Interorganizational Coordination in Dynamic Context: Networks in Emergency Response Management1

Full text

1 Earlier version of this paper presented at the NAACSOS 2003 Conference in Pittsburgh, PA
2 I wish to thank Dr. Louise K. Comfort for her helpful comments on earlier versions of this paper. I also thank Michael
Carrigan and Kilkon Ko in helping me conducting the content analysis of The New York Times reports and the FEMA
situation reports for this study. Further, I thank public managers from the Federal Emergency Management Agency, the
U.S. Army Corps of Engineers, the Department of Health and Human Services, and managers of the nonprofit
organizations in New York City who gave their time and thoughtful observations to this research.
http://www.insna.org/Connections-Web/Volume26-2/2.Kapucu.pdf
CONNECTIONS 26(2): 9-10
© 2005 INSNA
Interorganizational Coordination in Dynamic
Context: Networks in Emergency Response
Management1
Naim Kapucu, Ph.D.2
Department of Public Administration, University of Central Florida
This paper addresses the inter-organizational network in response to an extreme event.
Specifically, this paper analyzes interactions among public, private, and nonprofit organ-
izations that evolved in response to the September 11, 2001 terrorist attacks. The research
uses a theoretical framework primarily drawn from dynamic network theory and complex
adaptive systems theory. The study assumes that the increased efficiency that would likely
accrue in mitigation and response to disaster if agencies learned to collaborate more
productively. Organizational analysis techniques were used to identify the major organ-
izations that participated in the response system. The research found that effective response
and recovery require well-coordinated interorganizational networks and trust between
government agencies at all levels and between the public and private sectors.
INTRODUCTION
Public management increasingly takes place in settings of networked actors who necessarily rely on
each other. Building networks of effective action is particularly difficult in dynamic environments.
Yet, current administrative theorists devote relatively little attention to acting effectively in such
situations. The September 11 attacks and their aftermath, along with other major disaster events,
revealed much about institutional responses and collective behavior in extreme disaster conditions,
underscoring what is already known about the social processes that characterize such events, while at
the same time highlighting aspects of disasters that the literature has yet to explore fully.
In drawing lessons from the World Trade Center terrorist attacks in New York City, while the
response activities undertaken by official emergency agencies were crucial, those activities constituted
only part of the picture. Equally significant was the manner in which those agencies interacted with

Interorganizational Networks in Emergency Response Management / Kapucu
10
and obtained support from non-crisis organizations. It has long been recognized that disasters
represent occasions in which the boundaries between organizational and collective behavior are
blurred (Comfort, 1999). This paper discuses how to identify sets of structurally key players,
particularly in the context of networks of organizations in response to September 11, 2001. Specifi-
cally, this paper examines the interactions among organizations that evolved in response to the
September 11, 2001 terrorist attacks on World Trade Center (WTC) in New York City.
The paper addresses the following questions: How did interorganizational coordination among the
organizations evolved in response to the extreme event? What primary organizations were involved in
response to the attack? What were the primary nodes of interaction among the organizations in their
response to the attack? This study assumes that extreme events will lead to greater density of commu-
nication among organizations and less centralized networks. As organizations increase their interac-
tions, they share resources and information. As organizations from different sectors shares informa-
tion and resources, victims in impacted areas will be served better as a result of this collaboration.
METHODOLOGY
The case study and descriptive research methods were used in conducting this research (Yin, 1994).
The case study uses the data from the situation reports from the Federal Emergency Management
Agency (FEMA) and interviews with selected public and nonprofit managers involved in response to
September 11. Data collected through FEMA situation reports and interviews were used to develop a
list of organizations that participated in response operations performed by public, nonprofit, and
private organizations. Situation reports prepared by the Federal Emergency Management Agency
(FEMA, 2001) were used as the official account of organizational operations following the September
11 attacks. This analysis illustrates patterns of communication and information flows among actors.
The actual pattern of interaction reported by the organizations is compared with the designated
responsibilities of the public organizations under the Federal Response Plan (FEMA, 1999). This
comparison illustrates the differences between actual performance and designated roles in the Federal
Response Plan.
I identified all the organizations that participated regardless of any interaction from the FEMA
situation reports. Then, for the network analysis purposes, I identified only the reciprocal organiza-
tional interactions. With these interacting organizations I constructed the matrix for network
analysis. I also reduced the number of organizations by aggregation to construct a manageable
network matrix. The goal was to choose a level of “granularity” that corresponds to the problem at
hand. As in a traditional sociogram, one can aggregate constituents into larger units, if that proven
useful (Pentland, 1999). Second, based on identified actors from the content analysis of the FEMA
situation reports, I used the stratified random sampling method to construct a sample of organizations
that were actively involved in the response system. Third, semi-structured interviews (43) with the
staff, managers, and director of the participant organizations were conducted. Interviews helped to
clarify and expand some of the issues already discovered in the content analyses with regard to
interorganizational networks in response to the attack. The network data collected from my interviews
and the FEMA situation reports were analyzed using the UCINET 6.0 social network analysis
program.
In the network analysis, we are always interested in how an actor is embedded within a structure and
how the structure emerges from the micro-relations between individual parts. The other important
factor for the design of network data has to do with what ties or relations are to be measured for the
selected nodes (Scott, 2000). The other fundamental properties of a social network have to do with
how connected the actors are to one another. Networks that have few or weak connections, or where

Interorganizational Networks in Emergency Response Management / Kapucu 11
some actors are connected only by pathways of great length may display slow response to stimuli.
Networks that have more and stronger connections with shorter paths among actors may be more
robust and more able to respond quickly and effectively. Measuring the number and lengths of
pathways among the actors in a network allow us to index these important tendencies of whole
network (Hanneman, 2001; Wasserman and Faust, 1994). Indeed, most of the basic measures of
networks, measures of centrality, and measures of network groupings and substructures are based on
looking at the numbers and lengths of pathways among actors are used for the analysis of the collected
data. This paper uses the standard network centrality measures of degree, closeness, betweenness and
flow betweenness applied to groups and classes (Everett & Borgatti, 1999). Beside the group centrality,
I also measured cliques, subgroups, similarity and structural equivalence.
THEORETICAL BACKGROUND: INTERORGANIZATIONAL NETWORKS
The research uses a theoretical framework primarily drawn from dynamic network theory and
complex adaptive systems theory (Scott, 2000; Axelrod and Cohen, 1999; Comfort, 1999; Carley, 1999;
Holland, 1995; Wasserman and Faust, 1994; Alter & Hage, 1993; Nohria & Eccless, 1992). In complex
and turbulent environments, organizations frequently develop formal or informal relationships in
order to work together to pursue shared goals, address common concerns, and/or attain mutually
beneficial ends. In recent years, such interorganizational collaboration has become a prominent
aspect of the functioning of many different types of organizations. The number and significance of
collaborative forms of organizing, including interorganizational teams, partnerships, alliances, and
networks, have increased tremendously. The value of effective collaborative relationships as well as the
complexities and challenges they present have been recognized by many researchers, and they
continue to be a frequent subject of scholarly and practitioner-oriented literature (e.g., Linden, 2002;
Powell, 1990; Gray, 1989).
Many researchers have noted that network organizations reflect a qualitatively different form of
governance structure than the bureaucratic hierarchies they are beginning to replace (O’Toole, 1997;
Powell, 1990). In such a environment, understanding the dynamics of the interorganizational net-
works and the patterns of interaction have become urgent matters both for policy makers and those
who seek to understand the policy making process and implementation (Gidron et al, 1992).
In this paper, the term network is used to describe multiple-organizational relations involving
multiple nodes of interactions. A network is group of individuals or organizations who, on a voluntary
basis, exchange information and undertake joint activities and who organize themselves in such a way
that their individual autonomy remains intact. In this definition important points are that the rela-
tionship must be voluntary, that these are mutual or reciprocal activities, and that belonging to the
network does not affect autonomy and independence of the members.
A large body of theory and research about inter-organizational networks now exists to explain how
these relationships emerge, sustain, and create value for the whole society. A particularly interesting
generic type of network involves complex production relationships that benefit from being able to
form and dissolve quickly. The participants therefore wish to protect themselves against opportunistic
exploitation by their partners without having to suffer the delays and costs of formal contracting. This
means that there is some element of trust in the relationship so that post-transaction adjustments to
meet the parties’ needs and interests can be quickly addressed with minimal inter-personal and inter-
organizational resistance (Bardach, 1998).
Public administration scholar Harland Cleveland predicted in 1972 that organizations are moving
toward a more horizontal style of management in which leadership is shared and decisions are often

Interorganizational Networks in Emergency Response Management / Kapucu
12
made on the basis of expertise rather than positions. “The organizations that get things done will no
longer be hierarchical pyramids … they will be systems – interlaced webs of tension in control is loose,
power diffused (Cleveland, 1972, p. 13).
Ackoff (1974) points out that many important current problems are “messes” that actually involve sets
of interconnected problems. The multifaceted nature of these complex problems makes them
extremely difficult to conceptualize and analyze and thus immune to simple solutions (Chisholm,
1998). This interdependence and complexity often require extensive collaboration among different
types and various levels of organizations. Forming and developing inter-organizational networks
represents a response to this interdependence complexity.
Brinton Milward (1996) uses the “hollow state” to characterize what he regards as the increasingly
networked character of public management. Despite the evidence that networks are very important
for public administration, much of the discussion of this subject has been vague (Wamsley et al., 1990;
Provan and Milward, 2001). Helpful starts have been made in other fields. In particular, sociologists
and public choice specialists have developed rich conceptualizations regarding networks (Miller, 1994;
Cook and Whitmeyer, 1992; Ostrom, 1990). Public, nonprofit, and private sector resources may blend
in a variety of ways. These formats permit the mutual leveraging of resources and the blending of
public, nonprofit, and private attributes in ways that might not be possible in more traditional
structural arrangements. This governance perspective is connected to the concern about social capital
and the social underpinnings necessary to effective collaboration.
Networks in the field of public administration and organization theory are primarily based on the
organizations with clearly defined boundaries (Milward, 1996; Chisholm, 1998; Alter and Hage, 1993).
The effect of relations in organizations with permeable boundaries may be different. Modern organ-
izational environments are becoming more complex at an increasing rate (Weick, 2001; Emery and
Trist, 1965; quoted in Scott, 2001; Kauffman, 1993), largely through technical change (Simon, 1996).
This means that uncertainty also increases, and the ratio of externally to internally induced changes
also is increasing. There are instances where changing governance structures and technical changes
may actually reduce uncertainty (Comfort, 1999; Weick, 2001). The interactions of organizations in
a large system can generate greater complexity then the organizations themselves. Moreover,
organizations tend to move toward higher levels of complexity, largely through networks. Organiza-
tions must balance differentiation and coordination to successfully adapt to the rising environmental
complexity. Organizations also must determine the scope of their activities and degree of vertical
integration decisions. Depending on one’s theoretical perspective, these balancing conflicts are either
seen as inefficiencies (rational system) or necessary parts of the negotiation process (natural system)
(Scott, 2001).
Social network analysis is a well-developed and fast-growing area of organizational sociology, and it
provides tools and concepts for analyzing organizations as networks (Wasserman and Faust, 1994).
A meta-matrix, developed by Kathleen Carley (2002), represents a network of interactions that can be
analyzed using the same graph-theoretic techniques that have been applied to networks of individuals
and other entities. Meta-matrix analysis is a useful method in analyzing the structure of interorganiza-
tional response.
INTERORGANIZATIONAL NETWORKS IN EXTREME EVENTS
The dynamics of learning and adaptation, central to the complexities of an ecological system, are
increasingly used as an analogy to the collaborative relations between sectors in network based systems
of governance. Resilient social systems are characterized by reduced failure, measured in terms of lives

Interorganizational Networks in Emergency Response Management / Kapucu 13
lost, damage, and negative social and economic impacts, and reduced time to recovery – that is, more
rapid restoration of the social systems and institutions to their normal, pre-disaster levels of function-
ing. Aaron Wildawsky (1971, p. 77) describes resilience as “the capacity to cope with unexpected
dangers after they have become manifest, learning to bounce back.” The Resilience Multidisciplinary
Center for Earthquake Engineering Research (MCEER) has identified four general properties that can
be applied to all systems and to the elements that comprise systems: robustness (ability to withstand
the forces generated by a hazard agent without loss or significant deterioration of function; resource-
fulness (capacity to apply material, informational, and human resources to remedy disruptions when
they occur); redundancy (the extent to which elements, systems, or other units of analysis exist that
are capable of satisfying the performance requirements of a social unit in the event of loss or
disruption that threaten functionality); and rapidity (the ability to contain loses and restore system or
other units in a timely manner). Organizations can contribute to resilience in a society by incorpora-
tion other emergency response organizations and by integrating volunteers into emergency operations
as appropriate.
Meta Matrix People / Agents Knowledge Resources Tasks Organizations
People / Agents
Relations
Interaction
Network
Who knows
whom Structure
Knowledge
Network
Who knows what
Culture
Capabilities
Network
Who has
what resource
Capital
Assignment
Network
Who does what
Jobs
Work Network
Who works
where
Demography
Knowledge
Relation
Information
Network
What informs
what
Data
Skills Network
What knowledge
is needed to use
what resource
Technology
Needs Network
What is needed
to do that task
Needs
Competency
Network
What knowledge
is where
Culture
Resources
Relation
Substitution
Network
What resources
can be substituted
for which
Requirements
Network
What resources
are needed
to do that task
Needs
Capital
Network
What resources
are where
Resources
Tasks Relation
Precedence
Network
Which task must
be done before
which
Operations
Sectoral
Network
What tasks are
done where
Niche
Organizations
Relation
Inter-
Organizational
Network
Which
organization
works with witch
Partnerships
Source: Adapted from Kateen M. Carley 2002.
Figure 1. Meta Matrix
Extreme events are occurrences that are notable, rare, unique, and profound, in terms of their
impacts, effects, or outcomes. When extreme events occur at the interface between natural, social and
human systems, they are often called “disasters” (Red Cross, 2001). Quarantelli and Dynes (1977)

Interorganizational Networks in Emergency Response Management / Kapucu
14
define disaster as the disruption to society after the “event.” Everybody is affected in extreme events
and individuals and single organizations cannot prevent the harm caused by the event. In extreme
events standard procedures cannot be followed and they require dynamic system to adapt to
unanticipated and rapidly changing conditions. The September 11 2001 terrorist attack is an example
of an extreme event with significant impact upon humanity. Extreme events trigger greater density of
communication and interaction among organizations that stimulates collective action. A critical
aspect of this process is the formation of new and or stronger networks among multi-sector organiza-
tions.
1. Interorganizational networks in emergencies can play an important role in facilitating
the flow of information across organizational boundaries. Following are the principal
pathways through which social networks enhance performance of organizational
networks:
2. Social networks increase interaction among organizations that can lead to development
of trust which reduce transaction costs (Coleman, 1990),
3. Social networks spread risk by providing individual members with sources of support
during times of trouble, and allow the group as a whole to engage in overall higher levels
of risk-taking (Fukuyama, 1995),
4. Social networks facilitate the rapid dissemination of information among members and
reduce the asymmetries of information that can otherwise discourage profitable transac-
tions,
5. Social capital improves access to resources among network members,
6. Social networks allow members to solve collective action problems more easily with less
fear of defection and free riding (Ostrom, 1990)
The capacity of a society to understand and manage extreme events depends on its ability to
understand, anticipate, prepare for, and respond to them (Comfort, 1999). Moreover, increasing
organizational and technological interconnectedness may create more possibilities of multiorganiza-
tional partnerships for the surge of an extreme event. The WTC disaster illustrates how in disaster
settings high levels of cooperation and collaboration among organizational and community actors can
co-exist. Communities responding to disasters are seen as coping collectively with shared pain, loss,
and disruption and as temporarily suspending ongoing conflicts and disagreements in the interest of
meeting urgent needs and beginning the recovery process. Trustworthiness and social capital can,
especially, play an important role in extreme events within which there is no clear policy or guidelines
available to the participant organizations and individuals (Axelrod and Cohen, 1999).
INTERORGANIZATIONAL COORDINATION
Under the Federal Response Plan (FEMA, 1999), eight federal agencies in addition to FEMA play lead
roles in disaster operations, with 25 federal agencies assigned responsibilities under twelve specified
emergency support functions. The lead agencies include the Departments of Transportation (DOT),
National Communications Service (NCS), Defense (DOD), Agriculture (USDA), Health and Human
Services (HHS), Housing and Urban Development (HUD), Environmental Protection Agency (EPA),
and the General Accounting Office (GAO). Two departments have dual emergency support
functions. The USDA has the primary support function for firefighting, carried out by its sub-unit,
the U.S. Forest Service (USFS), as well as for food. FEMA is responsible for information management,
as well as urban-search-and-rescue operations. The American Red Cross (ARC) is designated as the
lead agency for mass care (Figure 2).

Interorganizational Networks in Emergency Response Management / Kapucu 15
Immediately after the attack, an intensive coordinated effort was begun by federal, state, and city
government, along with volunteer agencies, in the search, rescue, recovery, and identification of the
victims. Extensive assistance was directed toward the needs of victims and their families. While the
physical damage was concentrated in a relatively small area, the economic and social effects were
pervasive citywide. The pervasive threat of the attack created a situation of shared risk, that is, the risk
of the attack is shared by all members of society. This condition of shared risk offers an important
alternative perspective on response operations for extreme events. As the risk is shared, so is the
responsibility for assessing and responding to that threat (Comfort, 1999). Recognition of shared
responsibility immediately broadens the task of confronting the threat with organizations outside the
public sector. Individuals, private and nonprofit organizations become resources for this collective
response operation (Kapucu & Comfort, 2002).
Coordinating the activities of non-crisis organizations is a complex and difficult task. Public
managers are reluctant to rely upon nonprofit voluntary organizations during extreme events.
“Because they distrust the intentions of the volunteers, lack confidence in the volunteers skills and
resources, fear that volunteer may endanger themselves or others, are concerned that volunteer may
get into way of professional responders, and fear that there may be legal liability for volunteers’
actions” (Waugh, 2000; p. 47). As noted in Waugh (2000) that emergency management is the
quintessential government role. FEMA is the lead federal agency for responding to disasters and may
link with nonprofit organizations. According to FEMA regulations, in the event of a residentially
declared disaster or emergency, such as 9/11, FEMA is required to coordinate relief and assistance
activities of federal, state, and local governments; the American Red Cross; the Salvation Army; as well
as other voluntary relief organizations that agree to operate under FEMA’s direction. Disaster
response and recovery roles cross-cut 28 Federal agencies and the Red Cross, which participates with
FEMA in disaster operations guided by the Federal Response Plan (1999).
PATTERNS OF INTERORGANIZATIONAL NETWORKS
In this section of the paper, I measure degree, closeness, betweenness, and flow betweenness centrality
and clique and sub-groups (n-clique, c-clans, k-plexes). There are many measures of actor position
and overall network structure that are based on whether there are pathways between actors, the length
of the shortest pathway between two actors, and the numbers of pathways between actors. I employed
UCINET (Version 6.0) for the network analysis of the data. UCINET is a comprehensive program for
the analysis of social networks and other proximity data. The program contains several network
analytic routines and general statistical and multivariate analysis tools.
Size of the network is critical to the structure of organizational interactions because of the limited
resources and capacities that each organization has for building and maintaining networks. Usually,
the size of a network is indexed simply by counting the number of nodes. In any network there are
(k * k-1) unique ordered pairs of actors, where k is the number of actors. It follows from this that the
range of logically possible social structures increases (complexity) exponentially with size. If the size
of the network increases, the complexity of the relationships also increases.
The graph from the Federal Response Plan (FRP) is represented in Figure 2 below. We can perceive
a number of things in simply looking at the graph. There are a limited number of actors (28), and all
of them are connected very well in a very orderly manner as we would not expect from any complex
organizational networks. There appear to be some differences among the actors in how connected
they are (compare actors HUD and USDA, for example). If we look closely, we can see that some
actor’s connections are likely to be reciprocated (that is, if A shares information with B, B also shares
information with A) but some other actors are more likely to be senders than receivers of information.

Interorganizational Networks in Emergency Response Management / Kapucu
16
Figure 2. Networks in FEMA Emergency Response Plan
As a result of the variation in how connected organizations are, and whether the ties are reciprocated,
some actors may be at quite some “distance” from other actors. There appear to be groups of actors
who differ in this regard. For example, FEMA, HHS, USDA, ARC, and DOT that seem to be in the
center of the action while HUD, DOC, and TVA, seem to be more peripheral.
The graph from the FEMA situation reports is presented in Figure 3 below. We perceive a number of
things by simply looking at the graph as well. There are a limited number of actors here (41), and all
of them are “connected.” But, clearly not every possible connection is present, and there are
“structural holes.” There appear to be some differences among the actors in how connected they are
as usual. If we compare FEMA and NYCEMO with HUD and GSA for example, we can easily see the
difference. FEMA and NYCEMO are in the center of the activities. On the other hand, HUD and GSA
are not very central or well connected to other organizations. If we look closely, we can see that some
actor’s connections are likely to be reciprocated in this network but some others are not. FEMA,
NYCEMO, NYC government and mayor, and HHS seem to be in the center of the action; HUD, DOJ,
OSHA, FAA seem to be more peripheral in the network.
Findings from content analysis of the FEMA situation reports indicate that interactions were limited
and occurred primarily between organizations of similar types. For example, public organizations
tended to interact most frequently with other public organizations from the same jurisdiction; private
organizations with other private organizations; nonprofit organizations with other nonprofit
organizations. Interactions were infrequently reported across jurisdictional lines.

Interorganizational Networks in Emergency Response Management / Kapucu 17
Figure 3. Organizational Network -FEMA Situation reports
Group Centrality: Major Players
With larger populations or more connections, however, graphs may not be very helpful. Looking at
a graph can give a good intuitive sense of what is going on, but our descriptions of what we see are
imprecise. To get more precise, and to use computers to apply algorithms to calculate mathematical
measures of graph properties, it is necessary to work with the adjacency matrix and more complicated
calculations instead of the graph.
One of the methods used to understand networks and their participants is to evaluate the location of
actors in the network. Measuring the network location is finding the centrality of an actor. These
measures help determine the importance of a node in the network. I use centrality measures as a basic
tool for identifying key organizations in the response system network (Everett & Borgatti, 1999). The
centrality approaches (degree, closeness, and betweenness) describe the locations of individual
organization in terms of how close they are to the center of the action in a network.
Group degree centrality is defined as the number of non-group nodes that are connected to group
members (Everett & Borgatti, 1999). Actors who have more ties to other actors may have access to,
and be able to call on, more of the resources of the network as a whole. UCINET is used to do the
counting, and some additional calculations and standardizations that were suggested by Linton
Freeman (1979).

Interorganizational Networks in Emergency Response Management / Kapucu
18
Table 1. Freeman’s Degree Centrality Measures
Degree NrmDegree
2
34
41
13
27
35
39
3
24
10
FEMA
NY City Govt/ Mayor
Nonprofit Orgs
U.S. Military Armed Forces
NY State Govt
NYC OEM
Private Orgs
HHS
US Congress
USACE
329.000
87.000
58.000
42.000
42.000
32.000
32.000
28.000
22.000
21.000
822.500
217.500
145.000
105.000
105.000
80.000
80.000
70.000
55.000
52.500
Freeman’s degree centrality measures show that FEMA (actor #2) and New York City Govern-
ment/Mayor (actor #34) have the greatest degree, and can be regarded as the most influential in the
response operation. Nonprofit Organizations (actor #41) and the US Military and Armed Forces
(actor #13) are followed by New York State Government (actor #27). The similarity between the two
results, Freeman’s degree centrality measures and visual representation of the data in graph, can easily
be captured. That other organizations share information with these five would seem to indicate a
desire on the part of others to participate in network in response operations.
The following is the result from the degree group centrality calculated by UCINET for the optimal
groups in network (Table 2). FEMA, HHS, New York City Government, American Red Cross,
USACE, and nonprofit organizations were identified again as central organizations in the network.
Table 2. Degree Group Centrality
Observed # reached=41.000 (100.0%)
Group Members:
Observed no. reached = 30.000 (88.2%)
Group Members:
2 FEMA
3 HHS
6 DOT
25 USAR
27 NY State Government
28 CT Dpt of Health
37 NYFD
40 ARC
41 Nonprofit Orgs
3 FEMA
6 NYC Govt/mayor
7 Nonprofit Orgs
8 NY & NJ Port Authority
14 City Harvest, NY
18 USDA Forest Service
20 Salvation Army
21 Southern Baptist Kitchens
24 Catholic Charities of NY
Source: FEMA Situation Reports Source: Interviews
Closeness Centrality
Degree centrality measures might be criticized because they only take into account the immediate ties
that an actor has, rather than indirect ties to all others. One actor might be tied to a large number of
others, but these others might be rather disconnected from the network as a whole. In this case, the
actor could be quite central, but only in a local neighborhood (Wasserman and Faust, 1994). However,
closeness centrality emphasizes the distance of an actor to all others in the network by focusing on the
geodesic distance from each actor to all others. The sum of these geodesic distances for each actor is
the “farness” of the actor from all others. We can convert this into a measure of nearness or closeness
centrality by taking the reciprocal (one divided by the farness) and normalizing it relative to the most
central actor. Here are the UCINET results for closeness:

Interorganizational Networks in Emergency Response Management / Kapucu 19
Table 3. Closeness using FEMA situation reports data
Fairness nCloseness
2
13
34
3
35
27
20
10
18
37
30
29
6
16
FEMA
U.S. Military Armed Forces
NY City Govt/ Mayor
HHS
NYC OEM
NY State Govt
NCS
USACE
EPA
NYFD
NJ OEM
NJ Dpt of Health
DOT
HUD
42.000
67.000
67.000
68.000
69.000
70.000
70.000
72.000
73.000
73.000
81.000
81.000
93.000
98.000
95.238
59.701
59.701
58.824
57.971
57.143
57.143
55.556
54.795
54.795
49.383
49.383
43.011
40.816
Actor #2 (FEMA) is the closest, or most central, actor using this method, because the sum of FEMA’s
geodesic distances to other actors (a total of 41) is the least. Four other actors US Military Armed
Forces – USACE (actor #13), New York City Government/Mayor (actor #34), Health and Human
Services (actor # 3), and New York City Emergency Management Office (actor #35) are nearly as close
and thus are highly central organizations, HUD (actor #16) and the Department of Transportation
(DOT) (actor #6), on the other hand, have the greatest farness.
Betweenness Centrality
Suppose that FEMA wants to exchange resources and information and work with NYCEMO. FEMA
must go through an intermediate agency, NYC Government/Mayor for example. According to the
strict rules of bureaucratic hierarchy, FEMA must forward the request through another governmental
agency. The intermediate agency could delay the request, or even prevent the request from getting
through. This gives a coordinating position to the organization who lie “between” the two organiza-
tions with respect to others. FEMA might use other agencies or channels to work with NYCEMO.
Having more than one channel makes FEMA less dependent, a more central, and as more independ-
ent actor. Betweenness centrality views an actor as being in a favored position to the extent that the
actor falls on the geodesic paths between other pairs of actors in the network. UCINET, it is easy to
locate the geodesic paths between all pairs of actors, and to count up how frequently each actor falls
in each of these pathways. The results from UCINET are:
Table 4. Betweenness
Betweenness nBetweenness
02
34
37
13
20
27
3
18
41
35
5
39
15
10
7
40
FEMA
NY City Govt/ Mayor
NYFD
U.S. Military Armed Forces
NCS
NY State Govt
HHS
EPA
Nonprofit Orgs
NYC OEM
CDC
Private Orgs
DMAT
USACE
USDA
ARC
652.629
116.781
90.183
65.600
52.167
46.360
45.460
39.943
26.667
21.250
13.167
13.110
8.000
6.443
2.743
1.500
41.835
7.486
5.781
4.205
3.344
2.972
2.914
2.560
1.709
1.362
0.844
0.840
0.513
0.413
0.176
0.096

Interorganizational Networks in Emergency Response Management / Kapucu
20
It can be seen that there is a great deal of variation in actor betweenness. FEMA (actor #2) andNY City
Government/Mayor (actor #34) appear to be relatively a good bit more central than others by this
measure.
Flow Betweenness: Dynamics of Interorganizational Networks
The betweenness centrality measure I examined above characterizes actors as having positional
advantage to the extent that they fall on the shortest pathway between other pairs of actors. The idea
is that actors who are “between” other actors, and on whom other actors must depend to conduct
exchanges, will be able to translate this central intermediary role into power.
If the two actors want to have a network relationship, but the geodesic path between them is blocked
by an unwilling organization, and if there is another pathway, the two actors are likely to use it, even
if it is longer and less efficient. The flow approach to centrality expands the notion of betweenness
centrality. It assumes that actors will use all pathways that connect them to others proportionally to
the length of the pathways. Betweenness is measured by the proportion of the entire flow between two
actors that occurs on paths which connect them. For each actor, then, the measure adds up how
involved that actor is in all of the flows between all other pairs of actors (Wasserman and Faust, 1994).
Since the magnitude of this index number would be expected to increase with the size of the network
and with network density, it is useful to standardize it by calculating the flow betweenness of each
actor in ratio to the total flow betweenness that does not involve the actor (Everett & Borgatti, 1999).
Table 5. Flow betweenness
FlowBet nFlowBet
1
2
3
4
5
6
7
8
9
10
FEMA
HHS
DOD
CDC
DOT
USDA
GSA
DOE
USACE
SBA
795.727
97.754
2.167
27.294
0.000
4.497
6.167
0.000
7.176
0.000
51.008
6.266
0.139
1.750
0.000
0.288
0.395
0.000
0.460
0.000
By this more complete measure of betweenness centrality, FEMA (actor #2), U.S. Military and Armed
Forces (actor #13), HHS (actor #3), and New York City Office of Emergency Management (actor # 35)
are clearly the most important mediators. New York State Emergency Management Office (NYSEMO)
(actor #31) and American Red Cross (ARC) (actor #40), who were fairly important when we
considered only geodesic flows, appear to be rather less important by this calculation. While the
overall picture does not change a great deal, the elaborated definition of betweenness does give us a
somewhat different impression of who is most central in this network.
Cliques and Sub-groups: Groupings of Organizational Networks
Networks are also built up out of the combining of dyads and triads into larger, but still closely
connected sub-structures. Many of the approaches to understanding the structure of a network
emphasize how dense connections are compounded and extended to develop larger cliques or sub-
groupings (Wasserman and Faust, 1994). A clique is simply a sub-set of actors who are more closely
tied to each other than they are to actors who are not part of the group. This view of social networks
focuses attention on how connection of large networks structures can be built up out of small and
tight components.

Interorganizational Networks in Emergency Response Management / Kapucu 21
Divisions of actors into cliques is a very important aspect of networks in understanding how the
network as a whole is likely to behave. For example, suppose the actors in one network form two non-
overlapping cliques; and, suppose that the actors in another network also form two cliques, but that
the memberships overlap (some organizations are members of both cliques). Where the groups
overlap, it can be expected that conflict between them is less likely than when the groups do not
overlap (Hanneman, 2001). Where the groups overlap, resources can be mobilized and shared
effectively across the entire network; where the groups do not overlap, resource sharing may occur in
one group and not occur in others.
Knowing how an organization is embedded in the structure of groups within a net may also be
important to understanding its behavior. For example, some organizations may act as “bridges”
between groups (boundary spanners). Other organizations may have all of their relationships within
a single clique (locals). Some actors may be part of a tightly connected group, while others are
completely isolated from this group. Such differences in the ways that organizations are embedded in
the structure of groups within in a network can have profound consequences for the ways that these
actors see the network, and the behaviors that they are likely to practice to sustain or dysfunction the
colloboration.
Table 6. Cliques
1:
2:
3:
4:
5:
6:
7:
8:
9:
FEMA NCS NY State Govt NY City Govt/ Mayor Verizon
FEMA EPA NY State Govt NY City Govt/ Mayor
FEMA HHS NY State Govt NY City Govt/ Mayor NYC OEM
The President FEMA NY State Govt NY City Govt/ Mayor
FEMA DOD NY City Govt/ Mayor NYC OEM
FEMA CDC EPA NY City Govt/ Mayor
FEMA HHS CDC NY City Govt/ Mayor
FEMA USDA NY City Govt/ Mayor NYC OEM ARC
FEMA USACE EPA NY City Govt/ Mayor
Table 6 suggests a number of things: FEMA, Verizon, HHS, NY City Government/Mayor, NYCEMO,
USDA, and U.S. Military Armed Forces appear to be in the middle of the action in the sense that they
are members of many of the groupings, and serve to connect them, by co-membership.
Figure 4. Hierarchical Clustering of Equivalence Matrix
1 1 1 1 1 1 2 1 1 1 1 2 2
Level 8 9 5 6 1 2 5 8 7 0 1 2 4 0 9 3 4 3 1 6 7 2
----- - - - - - - - - - - - - - - - - - - - - - -
4.000 . . . . . . . . . . . . . . . . . XXX . . .
3.000 . XXX . XXX XXX . . XXXXX . . XXX XXX XXX .
2.667 . XXXXX XXX XXXXX . XXXXX . . XXXXXXX XXXXX
2.222 . XXXXX XXX XXXXX . XXXXX . . XXXXXXXXXXXXX
2.178 . XXXXX XXXXXXXXX . XXXXX . . XXXXXXXXXXXXX
1.915 . XXXXX XXXXXXXXX . XXXXX . XXXXXXXXXXXXXXX
1.810 . XXXXX XXXXXXXXXXX XXXXX . XXXXXXXXXXXXXXX
1.641 . XXXXX XXXXXXXXXXX XXXXX XXXXXXXXXXXXXXXXX
1.507 . XXXXX XXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXX
1.299 . XXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXX
1.249 . XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
1.057 XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
We see that actors #2 and #21 are joined first as being close because they share 4 clique memberships
in common. At the level of sharing only three clique memberships in common, actors #9, # 15, # 11,
# 12, # 5, # 8, # 1, # 4, # 13, # 14, 3 6, and # 7 join the core. If we require only one clique membership
in common to define group membership, then all actors are joined except # 18.

Interorganizational Networks in Emergency Response Management / Kapucu
22
3 Interview with NY & NJ Port Authority, 11/28/2003
4 The Tri-State Metropolitan Region consists of nearly 20 million people living in Connecticut, New Jersey, and New York.
Much of what was observed on 9/11 and in the days and weeks that followed in New York City’s
massive destruction and social disruption, was a complex organized response. The immediate impact
area was evacuated rapidly and in an orderly manner. After the collapse of the towers, the absence of
panic saved numerous lives. Assisted by emergency workers, occupants of the World Trade Center
and people in the surrounding area helped one another to safety, even at great risk to themselves.
Prior experience with the 1993 Trade Center bombing had led to significant learning among
organizational tenants and occupants of the Twin Towers, and planning and training contributed to
their ability to respond in an adaptive fashion to highly ambiguous and threatening conditions.3
It has long been recognized by academics that disasters represent occasions in which the boundaries
between organizational and collective behavior are blurred. Local capabilities are enhanced through
the active involvement of nonprofit organizations. In the World Trade Center disaster, all these
organizational patterns observed at Ground Zero: NYC emergency response organizations were
assisted by counterpart organizations from throughout the tri-state region4 and ultimately from
communities around the country, by nonprofit organizations offering whatever assistance they could.
Collective behavior brings charitable organizations with their needed resources to disaster areas while
simultaneously creating substantial management challenges.
CONCLUSION
The insight of both network and complexity theories can help constructs interorganizational networks
and help us understand their workings. Multi-sectoral collaboration involves creating new forms of
relationships among organizations. In order to foster linkages and the trust that would enable
accelerating coordination in emergency management response operations, the government should
provide incentives and information to promote multi-sectoral colloborations.
The idea of interdependency has long been at the heart of organization design in complex environ-
ments. Despite the richness of theoretical developments, there has been relatively little formal
investigation as to the extent to which interdependency among organizations can influence organiza-
tional adaptation over time in dynamic environments. This research represents a modest step towards
understanding how organizational design can be used to help track the interorganizational coordina-
tion in emergencies.
Effective response and recovery operations require colloborations and trust between government
agencies at all levels and between the public and nonprofit sectors. Ongoing collaboration raises trust,
and the importance of broad collaboration among various governmental levels and between govern-
ment, the private sector, the nonprofit sector, and the public cannot be overemphasized. In response
to 9/11 a resilient emergency response was achieved through integrating the resources and capacity
of emergency response organizations with other governmental agencies, private, and nonprofit
organizations.

Interorganizational Networks in Emergency Response Management / Kapucu 23
REFERENCES
Ackoff, Russell L. 1974. Redesigning the Future: A Systems Approach to Societal Problems. New York:
Wiley.
Alter, Christine and Jerald Hage. 1993. Organizations Working Together. Newbury Park: Sage.
Axelrod, Robert and Cohen, Michael D. 1999. Harnessing Complexity: Organizational Implications
of a Scientific Frontier. New York: the Free Press.
Barabâasi, Albert-Laszlâo, 2002. Linked: The New science of Networks. Cambridge, MA: Perseus
Publishing.
Bardach, Eugene 1998. Getting Agencies to Work Together: The Practice and Theory of Managerial
Craftsmanship. Washington, DC: The Brookings Institution.
Borgatti, S.P., Everett, M.G. and Freeman, L.C. 2002. UCINET 6.0VERSION for Windows: Software
for Social Network Analysis. Harvard: Analytic Technologies.
Carley, Katleen M. 1999. “On the Evolution of Social and Organizational Networks.” In Steven B.
Andrews and David Knoke (Eds.) Vol. 16, Special Issue of Research in the Sociology of Organiza-
tions, on “Networks In and Around Organizations.” JAI Press, Inc. Stamford, CT, pp. 3-30.
Chisholm, Rupert F. 1998. Developing Network Organizations: Learning from Practice and Theory.
New York: Addison-Wesley.
Cleveland, H. 1972. The Future Executive. New York: Herper Collins.
Coleman, James. 1990. Foundations of Social Theory. Cambridge, Mass.: Belknap, Harvard University
Press.
Comfort, Louise. 1999. Shared Risk: Complex Systems in Seismic Response. New York: Prgamon Press.
Cook, K. S. and J. M. Whitmeyer. 1992. “Two Approaches to Social Structure: Exchange Theory and
Network Analysis.” Annual Review of Sociology 18: 109-127.
Everett, M. G., & Borgatti, S. P. 1999. “The centrality of groups and classes.” Journal of Mathematical
Sociology. 23(3): 181-201.
Federal Emergency Management Agency (FEMA) Situation Reports. September 11- October 04, 2001.
Federal Emergency Management Agency (FEMA), 1999. Federal Response Plan. Available online at
http://www.fema.gov/rrr/frp/.
Freeman, Linton C. 1979 Centrality in social networks: I. Conceptual clarification. Social Networks 1:
215-239.
Fukuyama, Francis. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York: The Free
Press.
Gidron, Benjamin; Kramer, Ralph M.; and Salamon, Lester M. 1992. Government and the Third
Sector; Emerging Relationships in Welfare State. San Francisco: Jossey- Bass Publishers.
Gray, Barbara. 1989. Collaborating. San Francisco: Jossey-Bass.
Hanneman, Robert A. 2001. Introduction to Social Network Methods . Unpublished Textbook
(available online at www.faculty.ucr.edu/~hanneman)
Holland, J. 1995. Hidden Order: How Adaptation Builds Complexity. Reading, MA: Addison Wesley
Publishing, Company.

Interorganizational Networks in Emergency Response Management / Kapucu
24
Kapucu, Naim and Louise K. Comfort. 2002. “Inter-organizational Coordination in Extreme Events:
Public-nonprofit Partnerships in Dynamic Contexts,” Paper presented at APPAM Conference,
November 2002. Dallas, TX
Kauffman, Stuart A. 1993. The Origins of Order: Self-Organization and Selection in Evolution. New
York: Oxford University Press.
Linden, M. Russell. 2002. Working across Boundaries: Making Collaboration Work in Government and
Nonprofit Organizations. San Francisco: Jossey-Bass.
MCEER. 2002. Management of Complex Civil Emergencies and Terrorism ! Resistant Civil Engi-
neering Design. Proceedings from the MCEER Workshop on Lessons from the World Trade
Center Terrorist Attack.
Milward, H. Brinton, ed. 1996. Symposium on “The Hollow State: Capacity, Control and Performance
in Interorganizational Settings.” Journal of Public Administration Research and Theory. 6: 193-313.
Nohria, N. and R. Eccles, Eds. 1992. Networks and Organizations. Cambridge, MA: Harvard Business
School Press.
New York Times, September 12 – October 6, 2001.
Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for Collective Action .
Cambridge: Cambridge University Press.
O’Toole, L. J., Jr. 1997. “The implications for democracy in a networked bureaucratic world.” Journal
of Public Administration Research and Theory, 7 (3), 443-459.
Pentland, Brian T. 1999. “Organizations as Networks of Actions.” In Edited by Joel A. C. Baum and
Bill McKelvey. Variations in Organization Science: In Honor of Donald T. Campbell. Thousand
Oaks, CA: Sage Publications.
Powell, W. W. 1990. “Neither market nor hierarchy: Network form of organization.” In B. M. Staw
and L. L. Cummings (eds.), Research in Organizational Behavior, Vol. 12. Greenwich, CT: JAI
Press, 295-336.
Quarantelli, E. L. and Dynes. 1977. Disasters: Theory and Research. Beverly Hills, CA: Sage.
Red Cross. 2001. World Disaster Report. Geneva: International Federation of Red Cross and Red
Crescent Societies. (www.ifrc.org).
Scott, John. 2000. Social Network Analysis. Thousand Oak, CA: Sage Publications.
Scott, W. R. 2001. Organizations: Rational, Natural, and Open Systems . Englewood Cliffs, NJ:
Prentice-Hall.
Simon, Herbert A. 1996. The Sciences of the Artificial. Cambridge, MA: M.I.T. Press.
Wasserman, Stanley and Katherine Faust.1994. Social Network Analysis: Methods and Applications .
New York and Cambridge: Cambridge University Press.
Waugh, William L. Jr. 2000. Living with Hazards Dealing with Disasters: An Introduction to Emergency
Management. Armonk, NY: ME Sharpe.
Weick, Karl E. 2001. Making Sense of the Organization. Oxford, Massachusset: Blackwell Business.
Wildavsky, Aaron. 1991. Searching for Safety. New Brunswick, NJ: Transaction Publishers.
Yin, Robert. 1994. Case Study Research: Design, Methods. 2nd ed. Thousand Oaks, CA: Ssge Publica-
tions.