal congruency, at least in the perceived impact that the net-
work will have on organizations, but not necessarily so when
it comes to identifying with the network. Alternatively, goal
Table 3. Crisp-Set Qualitative Comparative Analysis Causal Recipes for Presence of Network Identification.
Causal recipes Description Raw coverage Unique coverage Consistency Number of cases
1. NEW × nonhss New, not non-health and
social services
0.06 0.06 1.00 1
2. STIESF × HIBRIDGE × nonhss Strong tie to SACYHN staff,
high bridging ties, not non-
health and social services
0.62 0.56 0.83 10 (9 unique)
3. STIESF × HIBOND × hilink Strong tie to SACYHN staff,
high bonding, not high
linking ties
0.12 0.06 1.00 2
Solution coverage 0.75 (12 of the 16)
Solution consistency 0.86 (12 of the 14)
Note. Consistent with QCA best practices, uppercase letters indicate presence of a condition and lowercase indicates absence of a condition. Parsimo-
nious solution. Model: hiident = f(nonhss, new, stiesf, hibond, hibridg, hilink). Consistency cutoff: .78. SACYHN = Southern Alberta Child and Youth
Health Network; QCA = qualitative comparative analysis.

12 American Review of Public Administration 00(0)
congruency may manifest for those organizations that are not
necessarily connected to the governance structure, but do
have a strong tie to network management and are connecting
with organizations in different service domains.
A key finding from this analysis is also that a strong tie to
the SACYHN staff is an important condition, but it is not a
necessary or a sufficient condition. Many of the organiza-
tions, including those that did not highly identify with the
network or perceive high organizational impact, had a strong
tie to the SACYHN staff and some organizations that did
highly identify with the network and perceived high organi-
zational impact did not have a strong tie to the SACYHN
staff. What the recipes overall reveal is the combination of a
strong tie to network management with bridging and this
combination as a substitute for ties to the governance struc-
ture. This finding that a strong tie to SACYHN staff needs to
be combined with bridging ties for the outcomes to be pres-
ent is valuable because it demonstrates the critical role of the
network staff in building and maintaining positive relation-
ships across domains. Ties to SACYHN staff were important
for facilitating the creation of bridging ties, as indicated by
the following quote from a regional committee member in
the education sector who perceived goal congruency:
I had a situation for example. I had report from a doctor and I
had no one to interpret the medical document . . . what I needed
to know is what is the impact of this diagnosis on the child in
terms of how they can learn because I was trying to place this
child. I got help from [SACYHN staff member], she went
through the medical system and got me people who could tell
me exactly what is this and what is the impact.
In addition, one likely reason for the combination of a strong
tie to SACYHN staff with bridging is that facilitating the ties
alone is not the end of the story; rather, the role of the staff is
also to address tensions to sustain the constructive nature of
the relationships. The following quote from one network
member perceiving goal congruence and whose organization
fits the relevant recipes in the results suggests the influence
of the SACYHN staff in facilitating diverse and positive
relationships: “So I think there is definitely, by bringing any
diverse group together, there’s potential for conflict, but
again it’s let’s make this work. And it’s the attitude of [execu-
tive director] that’s really facilitating that, and its worked.”
The ties to network managers may not be the ties that ulti-
mately result in goal congruence, but the work of network
managers is in fostering the means for the bridging ties to be
formed and then helping to ease the tensions that may arise
from those relationships.
Discussion
In this article, I examined goal congruence in a large, hetero-
geneous network and the portfolio of network members’ ties.
The literature emphasizes the role of network management in
building relationships and in mobilizing network members to
contribute to the work of the network. One facet of contribut-
ing to the work of the network is in justifying that the work
of the network is aligned with network members’ organiza-
tional mandates. The key question driving this research is
whether goal congruency is dependent solely on ties to net-
work management or whether various types of relationships
among network members may also be an avenue for shaping
perceptions of goal congruency.
I hypothesized that bonding ties, ties among those in the
same service sector, may be a logical type of relationship
associated with goal congruency; but I also argued that
Table 4. Crisp-Set Qualitative Comparative Analysis Causal Recipes for Presence of High Perceived Organizational Impact.
Causal recipes Description Raw coverage Unique coverage Consistency Number of cases
1. HILINK × hibond High linking ties and absence of high
bonding ties
0.19 0.12 1.00 5 (4 unique)
2. NEW × nonhss New, not non-health and social
services
0.04 0.04 1.00 1
3. STIESF × HIBRIDGE Strong tie to SACYHN staff and high
bridging ties
0.54 0.46 0.93 14 (13 unique)
4. HIBOND × nonhss
× stiesf × hibridge
High bonding ties, not non-health and
social services, absence of strong
tie to SACYHN staff and high
bridging ties
0.08 0 1.00 2 (0 unique)
5. HIBOND × nonhss
× hilink × hibridge
High bonding ties, not non-health
and social services, absence of high
linking and bridging ties
0.08 0 1.00 2 (0 unique)
Solution coverage 0.77 (20 of the 26)
Solution consistency 0.95 (20 of the 21)
Note. Consistent with QCA best practices, uppercase letters indicate presence of a condition and lowercase indicates absence of a condition. Intermedi-
ate solution. Model: hiorgimp = f(nonhss, new, stiesf, hibond, hibridg, hilink). Consistency cutoff: .89. SACYHN = Southern Alberta Child and Youth
Health Network; QCA = qualitative comparative analysis.

Lemaire 13
bonding ties may be complicated by competitive tensions.
Bonding ties did emerge as a sufficient path associated with
goal congruency in my analysis, but only for two organiza-
tions. The limited number of cases explained by bonding ties
indicates that other types of ties may be more beneficial con-
duits for influencing perceptions of goal congruency.
One viable means of attaining goal congruence between
organizational member goals and network goals is via bridg-
ing ties. Bridging ties were found to be associated with goal
congruence. In a case like SACYHN, where one of the stated
goals is to work across service domains, network members
working more closely with organizations outside their own
service domain were more likely to perceive goal congru-
ence between their own organization and the network’s.
Similar to arguments in classic research and theorizing on
organizations, by connecting with those in different domains,
the perspective of network members is expanded regarding
what the greater network goals are. By working across
domains, network members’ views of what the network goals
are to be and the role their own organization plays in those
goals can be expanded by learning more about the overlap
among service domains. This finding is supported by the dis-
cussions in the interviews. One of the outcomes of the
SACYHN network according to those interviewed was in
finding those areas of overlap among different agencies and
their mandates. What this led to was the discovery of the
directions where agencies could move together without los-
ing sight of their own organizational mandates. As noted in
an interview with a SACYHN management staff member,
There was less understanding before [the network], of internal
limitations, of mandates. Now there is much greater trust, much
greater understanding, much greater willingness to air internal
business. We have gotten much better at identifying the issues
and needs around children and youth in our region and willingness
to share our resources and share our ideas collectively.
The network becomes less about the priorities of their own
service sector and more about the role their own organization
plays in attaining the broader goals of the network as a whole.
As noted by one health representative who perceived goal
congruency and whose organization had a diverse tie portfo-
lio, “SACYHN is good at keeping people engaged in the big-
ger picture because it is easy to get side-tracked into your
daily focus.” These comments were made to stress the bene-
fit that SACYHN and its focus on building inter-sectoral
relationships had had on organizations; that is, the bridging
ties and the impact those ties had had on helping organiza-
tional representatives see their own goals as part of the
broader network goals.
I also hypothesized that linking ties, which are the ties to
the network’s governance structure, would also be a means
of attaining goal congruence. Although a network member
may not be directly involved in the governance of the net-
work, if the governance norms are to align priorities with all
network members, then having strong ties to the governance
structure allows for that flow of influence between network
members in general and those involved in setting the direc-
tion of the network. Through these ties, network members
can be assured that the governance leaders recognize their
priorities and the governing leaders can ensure they are
learning about the priorities of all network members. In the
case of SACYHN, alignment of provincial and regional pri-
orities was given precedence and the norms established were
in using the network as a means of learning about the priori-
ties at each level and working to align them. For instance, as
one provincial representative and Steering Committee mem-
ber who perceived goal congruence noted in an interview,
I think SACYHN has done quite a good job and I think that even
though a political head like [the former Honorary Chair] could
have influenced the agenda quite markedly so that it didn’t meet
the communities’ needs, I don’t think that happened. I think that
the table, the Steering Committee definitely heard the issues and
set the priorities around those issues.
At least in the case of SACYHN, where early Steering
Committee leaders had established the norms of aligning
provincial and regional priorities, network members with a
high number of linking ties to the Steering Committee tended
to also perceive congruence in their own organizational goals
and the goals of the network.
Attaining commitment to network-level goals is essential
to ensure network involvement, because of the voluntary
aspect of networks. Therefore, in governing a network, those
setting the direction of the network must ensure that they set
direction that will be supported by all network members. The
way to ensure goal congruence is either by aligning the goals
of the network with the goals of organizational members or
by expanding the perspectives of organizational members to
recognize the importance of broader, network-level goals for
their own organization. I have argued and found support for
the importance of bridging and linking ties for these strate-
gies. Bridging ties across domains are essential for network
members to learn from others in different domains and
expand their perspectives of network goals. Linking ties to
the network governance structure are also important for bidi-
rectional influence to flow from those involved in setting
purpose and the other network members not involved in net-
work-governing decisions. Therefore, choosing the appropri-
ate governance form for a network is only one step in the
process. Governing the relationships across the network is
also necessary to ensure high enough goal congruence to
support that governance form (Provan & Kenis, 2008).
In the case of SACYHN, network members not in the
health or social service sectors perceived lower goal congru-
ency between their own organizational goals and network
goals overall. There are many possible reasons for this, and it
cannot be determined from this study why members not in
the health and social service sectors perceptions differed

14 American Review of Public Administration 00(0)
from those in the other sectors. However, for those non-
health and social service organizations that did perceive
greater congruency between their organizational goals and
the network’s, bridging ties were key. Bridging ties across
domains may be especially essential for those network mem-
ber organizations in domains not as closely related to the net-
work purpose.
If bridging and linking ties are indeed essential for the
governing of purpose, who ensures those relationships are in
place? I initially argued that strong ties to the network man-
agers are not the channels through which goal congruence is
attained; however, strong ties to network managers were
indeed found to be essential, especially in regard to bridging
ties. The ties to network managers may not be the ties that
ultimately result in goal congruence, but the work of network
managers is in fostering the means for the bridging ties to be
formed and sustained. As Saz-Carranza and Ospina (2011)
found, network managers were the ones who created the
spaces for diverse members to link with one another. Paquin
and Howard-Grenville (2013) also found what they called
“orchestrators” are important for building broad interest and
support for network activities, accomplished via facilitating
interaction spaces and encouraging both serendipitous ties
and strategic ties. This is consistent with the work of the
SACYHN staff, as the following example describes:
Very high quality pediatric education of all kinds that nobody
ever had access to unless they worked at the hospital. And so we
[SACYHN staff] just used that as an opportunity to begin to
invite other child serving sectors to attend and we didn’t get
permission for that . . . And then finally, we actually formalized
that. We actually hosted an education think tank and we invited
a lot of the people who were responsible for putting on all that
education together and talked about what we had been doing
kind of underground and formalized an agreement with them.
What I found with this study of the SACYHN case is that
as others have argued (e.g., Agranoff, 2007; Kickert, Klijn,
& Koppenjan, 1997), network management is fundamental
and ties to network management are likely essential; but one
of the primary roles of network management is in facilitating
and sustaining the relationships among network members
that are also essential for network functioning. As Macciò
and Cristofoli (2017) found, even with an established NAO,
network endurance is reliant on a skillful network manager
who can “create an interactive environment, resolve tensions
among partners and build consensus and commitment to
common goals” (p. 12). Even if a whole network has a gov-
ernance form and the network purpose has been established,
constantly maintaining participation in that work takes effort.
Governing a network not only requires formulating the pur-
pose of the network, but also in governing the structure of
relationships is necessary to ensure the work of the network
aligns with network member’s own goals. The analysis pre-
sented in this article offers one approach for considering how
the structure of relationships may be a critical component in
the governance of purpose.
Conclusion
Network orchestration is vital to attaining the architecture of
relationships, such as bridging and linking ties, that is key for
elements of an effective network, such as goal congruence.
Understanding which relationships are critical for network
managers to facilitate, and why, is imperative for illuminat-
ing the black box of network functioning (Agranoff &
McGuire, 2001). This work advances theory on the types of
relationships that are essential to ensure goal congruence in a
network. It offers evidence indicating that facilitating bridg-
ing and linking ties may be as fundamental ingredients as are
the ties to network managers, at least in regard to goal
congruency.
However, further research is needed to test the arguments
put forth in this article. The limited empirical evidence
offered here is not conclusive in supporting the role of bridg-
ing and linking ties for attaining goal congruence. In differ-
ent network contexts, alternative combinations of linking,
bridging, and bonding may be necessary to explain outcomes
for different groups of network members. There are also dif-
ferent ways that the forms of social capital could be mea-
sured, and relying on network data is not free of measurement
error; and as Hug (2013) notes, measurement error can affect
the conclusions drawn from QCA results. In addition, this
analysis only focused on tie portfolios and two context con-
ditions, but further analysis should examine whether ties are
indeed the mechanism influencing goal congruency or
whether these results are due to chance (Braumoeller, 2015).
QCA is unreliable when limited diversity is a factor
(Seawright, 2014); therefore, this analysis only focused on
examining the key focus of tie portfolios. The next step
would be to examine tie portfolio in relation to other organi-
zational or network-level factors. Furthermore, the snapshot
of this network at one point in time without a comparison
case does not allow for empirical validation of the argument.
Further examination of the ideas presented here would need
to be empirically studied under a cross-comparative or longi-
tudinal research design.
There is still a great deal of research needed to under-
stand the management and governance of whole goal-
directed networks. This work makes only a small contribution
given the limitations of a cross-sectional case study. The
case study, however, does involve the analysis of 38 organi-
zations, which is still a small N, but finding consistent pat-
terns in tie portfolios and goal congruency across 38
organizations suggests a worthy venue for future study.
More research should look into whether these same tie port-
folio patterns hold in other network context, across varia-
tions in network governance form, and in relation to other
factors. Our understanding of network governance and

Lemaire 15
management will not be complete until we better understand
the specifics of relationship building that may be necessary
to support network governance and the role of network man-
agement in building and sustaining effective goal-directed
service delivery networks. Evidence has been accumulated
regarding the importance of network managers and building
relationships, but a better of understanding of which rela-
tionships are the levers that are linked to which outcomes is
necessary to progress scholarship and practice on how to
make public sector networks work.
Appendix A
Truth Table for Network Identification.
Hilink Hibridg Hibond Stiesf New Nonhss Number Hiident Raw consist. PRI consist. SYM consist.
1 1 0 1 0 0 2 1 1 1 1
0 1 1 1 0 1 1 1 1 1 1
0 1 1 1 0 0 1 1 1 1 1
0 0 0 0 1 0 1 1 1 1 1
1 1 1 1 0 0 9 1 0.777778 0.777778 1
0 0 1 0 0 0 2 0 0.5 0.5 1
1 0 0 1 0 0 2 0 0.5 0.5 1
0 0 0 0 1 1 3 0 0.333333 0.333333 1
0 0 0 1 0 1 3 0 0.333333 0.333333 1
1 0 1 1 0 0 2 0 0 0
0 0 1 0 0 1 1 0 0 0
0 1 0 1 0 1 1 0 0 0
0 1 1 0 0 0 1 0 0 0
0 0 0 0 0 1 3 0 0 0
0 0 0 0 0 0 4 0 0 0
1 0 0 1 0 1 1 0 0 0
1 1 1 1 0 1 1 0 0 0
Appendix B
Truth Table for Organization Impact.
Hilink Hibridg Hibond Stiesf New Nonhss Number Hiorgimp Raw consist. PRI consist. SYM consist.
1 1 1 1 0 1 1 1 1 1 1
1 0 0 1 0 1 1 1 1 1 1
0 1 1 1 0 1 1 1 1 1 1
0 1 1 1 0 0 1 1 1 1 1
0 1 0 1 0 1 1 1 1 1 1
0 0 1 0 0 0 2 1 1 1 1
1 0 0 1 0 0 2 1 1 1 1
0 0 0 0 1 0 1 1 1 1 1
1 1 0 1 0 0 2 1 1 1 1
1 1 1 1 0 0 9 1 0.888889 0.888889 1
0 0 0 0 1 1 3 0 0.666667 0.666667 1
0 0 0 1 0 1 3 0 0.666667 0.666667 1
1 0 1 1 0 0 2 0 0.5 0.5 1
0 0 0 0 0 0 4 0 0.25 0.25 1
0 0 0 0 0 1 3 0 0 0
0 0 1 0 0 1 1 0 0 0
0 1 1 0 0 0 1 0 0 0
Note. PRI = proportional reduction in consistency; SYM = symmetric consistency

16 American Review of Public Administration 00(0)
Acknowledgments
I would like to thank Keith Provan, Brint Milward, Janice Popp,
Carol Adair, and all of the Southern Alberta Child and Youth Health
Network (SACYHN) coordinators for their help with the formula-
tion of the research and the collection of the data. I would also like
to additionally thank Keith and Brint for their helpful comments on
the manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect
to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support
for the research and/or authorship of this article: The Southern
Alberta Child and Youth Health Network provided partial support
for research expenses.
ORCID iD
Robin H. Lemaire https://orcid.org/0000-0002-5191-6364
Notes
1. Szreter and Woolcock (2004) proposed linking capital as a
form of bridging capital because it is capital stemming from
relationships across different groups (i.e., across different
power structures). However, in this article linking capital is
used to categorize relationships related to the governance
structure and because the governance structure is comprised
of a subset of members of the network, at times those could
be considered bridging ties but at other times those could be
considered bonding ties. As the goal of this article is to use the
forms as ways to capture the different types of relationships
in the network without teasing out whether linking is indeed a
form of bridging, linking will be treated as a separate category.
2. Respondents were specifically asked in the questionnaire to
respond on behalf of their organization, considering ties and per-
ceptions of colleagues. Specific respondents were recruited using
the criteria of organizational role and involvement in the network.
For larger organizations, multiple respondents were recruited to
better capture the different knowledge and perspectives across
the organization. Thus, aggregating individual responses to the
organization level is consistent with decades of organizational
research that has relied on the aggregate of individual responses
to represent organizations (Klein, Dansereau, & Hall, 1994).
3. A one-way analysis of variance (ANOVA) was used to analyze
the intraclass correlations of responses within organizations,
with a statistically significant difference between groups (F =
2.17, p = .032). The ICC was also calculated as .54, in the range
of fair according to the guidelines provided by Cicchetti (1994).
4. Another rationale for the use of cs-QCA (crisp-set qualita-
tive comparative analysis) is that it is simpler to explain to the
general readership. However, as reviewers often question the
choice of crisp set, preliminary analysis using fuzzy-set QCA
(fs-QCA) was conducted and produced similar results. As the
fuzzy-set results did not add additional value in comparison to
the crisp-set results and the crisp-set is simpler and consistent
with the research question, I opted for the crisp-set approach.
5. Eliminating these four as cases means they were not included
in the analysis of the outcome, but ties to these four were still
incorporated in the analysis of the other cases to preserve their
role in the structure of the network.
6. In regard to the original survey scale, 3 = “My organization
identifies itself somewhat with SACYHN.” Higher than 3
indicated stronger positive identification, whereas lower than
3 indicated weak or absent identification.
7. In regard to the original survey scale, 3 indicated neutral
impact, greater than 3 indicated positive impact, and less than
3 indicated negative impact.
8. All 16 organizations in the network identification outcome set were
also in the organizational impact outcome set; however, the organi-
zational impact outcome set captured 10 additional organizations.
9. The mean was 3.84; therefore four activities indicated a tie that
was higher than average.
References
Agranoff, R. (2007). Managing within networks. Washington, DC:
Georgetown University Press.
Agranoff, R., & McGuire, M. (2001). Big questions in network
management research. Journal of Public Administration
Research and Theory, 11, 295-326.
Barnard, C. I. (1938). The functions of the executive. Cambridge,
MA: Harvard University Press.
Bartelings, J. A., Goedee, J., Raab, J., & Bijl, R. (2017). The nature of
orchestrational work. Public Management Review, 19, 342-360.
Baumgartner, M. (2015). Parsimony and causality. Quality &
Quantity, 49, 839-856.
Braumoeller, B. F. (2015). Guarding against false positives in qual-
itative comparative analysis. Political Analysis, 23, 471-487.
Bunger, A. C. (2013). Administrative coordination in nonprofit
human service delivery networks: The role of competition and
trust. Nonprofit and Voluntary Sector Quarterly, 42, 1155-1175.
Cicchetti, D. V. (1994). Guidelines, criteria, and rules of thumb for
evaluating normed and standardized assessment instruments in
psychology. Psychological Assessment, 6, 284-290.
Fiss, P. C. (2011). Building better causal theories: A fuzzy set
approach to typologies in organization research. Academy of
Management Journal, 54, 393-420.
Hug, S. (2013). Qualitative comparative analysis: How inductive
use and measurement error lead to problematic inference.
Political Analysis, 21, 252-265.
Kanter, R. M., & Khurana, R. (2009). Types and positions: The
significance of Georg Simmel’s structural theories for organi-
zational behavior. In P. S. Adler (Ed.), The oxford handbook of
sociology and organization studies: Classical foundations (pp.
291-306). New York, NY: Oxford University Press.
Kettl, D. (2008). The next government of the United States. New
York, NY: W.W. Norton.
Kickert, W. J. M., Klijn, E.-H., & Koppenjan, J. F. M. (1997).
Managing complex networks. London, England: SAGE.
Klein, K. J., Dansereau, F., & Hall, R. J. (1994). Levels issues in
theory development, data collection, and analysis. Academy of
Management Review, 19, 195-229.
Krishna, A. (2002). Active social capital: Tracing the roots of develop-
ment and democracy. New York, NY: Columbia University Press.
Krogslund, C., Choi, D. D., & Poertner, M. (2015). Fuzzy sets on
shaky ground: Parameter sensitivity and confirmation bias in
fsQCA. Political Analysis, 23, 21-41.

Lemaire 17
Kumar, N., Stern, L. W., & Anderson, J. C. (1993). Conducting
interorganizational research using key informants. Academy of
Management Journal, 36, 1633-1651.
Lemaire, R. H., & Provan, K. G. (2018). Managing Collaborative
Effort: How Simmelian Ties Advance Public Sector Networks.
American Review of Public Administration, 48, 379-394.
Lipset, S. M., Trow, M. A., & Coleman, J. S. (1956). Union democ-
racy. New York, NY: Free Press.
Macciò, L., & Cristofoli, D. (2017). How to support the endur-
ance of long-term networks: The pivotal role of the network
manager. Public Administration, 95, 1060-1076. doi:10.1111/
padm.12349
Marsden, P. V. (1990). Network data and measurement. Annual
Review of Sociology, 16, 435-463.
McGuire, M. (2002). Managing networks: Propositions on what
managers do and why they do it. Public Administration Review,
62, 599-609.
McGuire, M., & Agranoff, R. (2011). The limitations of public
management networks. Public Administration, 89, 265-284.
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of
a feather: Homophily in social networks. Annual Review of
Sociology, 27, 415-444.
Meyers, M. K., Riccucci, N. M., & Lurie, I. (2001). Achieving
goal congruence in complex environments: The case of wel-
fare reform. Journal of Public Administration Research and
Theory, 11, 165-201.
Michels, R. (1915). Political parties: A sociological study of the
oligarchical tendencies of modern democracy (E. Paul & C.
Paul, Trans.). New York, NY: The Free Press.
Milward, H. B., & Provan, K. G. (2006). A manager’s guide to choosing
and using collaborative networks. Washington, DC: International
Business Machines Center for The Business of Government.
Narayan, D. (2000). Voices of the poor. New York, NY: Oxford
University Press.
O’Toole, L. (2015). Networks and Networking: The public adminis-
trative agendas. Public Administration Review, 75, 361-371.
Paquin, R. L., & Howard-Grenville, J. A. (2013). Blind dates and
arranged marriages: Longitudinal processes of network orches-
tration. Organization Studies, 34, 1623-1653.
Provan, K. G., Fish, A., & Sydow, J. (2007). Interorganizational
networks at the network level: A review of the empirical litera-
ture on whole networks. Journal of Management, 33, 479-516.
Provan, K. G., Huang, K., & Milward, H. B. (2009). The evolution of
structural embeddedness and organizational social outcomes in a
centrally governed health and human services network. Journal
of Public Administration Research and Theory, 19, 873-893.
Provan, K. G., & Kenis, P. (2008). Modes of network governance:
Structure, management, and effectiveness. Journal of Public
Administration Research and Theory, 18, 229-252.
Provan, K. G., & Lemaire, R. H. (2012). Core concepts and key
ideas for understanding public sector organizational networks:
Using research to inform scholarship and practice. Public
Administration Review, 72, 638-648.
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 net-
works. Public Administration Review, 61, 414-423.
Putnam, R. D. (1993). Making democracy work. Princeton, NJ:
Princeton University Press.
Putnam, R. D. (2000). Bowling alone: The collapse and revival of
American community. New York, NY: Simon & Schuster.
Raab, J., Lemaire, R. H., & Provan, K. G. (2013). Chapter 10 the
configurational approach in organizational network research.
In P. Fiss, B. Cambré, & A. Marx (Ed.), Configurational the-
ory and methods in organizational research (pp. 225-253).
Bingley, UK: Emerald Group.
Ragin, C. C. (1987). The comparative method. Berkeley: University
of California Press.
Ragin, C. C. (2000). Fuzzy set analysis. Chicago, IL: University of
Chicago Press.
Ragin, C. C. (2008). Redesigning social inquiry: Fuzzy sets and
beyond. Chicago, IL: University of California Press.
Ragin, C. C., & Davey, S. (2016). Fuzzy-set/qualitative compara-
tive analysis. Irvine: Department of Sociology, University of
California.
Saz-Carranza, A., & Ospina, S. M. (2011). The behavioral dimen-
sion of governing interorganizational goal-directed networks:
Managing the unity-diversity tension. Journal of Public
Administration Research and Theory, 21, 327-365.
Seawright, J. (2014). Comment: Limited diversity and the unreli-
ability of QCA. Sociological Methodology, 44, 118-121.
Szreter, S., & Woolcock, M. (2004). Health by association? Social
capital, social theory, and the political economy of public
health. International Journal of Epidemiology, 33, 650-667.
Stevenson, W. B., & Greenberg, D. (2000). Agency and social networks:
Strategies of action in a social structure of position, opposition, and
opportunity. Administrative Science Quarterly, 45, 651-678.
Thiem, A., Baumgartner, M., & Bol, D. (2016). Still lost in transla-
tion! A correction of three misunderstandings between configu-
rational comparativists and regressional analysts. Comparative
Political Studies, 49, 742-774.
Tolbert, P. S., & Hiatt, S. R. (2009). On organizations and oligar-
chies: Michels in the twenty-first century. In P. S. Adler (Ed.),
The Oxford handbook of sociology and organization studies:
Classical foundations (pp. 174-199). New York, NY: Oxford
University Press.
Vangen, S., & Huxham, C. (2012). The tangled web: Unraveling
the principle of common goals in collaborations. Journal of
Public Administration Research and Theory, 22, 731-760.
Venkatraman, N., & Grant, J. H. (1986). Construct measurement
in organizational strategy research: A critique and proposal.
Academy of Management Review, 11, 71-87.
Wasserman, S., & Faust, K. (1994). Social network analysis:
Methods and applications. Cambridge, England: Cambridge
University Press.
Zaheer, A., McEvily, B., & Perrone, V. (1998). Does trust matter?
Exploring the effects of interorganizational and interpersonal
trust on performance. Organization Science, 9, 141-159.
Author Biography
Robin H. Lemaire is an associate professor in the Center for Public
Administration and Policy in the School of Public and International
Affairs at Virginia Tech. Her research focus is on organization
theory and the management of public and nonprofit organizations,
with a particular interest in inter-organizational networks and net-
work analysis.