Upload
maria-e
View
212
Download
0
Embed Size (px)
Citation preview
This article was downloaded by: [Aston University]On: 24 January 2014, At: 10:43Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Society & Natural Resources: AnInternational JournalPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/usnr20
Does Community-Based CollaborativeResource Management Increase SocialCapital?Cheryl L. Wagner a & Maria E. Fernandez-Gimenez aa Department of Forest, Rangeland, and Watershed Stewardship ,Colorado State University , Fort Collins, Colorado, USAPublished online: 19 Mar 2008.
To cite this article: Cheryl L. Wagner & Maria E. Fernandez-Gimenez (2008) Does Community-Based Collaborative Resource Management Increase Social Capital?, Society & Natural Resources: AnInternational Journal, 21:4, 324-344, DOI: 10.1080/08941920701864344
To link to this article: http://dx.doi.org/10.1080/08941920701864344
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.
This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Does Community-Based Collaborative ResourceManagement Increase Social Capital?
CHERYL L. WAGNER AND MARIA E. FERNANDEZ-GIMENEZ
Department of Forest, Rangeland, and Watershed Stewardship,Colorado State University, Fort Collins, Colorado, USA
One potential benefit of collaboration is increased social capital, defined as trust,norms of reciprocity, and networks. Social capital may improve a group’s abilityto collaborate, manage risk, innovate, and adapt to change. We used mail surveysof group participants and key informant interviews to assess whether social capitalchanged over time in eight collaborative groups in northwest Colorado. The majorityof social capital measures increased over time in all groups. However, social capitalbetween agencies and group members increased more than social capital amongother participants for most groups. Interviews revealed that commitment and conti-nuity; understanding, empathy, and respect; transparency; and dependability andpredictability are important mechanisms for building social capital in collaborativesettings.
Keywords collective action, natural resource agencies, norms of reciprocity,trust, values
Community-based collaborative resource management (CBCRM) is on the riseacross the United States (Gray et al. 2001; Weber 2003; Wondolleck and Yaffee2000). This is especially true in the West, where conflict over the Old West’s historyof natural resource extraction is increasingly at odds with the New West’s interest inrecreation and conservation (Brick et al. 2001). We define CBCRM as groups ofdiverse stakeholders who convene voluntarily to work on natural resource policy,planning, or management issues specific to a particular location.
Broadly defined, social capital is relationships of trust, norms of reciprocity, andnetworks among individuals that can be drawn upon for individual or collectivebenefit (Coleman 1990; Putnam 1993). The crux of social capital is that relationships‘‘constitute an important asset, one that can be called on in a crisis, enjoyed for itsown sake, and leveraged for material gain’’ (Woolcock and Narayan 2000, 226).Social capital is considered a resource because individuals or groups that possesssocial capital can use it to acquire other resources (Coleman 1990; Bourdieu 1986).
In the context of CBCRM, social capital is an asset that groups or stakeholderscan use to obtain the results they seek and accomplish goals that are otherwise unat-tainable (Putnam 1993). Social capital is important because it can provide access to
Received 6 August 2006; accepted 4 December 2007.Address correspondence to Maria E. Fernandez-Gimenez, Department of Forest,
Rangeland, and Watershed Stewardship, Colorado State University, Fort Collins, CO80523-1472, USA. E-mail: [email protected]
Society and Natural Resources, 21:324–344Copyright # 2008 Taylor & Francis Group, LLCISSN: 0894-1920 print/1521-0723 onlineDOI: 10.1080/08941920701864344
324
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
other forms of capital such as financial capital (e.g., grants, funding), human capital(e.g., information, skills, scientific expertise), or natural capital (e.g., forage, habitat,water), and because it may improve a group’s or community’s ability to come upwith innovative solutions to problems, manage risk, and adapt to change (Adger2003; Olsson et al. 2004).
Social capital may act as both an input to and an outcome of collaborativeprocesses. Social capital facilitates the cooperation necessary for collective action,and thus can be seen as an input (Adger 2003; Leach and Sabatier 2005; Mullenand Allison 1999). Because it facilitates collaboration, social capital increases thelikelihood that a collaborative effort will succeed (Leach and Sabatier 2005; Schuettand Selin 2002). Trust and norms of reciprocity among participants increase thewillingness of parties to cooperate because ‘‘people have the confidence to invest incollective activities, knowing that others will also do so’’ (Pretty and Ward 2001,210). Because collective action is a mechanism that may build social capital, socialcapital can also be seen as a potential outcome of collaboration (Leach 2002; Ostrom1997; Putnam 1993). In other words, social capital and collective action are mutuallyreinforcing and their interaction creates a positive feedback loop.
Collective action may build social capital because social capital is ‘‘created andmaintained by the very process of working together’’ (Ostrom 1997, 178). Further,unlike other forms of capital, social capital increases with use (Bourdieu 1986).The forum for interaction provided by CBCRM supplies the setting for individualsto work together, and in so doing, to build social capital. Moreover, successful col-lective action ‘‘spawns further social capital when acts of reciprocity reinforce normsof reciprocity, and when small but visible accomplishments encourage new stake-holders to join their partnership and current participants to redouble their efforts’’(Leach and Sabatier 2005, 234).
Social capital can also harm the progress and well-being of groups and com-munities. CBCRM efforts, particularly those that involve historically conflictinginterests, are often in danger of being co-opted by individuals in power and influ-enced by politics. Individuals may use their social capital to exclude others or toform alliances that benefit elite factions rather than the common good (Singleton2002; Walker and Hurley 2004). Some critics assert that theoretical and empiricalwork on social capital has largely failed to address the influence of power and poli-tics on social networks and the access to resources that they provide (Wall et al.1998). Another key criticism of the social capital literature is the assumption ofrational action or ‘‘universal selfishness.’’ Ahn and Ostrom (2002) and Ostrom(1998) have argued for a more complex ‘‘second generation’’ theory of socialcapital that acknowledges diverse motivations beyond enlightened self-interest,which emerge for various reasons including culture and values as well as individualexperience.
Social capital theorists distinguish among several types of network connections,each of which has different implications for CBCRM. Horizontal ties are connec-tions between individuals of ‘‘equivalent status and power’’ (Putnam 1993). Horizontalties are further broken down into ‘‘bonds’’ and ‘‘bridges.’’ Bonds, or strong ties, referto intimate connections between individuals who are often similar to one another,such as family and close friends (Granovetter 1973). The exclusionary function ofbonding social capital may be a barrier to effective collaboration. Bridges, or weakties, are informal or formal connections between agents that are different from oneanother, such as between friends, acquaintances, and colleagues or organizations
Does Collaboration Increase Social Capital? 325
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
with other organizations (Putnam 1993). Bridges are important in CBCRM for theirability to encourage wider cooperation (Putnam 1993).
Vertical ties, sometimes called linkages, connect ‘‘unequal agents in asymmetricrelations of hierarchy and dependence’’ (Putnam 1993, 173). Vertical ties connectagents across levels of organization, for example between a collaborative groupand a state or federal agency (Woolcock 1998). Such ties are important in the contextof CBCRM because they enable individuals or groups to draw on external resourcesas well as providing a means through which individuals or groups can influence pol-icy (Pretty and Ward 2001; Woolcock 1998). Likewise, vertical ties may be importantfor agencies because they are a link between the organization and the community.Such links provide agencies access to community viewpoints, concerns, and knowl-edge and furthermore, provide means for agencies to distribute their information.
Past research has shown that collaboration can increase trust and facilitatenetworking (Paulson 1998; Pretty and Ward 2001; Selin et al. 2000; Sturtevant andHorton 2000); however, empirical research that specifically examines whether collab-oration builds social capital is lacking. Further, the multiple dimensions of socialcapital (trust, norms of reciprocity, and networks) and the different kinds of networkconnections (bonds, bridges, and linkages), contribute to the conceptual complexityof social capital, making it difficult to measure. This study sought to address thesechallenges by developing and implementing a survey to assess group-level measuresof social capital in CBCRM groups. By group-level social capital, we mean the aggre-gate perceptions of individual participants in each group, of relationships amongmembers of that group. This definition of group social capital corresponds withthe definition of organizational social capital used by Leana and van Buren(1999), ‘‘a resource reflecting the character of social relations within the organization,realized through members’ levels of collective goal orientation and shared trust’’(540). We refer to groups rather than organizations, because many of the CBCRMgroups we studied did not have formal organizational status. Thus, although thegroups we studied are community-based,1 our unit of analysis is the group, not thecommunity or the individual. The specific objectives of our study were (1) to assesswhether the level of social capital in CBCRM groups changed over time, (2) to deter-mine which dimensions of social capital were most affected over time, and (3) toidentify specific mechanisms through which collaboration built social capital.
Site Description and Study Groups
This research focused on Moffat, Routt, and Jackson counties in northwestColorado (Figure 1). Historically, livestock grazing was one of the dominant usesof public and private lands in this area and ranchers were among the primary andmost influential public land users. However, with the recent rapid growth and devel-opment in this region, conflict over access, use, and management of the area’sresources is increasing. In particular, other users of public lands, such as recreation-ists and oil and gas companies, have increasingly demanded access to public landsand expressed interest in their management. Consequently, public land managersand communities are seeking ways to resolve the conflict over access to and manage-ment of these public lands. The numerous CBCRM groups in this area are largely aresponse to the increased conflict among resource users. In this study, we includedeight CBCRM groups based in northwest Colorado. These groups range in size, mis-sion, duration and frequency of interaction and are described in Table 1.2 Despite
326 C. L. Wagner and M. E. Fernandez-Gimenez
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
important differences in the socio-economic composition and conditions of the threestudy counties, many of the natural resource conflicts in the region share commonthemes, and a number of actors participate in multiple groups within the region.3
Methods
Sampling Frame
We compiled a list of all active, idle, and disbanded collaborative groups consistingof a diverse group of stakeholders focused on natural resource management inMoffat, Routt, and Jackson counties by contacting government agencies, cooperat-ive extension offices, and community organizations. From the sampling frame of 24collaborative groups, 7 groups were excluded due to lack of information. From theremaining 17 groups we purposively chose groups that were sufficiently large tocompare (4 groups were excluded because they had fewer than 10 participants), wererepresentative of collaborative groups in the study area (2 groups excluded), andwanted to participate (1 group excluded). The remaining 10 groups were includedin the initial study and survey, but 2 were dropped from the final analysis due tolow response rates (less than 45%) that resulted in small sample sizes.
Survey Design and Implementation
In order to assess change in social capital over time the survey was designed with twosets of identical questions. The first section asked the respondent to answer thequestions thinking about the group as it was when he=she first joined it. The secondsection asked the respondent to answer the same set of questions, thinking about the
Figure 1. May of study area.
Does Collaboration Increase Social Capital? 327
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Ta
ble
1.
Su
mm
ary
of
cha
ract
eris
tics
of
the
eig
ht
com
mu
nit
y-b
ase
dco
lla
bo
rati
ve
gro
up
sin
clu
ded
inth
est
ud
y
Gro
up
na
me
Gro
up
mis
sio
n=
focu
sY
ears
act
ive
Nu
mb
era
nd
typ
eo
fp
art
icip
an
tsM
eeti
ng
freq
uen
cyC
ou
nti
esin
vo
lved
No
rth
wes
tC
olo
rad
oS
tew
ard
ship
(NW
CO
S)
Imp
rov
ep
ub
lic
lan
ds
dec
isio
nm
ak
ing
by
pro
mo
tin
gco
op
era
tio
na
mo
ng
div
erse
inte
rest
s.A
pri
ma
ryfo
cus
isto
con
trib
ute
toth
ed
evel
op
men
to
fth
eB
LM
Res
ou
rce
Ma
na
gem
ent
Pla
n(R
MP
)fo
rth
eL
ittl
eS
na
ke
Res
ou
rce
Are
ain
NW
Co
lora
do
.
20
03
-P
rese
nt
11
2(l
oca
lre
sid
ents
,a
gen
cyst
aff
,st
ate
wid
eo
rga
niz
ati
on
s)
At
lea
sto
nce
am
on
th,
oft
enm
ore
Pri
ma
rily
Mo
ffa
t
Ow
lM
ou
nta
inP
art
ner
ship
(OM
P)
Bri
ng
tog
eth
era
gen
cies
an
dp
riv
ate
lan
do
wn
ers
toa
dd
ress
reso
urc
eco
nfl
icts
on
pu
bli
ca
nd
pri
va
tela
nd
s.Is
sues
the
gro
up
ha
sw
ork
edto
ad
dre
ssin
clu
de
liv
esto
ckg
razi
ng
on
pu
bli
cla
nd
s,n
ox
iou
sw
eed
s,a
nd
liv
esto
ck=
wil
dli
fein
tera
ctio
ns.
19
93
-P
rese
nt
17
(lo
cal
resi
den
tsa
nd
ag
ency
sta
ff)
Ab
ou
to
nce
am
on
thJa
ckso
n
Co
mm
un
ity
Ag
ricu
ltu
reA
llia
nce
(CA
A)
Pro
mo
tep
rese
rva
tio
no
fa
gri
cult
ura
lla
nd
sa
nd
incr
ease
aw
are
nes
so
fth
eY
am
pa
Va
lley
’sa
gri
cult
ura
lh
erit
ag
e.T
he
gro
up
ha
sw
ork
edto
ach
iev
eth
isb
yb
rid
gin
gd
iver
sese
cto
rso
fth
eco
mm
un
ity
thro
ug
hv
ari
ou
sp
rog
ram
s,ev
ents
,a
nd
cou
rses
.
20
00
-P
rese
nt
26
(lo
cal
resi
den
ts,
org
an
iza
tio
ns
an
da
gen
cies
)
Bet
wee
no
nce
am
on
tha
nd
qu
art
erly
Ro
utt
Em
era
ldM
ou
nta
inP
art
ner
ship
(EM
P)
Wo
rkto
wa
rdp
rese
rvin
gE
mer
ald
Mo
un
tain
for
recr
eati
on
,w
ild
life
,a
nd
gra
zin
g.
Th
eg
rou
ph
as
bee
nw
ork
ing
wit
hth
eB
LM
too
rch
estr
ate
ala
nd
exch
an
ge
ino
rder
top
rote
ctth
em
ou
nta
infr
om
futu
red
evel
op
men
t.
19
95
-P
rese
nt
27
(lo
cal
resi
den
tsa
nd
ag
ency
sta
ff)
Ab
ou
to
nce
am
on
thR
ou
tt
328
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Co
lum
bia
nS
ha
rpT
ail
Gro
use
Wo
rkin
gG
rou
p(C
ST
G)
Dev
elo
ped
am
an
ag
emen
tp
lan
for
pro
tect
ing
Sh
arp
Ta
ilG
rou
sein
NW
Co
lora
do
.T
he
gro
up
dev
elo
ped
spec
ific
con
serv
ati
on
act
ion
sth
at
inco
rpo
rate
dth
en
eed
s,v
alu
es,
an
din
tere
sts
of
div
erse
sta
keh
old
ers.
20
00
–2
00
16
0(l
oca
lre
sid
ents
,a
gen
cies
an
dN
GO
s)
Met
at
lea
sto
nce
am
on
thM
off
at
an
dR
ou
tt
Ax
ial
Ba
sin
Co
ord
ina
ted
Res
ou
rce
Ma
na
gem
ent
(AB
-CR
M)
Wo
rkto
reso
lve
reso
urc
eu
seco
nfl
ict
bet
wee
na
gen
cies
an
dla
nd
ow
ner
s.T
he
gro
up
dev
elo
ped
am
an
ag
emen
tp
lan
tore
solv
eco
nfl
ict
reg
ard
ing
av
ail
ab
ilit
yo
ffo
rag
ere
sou
rces
for
wil
dli
fea
nd
liv
esto
ck.
19
92
-P
rese
nt
14
(lo
cal
resi
den
tsa
nd
ag
enci
es)
Init
iall
ym
eto
nce
am
on
th,
no
wm
eet
twic
ea
yea
r
Mo
ffa
t
No
rth
wes
tC
olo
rad
oS
ag
eG
rou
seW
ork
ing
Gro
up
(SG
WG
)
Dev
elo
pa
ma
na
gem
ent
pla
nfo
rp
rote
ctin
gsa
ge-
gro
use
inN
WC
olo
rad
ow
hil
em
eeti
ng
the
nee
ds
of
the
com
mu
nit
ya
nd
div
erse
sta
keh
old
ers.
Th
eg
rou
pd
evel
op
edco
nse
rva
tio
na
ctio
ns
tha
tco
uld
be
tak
ento
mee
tth
eir
go
als
.
19
96
-P
rese
nt
47
(lo
cal
resi
den
ts,
ag
enci
esa
nd
NG
Os)
Ab
ou
to
nce
am
on
thM
off
at
an
dR
ou
tt
Ya
mp
aR
iver
Sy
stem
Leg
acy
Pro
ject
(YR
LP
)
Wo
rkto
pro
tect
eco
log
ica
lh
ealt
ho
fth
eY
am
pa
Riv
era
nd
ad
jace
nt
ag
ricu
ltu
ral
lan
ds,
wh
ile
pro
vid
ing
op
po
rtu
nit
ies
for
recr
eati
on
.T
he
gro
up
wo
rked
top
urc
ha
seco
nse
rva
tio
nea
sem
ents
an
dto
neg
oti
ate
coo
per
ati
ve
ma
na
gem
ent
ag
reem
ents
.
19
96
-P
rese
nt
12
(lo
cal
resi
den
ts,
ag
enci
esa
nd
NG
Os)
Ab
ou
to
nce
am
on
thM
off
at
an
dR
ou
tt
329
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
group as it is today, or when he=she last participated. With this design, we wereable to mimic a pre- and posttest design in circumstances in which a pretest wouldotherwise be impossible.
As with many surveys, recall bias may be a source of error. In addition, itis possible that a respondent’s answers to the first set of questions might haveinfluenced their answers to the second set. In order to check the data for accuracyand consistency, we conducted semistructured interviews with 12 key informants,based on their knowledge of or involvement in one or more of the surveyed groups.In addition, we asked respondents about two points in time, the beginning and theend of their participation, which were likely the most salient and therefore whichrespondents are most likely able to recall (Eisenhower et al. 2004).
We incorporated elements into the survey design in order to ensure contentvalidity of social capital as it is conceptualized and operationalized in this study.We thoroughly reviewed the existing literature on social capital measures, concen-trating on measures that have been evaluated for reliability and validity (Krishnaand Shrader 1999; O’Brien et al. 2004; Onyx and Bullen 2001; Stone and Hughes2002). Based on this literature review we included four dimensions of social capital:trust, rules and reciprocity, values and beliefs, and communication quality and quan-tity. As previously indicated, an important aspect of social capital is social networks,but time and resources constraints prevented a comprehensive network analysis.Instead, we used communication quality as a proxy for network quality becausesocial networks are ‘‘produced through communication’’ and, in turn, ‘‘determinethe communication of its members’’ (Pace and Faules 1994).
For each of the four social capital dimensions we selected and developed mea-sures that were relevant to a group-level measure of social capital. Each dimensionwas measured for two network types: connections among group participants (groupbridges) and connections between the group and government agencies (linkages).The majority of items were based on a 7-point Likert-like scale in which responsevalues were 1 ¼ not at all, 4 ¼ somewhat, and 7 ¼ to a great extent. This allowedus to determine the extent respondents felt each statement described the participantsof the group. (For the specific items used for each dimension, refer to the tables inthe results section.) We submitted the survey instrument for review by three otherscholars of social capital and community-based collaboration. We also pilot testedthe survey on two collaborative groups outside of the study area (52 respondents),after which we made revisions to improve the survey’s applicability and clarity.4
Using Dillman’s tailored design method (Dillman 2000), we sent questionnairesto all participants of the 10 original study groups, as identified from group mailinglists. Group participants included resource users, other citizens, representatives ofvarious organizations and interest groups (e.g., environmental and agriculturaladvocacy groups), and agency staff. Participants involved with more than one groupin our sample received multiple surveys, one for each group in which they partici-pated. In all we sent out a total of 422 surveys to 339 people. We revised the partici-pant lists to exclude individuals who never attended a group meeting, and those withincorrect addresses, resulting in a revised list of 256 people. For the eight groupsincluded in the analysis, we received 174 completed surveys from 143 people foran overall response rate of 55%. Individual group samples sizes for the eight groupsranged from 7 to 56 respondents and response rates ranged from 50 to 77%.
We conducted a nonresponse bias analysis to determine whether nonrespon-dents differed from respondents. We used a stratified random sampling procedure
330 C. L. Wagner and M. E. Fernandez-Gimenez
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
to select 10% of the nonrespondents from the six groups that had response ratesbelow 65%. We phone surveyed the nonrespondents and compared their responsesto respondents for 4 items. Differences between respondents and non-respondentsfor all items were not statistically significant (p > .05) and had small to mediumCohen’s D effect sizes (0.06 to 0.67), and thus we concluded our sample was repre-sentative and we could generalize to the group (Cohen 1988).
Interviews
In order to validate and enrich the survey data, we interviewed 12 participants fromthe groups in our study. We selected group participants who had comprehensiveknowledge of a particular group or who were actively involved in numerous groups.We specifically selected active participants in order to better understand the specificactions that individuals and groups can take to enhance group-level social capital.We sought individuals active in multiple groups for their comparative perspectiveon several of the studied groups. However, we acknowledge that by selecting activeparticipants in multiple groups our interviews were likely biased toward individualswith high levels of individual social capital. We interviewed one to seven participantsin each of the eight groups, with an average of three participants from each group.Interviews were semistructured and focused on the change in group participants’relationships over time, and the behaviors and activities that contributed to ordetracted from these relationships.
Data Analysis
Internal reliability was very strong for the scales measuring trust at the outset of thegroup (Cronbach alpha [a] ¼ .94) and current trust (a ¼ .94) and for similar valuesand beliefs at the outset (a ¼ .97) and currently (a ¼ .96). All items had correcteditem-total correlations of at least .7 and deleting items could not significantlyimprove Cronbach alpha scores.
We evaluated the change over time of social capital for two network types,connections among group participants (group bridges) and connections betweenthe group and government agencies (agency linkages). We analyzed the groupbridges in two ways, pooled and group by group. For the pooled analysis we com-bined all participants from the eight groups and used paired t-tests to examinewhether there was an overall temporal trend of change. We adjusted p values usingBonferroni’s correction to account for multiple tests by multiplying p values by thenumber of tests within each of the dimensions of social capital (Wright 1992). Weused Cohen’s D effect sizes to assess the standardized difference, or strength ofthe relationship, between the initial and current measures in relation to the amountof variation in the data (Cohen 1988). When evaluating Cohen’s D, 0.2, 0.5, and 0.8are considered small, medium, and large effects, respectively (Cohen 1988). For thegroup by group analysis, we analyzed the group bridges of each group, therebyallowing us to determine if there were differences among groups. We used descriptivestatistics rather than paired t tests because of small group sample sizes.
We analyzed the agency linkages using the pooled sample of all eight group part-icipants in two ways: a collective analysis of all agencies, using paired t-tests andadjusting p values using Bonferroni’s correction; and a separate analysis of eachagency, using descriptive statistics rather than paired t tests due to small sample sizes.
Does Collaboration Increase Social Capital? 331
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Interview Analysis
Following transcription of all the recorded interviews, we coded all interviews usingNVivo qualitative data management software (QSR International Pty 1999–2000).Using techniques described by Coffey and Atkinson (1996), we deductively codedbased on the dimensions of social capital used in the survey. We also inductivelycoded all interviews to assess emergent themes about social capital development(Strauss and Corbin 1998). We analyzed the coded data by reading through thecodes looking for emergent subthemes that would provide explanation for our sur-vey results and insight into the development of social capital in collaborative groups.
Results
Group Bridges
The pooled analysis of group bridges showed a significant increase for 12 of the 16group social capital variables (p < .05), which suggests group bridging social capitalincreased over time (Table 2). The differences over time ranged from 0.21 to 0.60 andCohen’s D effect sizes for the statistically significant variables ranged from 0.16 to0.30, indicating small effects.
A comparison of participants’ responses by group shows that the direction andextent of change over time for the group bridging social capital variables differsamong collaborative groups (Table 3). For five of the eight groups (NWCOS,OMP, CAA, SGWG, YRLP), 15 or more of the 16 group social capital variablesincreased over time. The other three groups (EMP, CSTG, AB-CRM) experiencedsome increases over time (8 or 9 variables), but also experienced notable decreases(4 to 6 variables) and instances of no change (1 to 4 variables). Though the specificsocial capital variables that changed over time varied across groups, each of the 16group bridging variables increased in at least one group. Notably, two variables,‘‘respect others’ viewpoints’’ and ‘‘consider all participants’ input equally,’’ increasedin all eight groups, and eight other variables increased in seven of the eight groups.The variables ‘‘commitment to the group’’ and ‘‘trust’’ decreased in the most groups,four and three out of the eight, respectively.
Agency Linkages
When all agencies were combined, the analysis of agency linkages showed a signifi-cant increase for all eight agency linkage variables (p < .05) (Table 4). Cohen’s Deffect sizes for the statistically significant variables ranged from 0.28 to 0.43, indicat-ing small effects. Differences over time for the agency social capital variables rangedfrom 0.45 to 0.62.
When each agency was considered separately, we found increases of at leastseven of the eight agency social capital variables for six out of seven agencies.Increases over time for the agency social capital variables ranged from 0.16 to 1.4(excluding one variable that did not change). The majority of agency social capitalvariables increased from relatively high initial levels (>4).
One agency, the City of Steamboat Springs, deviated from the overall trend. Forthis agency, two variables, ‘‘trust’’ and ‘‘open and clear communication’’ decreasedover time, and one variable, ‘‘respect each other’s knowledge,’’ did not change. Theremaining five variables increased over time, but, compared to the other agencies, the
332 C. L. Wagner and M. E. Fernandez-Gimenez
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Ta
ble
2.
Co
mp
ari
son
of
bri
dg
ing
soci
al
cap
ita
l‘‘
wh
enI
firs
tjo
ined
the
gro
up
’’(t
1)
ver
sus
‘‘to
da
yo
rw
hen
Ila
stp
art
icip
ate
d’’
(t2
)p
oo
lin
ga
llp
art
icip
an
tso
fei
gh
tco
lla
bo
rati
ve
gro
up
s
t1t2
Ch
an
ge
Std
.d
ev.
tv
alu
ep
va
lue
Co
hen
’sD
Tru
stT
rust
(in
dex
)a4
.98
5.1
90
.21
.94
3�
2.7
22
.00
7�
0.1
6R
ule
sa
nd
reci
pro
city
Wo
rkto
get
her
acc
ord
ing
toco
mm
on
gro
un
dru
lesa
4.8
55
.15
0.3
01
.30
6�
2.8
66
.04
0�
0.2
1
Ret
urn
act
so
fg
oo
dw
illa
4.8
04
.98
0.1
81
.11
8�
2.0
31
.35
20
.13
Are
hel
pfu
la5
.06
5.2
40
.18
1.0
71
�2
.05
1.3
36
0.1
4A
reco
mm
itte
da
4.7
64
.93
0.1
71
.29
7�
1.6
21
.85
60
.12
Rec
og
niz
eg
rou
pv
alu
ea4
.94
5.2
90
.35
1.4
07
�3
.09
2.0
16�
0.2
4S
ho
wco
nce
rnfo
rg
rou
pw
elfa
rea
4.2
54
.58
0.3
31
.31
9�
3.1
16
.01
6�
0.2
0A
rew
illi
ng
toco
mp
rom
isea
4.0
54
.41
0.3
61
.18
2�
3.7
73
<.0
01�
0.2
1S
ha
rere
sou
rces
a4
.54
4.9
20
.38
1.2
50
�3
.73
6<
.00
1�
0.2
8S
imil
ar
va
lues
an
db
elie
fsS
imil
ar
Va
lues
(in
dex
)a4
.24
4.4
10
.17
1.0
88
�1
.88
8.0
61
0.1
1C
om
mu
nic
ati
on
qu
ali
tya
nd
qu
an
tity
Are
wil
lin
gto
list
ena
4.6
65
.03
0.3
71
.38
6�
3.3
14
.00
6�
0.2
5R
esp
ect
each
oth
ers’
vie
wp
oin
tsa
4.5
34
.90
0.3
71
.23
6�
3.7
15
<.0
01�
0.2
3C
on
sid
era
llp
art
icip
an
t’s
inp
ut
equ
all
ya
4.1
64
.49
0.3
31
.21
7�
3.4
44
.00
6�
0.1
9H
av
eo
pen
an
dcl
ear
com
mu
nic
ati
on
a4
.57
5.0
10
.44
1.2
56
�4
.35
5<
.00
1�
0.3
0S
ha
rein
form
ati
on
a4
.65
5.2
50
.60
2.6
14
�2
.81
4.0
36�
0.2
8H
av
est
ron
gw
ork
ing
rela
tio
nsh
ips
wit
ho
ther
pa
rtic
icp
an
tsb
4.0
54
.48
0.4
31
.52
5�
3.5
10
.00
6�
0.3
0
No
te.
Ast
eris
kin
dic
ate
ssi
gn
ific
an
ta
tp<
.05
aft
era
dju
stin
gfo
rm
ult
iple
test
su
sin
gB
on
ferr
on
i’s
corr
ecti
on
(ori
gin
al
pv
alu
em
ult
ipli
edb
yth
en
um
ber
of
item
sin
the
resp
ecti
ve
soci
al
cap
ita
ld
imen
sio
n).
aS
even
-po
int
sca
lein
dic
ati
ng
exte
nt
resp
on
den
tsfe
elth
est
ate
men
td
escr
ibes
the
pa
rtic
ipa
nts
(1¼
no
ta
ta
ll,
7¼
toa
gre
at
exte
nt)
.bS
even
-po
int
sca
lein
dic
ati
ng
nu
mb
ero
fp
art
icip
an
ts(1¼
no
ne,
4¼
ha
lf,
7¼
all
).
333
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Ta
ble
3.
Co
mp
ari
son
of
bri
dg
ing
soci
al
cap
ita
l‘‘
wh
enI
firs
tjo
ined
the
gro
up
’’(t
1)
ver
sus
‘‘to
da
yo
rw
hen
Ila
stp
art
icip
ate
d’’
(t2
)fo
rei
gh
tco
lla
bo
rati
ve
gro
up
s
Gro
up
So
cia
lca
pit
al
item
NW
CO
SO
MP
CA
AE
MP
CS
TG
AB
-CR
MS
GW
GY
RL
P(n
)5
61
22
01
63
09
24
7
Tru
stT
rust
Ind
exd
4.0
9a
5.4
76
.21
5.4
15
.52
5.2
44
.79
5.0
04
.27
b6
.11
6.3
55
.25
5.3
85
.11
5.3
25
.76
0.1
8c
0.6
40
.14
�0
.16
�0
.14
�0
.13
0.5
30
.76
Ru
les
an
dre
cip
roci
tyW
ork
tog
eth
era
cco
rdin
gto
com
mo
ng
rou
nd
rule
sd3
.86
5.4
55
.89
5.2
55
.36
4.8
34
.79
5.5
74
.43
5.8
26
.22
5.3
85
.08
5.0
05
.11
6.4
30
.57
0.3
70
.33
0.1
3�
0.2
80
.17
0.3
20
.86
Ret
urn
act
so
fg
oo
dw
illd
3.9
85
.00
6.1
75
.06
5.2
14
.83
4.7
44
.86
4.1
05
.67
6.2
25
.06
5.2
14
.67
5.2
15
.57
0.1
20
.67
0.0
50
.00
0.0
0�
0.1
60
.47
0.7
1
Are
hel
pfu
ld4
.24
5.5
86
.17
5.2
55
.39
5.0
05
.26
5.1
44
.44
6.1
76
.22
5.2
55
.43
5.3
35
.32
5.8
60
.20
0.5
90
.05
0.0
00
.04
0.3
30
.06
0.7
2
Are
com
mit
ted
toth
eg
rou
pd
4.0
05
.33
5.3
95
.50
5.1
34
.67
4.5
85
.29
4.3
56
.00
5.7
25
.38
5.0
84
.33
4.5
35
.29
0.3
50
.67
0.3
3�
0.1
2�
0.0
5�
0.3
4�
0.0
50
.00
Rec
og
niz
eg
rou
pv
alu
ed4
.06
5.2
55
.83
5.5
05
.40
4.8
34
.89
5.8
64
.57
6.0
86
.22
5.7
55
.32
5.0
05
.11
6.4
30
.51
0.8
30
.39
0.2
5�
0.0
80
.17
0.2
20
.57
334
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Sh
ow
con
cern
for
gro
up
wel
fare
d3
.25
5.1
75
.67
4.4
44
.56
4.5
04
.47
4.0
03
.39
5.9
26
.00
4.8
14
.76
4.5
04
.95
5.2
90
.14
0.7
50
.33
0.3
70
.20
0.0
00
.48
1.2
9
Are
wil
lin
gto
com
pro
mis
eo
nd
iffi
cult
issu
esd
2.7
64
.83
5.4
44
.75
4.4
34
.83
4.1
14
.71
3.0
45
.75
6.0
64
.81
4.6
54
.83
4.6
85
.00
0.2
80
.92
0.6
10
.06
0.2
20
.00
0.5
80
.29
Sh
are
reso
urc
esd
3.6
95
.36
5.2
25
.25
4.8
35
.00
4.2
65
.14
4.2
45
.91
5.8
35
.19
5.0
45
.17
4.6
35
.29
0.5
50
.55
0.6
1�
0.0
60
.22
0.1
70
.39
0.1
4
Va
lues
an
db
elie
fsS
imil
ar
Va
lues
an
dB
elie
fsIn
dex
d3
.35
4.4
75
.35
4.9
04
.52
4.3
34
.31
4.9
03
.45
5.3
35
.35
4.7
74
.64
4.4
44
.52
5.5
20
.10
0.8
60
.00
�0
.13
0.1
20
.11
0.2
10
.62
Co
mm
un
ica
tio
nq
ua
lity
an
dq
ua
nti
tyA
rew
illi
ng
toli
sten
d3
.63
5.0
05
.83
5.3
15
.20
5.1
74
.68
4.7
14
.02
6.0
06
.17
5.1
95
.32
5.1
75
.21
5.8
60
.39
1.0
00
.34
�0
.12
0.1
20
.00
0.5
31
.15
Res
pec
to
ther
s’v
iew
po
ints
d3
.35
5.3
36
.00
4.7
54
.92
5.1
74
.68
5.0
03
.84
6.0
86
.22
4.8
14
.96
5.3
35
.32
5.5
70
.49
0.7
50
.22
0.0
60
.04
0.1
60
.64
0.5
7C
on
sid
era
llp
art
icip
an
ts’
inp
ut
equ
all
yd
3.0
04
.92
5.6
74
.63
4.4
44
.67
4.2
64
.57
3.3
35
.58
5.9
44
.94
4.5
24
.83
4.7
95
.14
0.3
30
.66
0.2
70
.31
0.0
80
.16
0.5
30
.57
Ha
ve
op
ena
nd
clea
rco
mm
un
ica
tio
nd
3.5
95
.00
5.2
25
.19
5.0
45
.17
4.7
94
.86
4.3
15
.82
5.7
25
.25
5.0
85
.17
5.1
15
.71
0.7
20
.82
0.5
00
.06
0.0
40
.00
0.3
20
.85
(Co
nti
nu
ed)
335
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Ta
ble
3.
Co
nti
nu
ed
Gro
up
So
cia
lca
pit
al
item
NW
CO
SO
MP
CA
AE
MP
CS
TG
AB
-CR
MS
GW
GY
RL
P(n
)5
61
22
01
63
09
24
7
Op
enly
sha
rein
form
ati
on
d3
.49
5.0
85
.83
5.1
35
.28
5.5
04
.83
4.8
64
.75
6.0
86
.22
5.1
95
.12
5.3
35
.33
5.2
91
.26
1.0
00
.39
0.0
6�
0.1
6�
0.1
70
.50
0.4
3
Nu
mb
ero
fp
art
icip
an
tsw
ith
wh
ich
resp
on
den
th
as
stro
ng
wo
rkin
gre
lati
on
ship
se
3.5
54
.17
4.5
84
.50
3.8
94
.57
4.2
44
.29
3.8
65
.58
5.1
65
.06
3.7
64
.67
4.8
35
.57
0.3
11
.41
0.5
80
.56
�0
.13
0.1
00
.59
1.2
8
aM
ean
sco
rea
tti
me
1‘‘
wh
enfi
rst
join
edth
eg
rou
p.’
’b
Mea
nsc
ore
at
tim
e2
‘‘to
da
y,
or
wh
enla
stp
art
icip
ate
d.’
’cM
ean
dif
fere
nce
bet
wee
nt1
an
dt2
.d
Sev
en-p
oin
tsc
ale
ind
ica
tin
gex
ten
tre
spo
nd
ents
feel
the
sta
tem
ent
des
crib
esth
eg
rou
pp
art
icip
an
ts(1¼
no
ta
ta
ll,
4¼
som
ewh
at,
7¼
toa
gre
at
exte
nt)
.eS
even
-po
int
sca
lein
dic
ati
ng
nu
mb
ero
fp
art
icip
an
ts(1¼
no
ne,
4¼
ha
lf,
7¼
all
).
336
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
differences (0.16 to 0.42) were much lower. Notably, however, this agency had highmean scores at time 1 and time 2 (over 5) for all measures, suggesting that verticalties between agencies have been strong throughout the period of interest in ourstudy.
Discussion
Theory suggests that the interactions among individuals that occur during collabor-ation may increase social capital (Ostrom 1997). We found an increase over time inthe majority of social capital measures across all eight groups in our study, thoughthe extent of the increases varied significantly between metrics of social capital andacross groups. These findings are consistent with prior research, which suggests thatcommunity-based collaboration may build social capital (Carr et al. 1998; Leach2002; Pretty and Ward 2001; Selin et al. 2000; Sturtevant and Horton 2000). Inthe following sections we discuss in more detail several patterns that emerged fromour survey results and interviews that provide a richer understanding of the processof social capital development in collaborative groups.
Respect and Equal Input: The Foundations of Social Capital?
Two social capital measures, ‘‘consider all input equally’’ and ‘‘respect others’ view-points,’’ increased in all eight groups. Since participants of collaborative groupsoften come to the table with the expectation of listening and sharing viewpoints, itseems reasonable that over time, participants would work toward considering all
Table 4. Comparison of agency linking social capital ‘‘when I first joined the group’’(t1) versus ‘‘today or when I last participated’’ (t2) for 7 agencies combined
Social capital itema t1 t2 Change Std. dev. t value p value Cohen’s D
Trust each other 4.49 5.11 0.62 1.071 �8.489 <.001� 0.43Are willing to
compromiseon difficult issues
4.09 4.61 0.52 1.190 �6.378 <.001� 0.34
Are committed tothe group
4.75 5.30 0.55 1.279 �6.323 <.001� 0.39
Share resources 4.70 5.15 0.45 1.349 �4.797 <.001� 0.30Share similar values 4.24 4.69 0.45 1.227 �5.346 <.001� 0.28Have open and clear
communication4.60 5.18 0.58 1.147 �7.391 <.001� 0.41
Respect each other’sknowledge
4.67 5.18 0.51 1.129 �6.661 <.001� 0.36
Share information 4.71 5.17 0.46 1.273 �5.316 <.001� 0.31
Note. Asterisk indicates significant at p < .05 after adjusting for multiple tests usingBonferroni’s correction (original p value was multiplied by 8).
aSeven-point scale indicating extent respondents feel the statement describes the relationshipbetween the group and government agencies (1 ¼ not at all, 4 ¼ somewhat, 7 ¼ to a greatextent).
Does Collaboration Increase Social Capital? 337
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
participants’ input equally and respecting others’ viewpoints. Respect and listeningmay indeed be the foundations of social capital in collaborative processes. Even ifa group never gets to building trust and compromising, listening to and respectingothers’ opinions and perspectives is a likely outcome of meeting face-to-face. Asone participant noted, ‘‘I think that by the end of it, everybody, at least had a respectand understanding for the other stakeholders, what their concerns, issues, and fearswere.’’
Even so, collaborative situations may not easily overcome concentrations ofpower and individual influence. Though ‘‘consider all input equally’’ increased inall eight groups, the increase was slight and this indicator remained one of the lowestscored in all groups. If respect and equality are indeed foundations of social capital,failure to build strong foundations may hamper further social capital development.For example, uneven influence among participants within a group may affect otheraspects of social capital such as trust. One participant commented:
‘‘My perception is that interests aren’t being addressed and incorporatedevenly . . . whether it is true or not, it is my perception and it doesundermine trust in the group and a little bit of trust in the [agency].’’
Laying the groundwork and providing an environment that treats individualsfairly and equally is crucial for relationships among participants to mature andstrengthen. This may be particularly important for agencies because they areexpected to be objective. Agencies must make concerted efforts to ensure that theyare perceived as impartial to any given stakeholder group.
Agencies: The Principal Beneficiaries?
Across all eight groups, we found that agency linkage networks tended to strengthenmore than group bridging networks. This pattern may be explained, in part, by thecontext of the groups in our study. In northwest Colorado, the prevailing conserva-tive Western culture is often distrustful and skeptical of the motives and objectives ofgovernment agencies. Through repeated interaction between agency staff andcommunity members, collaborative groups often provide opportunities for agenciesand community members to communicate in a more productive manner. Communi-cation provides the opportunity for exchange of viewpoints, which in turn, mayincrease understanding, trust, and respect. As one participant commented:
I have more appreciation for what the agency has to try and do. I have anew respect for what they have to deal with . . . sometimes I got theimpression before that they were more lopsided in favor of one way orthe other. But I think as they go through this process, I’m seeing, I think,that they are trying hard.’’
Just as participants may gain understanding of agency viewpoints, agency staffmay gain understanding of stakeholder viewpoints. One agency participant noted:
‘‘Maybe the [group] is helping them too . . . I have heard them say onmore than one occasion, it has helped bring some of these viewpoints
338 C. L. Wagner and M. E. Fernandez-Gimenez
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
more into perspective for them. So maybe it is easier for them to juggle allthese viewpoints when they have a clearer understanding of what every-one’s concerns are.’’
This suggests that the benefits that agencies derive from participating in colla-borative groups and establishing connections with citizens are significant. If agencypersonnel are willing and capable of committing to collaborative processes, agenciesmay benefit by increasing their standing and effectiveness. The specific benefits toagencies include an increased understanding of stakeholder viewpoints, an avenueto communicate agency policies directly to stakeholders, and linkages through whichthey can establish trust and reciprocal exchanges of information and resources,cornerstones of social capital. These benefits may in turn increase the capacity ofagencies to accomplish their management objectives.
Straining Relationships: Declines of Social Capital
In some groups, the dominant trend of increasing social capital was accompanied bynotable decreases for some social capital metrics. These findings are inconsistentwith most previous research that has focused on cases in which collaborationincreases social capital (Carr et al. 1998; Leach 2002; Pretty and Ward 2001; Selinet al. 2000; Sturtevant and Horton 2000), though there are a few notable exceptions(Singleton 2002; Walker and Hurley 2004).
Controversial group actions are one potential explanation for the decreases insocial capital found in our study groups. For example, in the EMP, a few groupactions clearly caused strain among participants and were likely the source of thedecreases in social capital measures. Some participants of the EMP perceivedthat the county had decreased its commitment to the group. One participant notedthat although the county was very involved in the beginning, once the partnershipdecided to pursue a specific land exchange:
‘‘The county commissioners backed off from their . . . open support. Theystill support the objectives of the partnership . . . but I think thatis a relationship that is really sensitive and we have had to workaround that.’’
In this case, the county hesitated committing to a decision that was not supported bysome members of the public. The group’s feelings about the county are also evidentthrough survey results. Initially, 63% of group respondents listed the county asone of the two most important agencies to the group, compared to only 13% whoindicate that the county is currently important to the group.
This example illustrates that social capital development is complex and dynamic.These results support social capital theory that asserts, once built, individuals mustcontinually foster, maintain, and utilize social capital (Bourdieu 1986). The impli-cation for participants of collaborative groups is that the mere fact that a collabora-tive group exists does not ensure that an increase in social capital will result.Participants must work continually and actively at maintaining and developing trust,reciprocity, and connections, which in turn facilitate interpersonal communicationand group success.
Does Collaboration Increase Social Capital? 339
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Building Social Capital: What Facilitates the Process?
In some ways, building social capital may be a passive by-product of individualsworking together face-to-face and getting to know one another. However, indivi-duals and groups can also actively engage in activities and take actions to encouragesocial capital development. Eight factors that encourage and facilitate social capitaldevelopment emerged from our interviews: commitment and continuity; understand-ing, empathy, and respect; transparency; and dependability and predictability, eachof which is discussed next.
Participant commitment to a group and continuity of participation ensureopportunities for repeated interaction, which is crucial for building trust and expec-tations of reciprocity (Ostrom 1990). As one agency employee commented, ‘‘Whenthey saw that I was committed to the process, and that I was going to be there, ithelped to develop that level of trust.’’ Building norms of reciprocity requires commit-ment and continuity because they provide opportunities for individuals to recipro-cate exchanges in the future. A continuing relationship with expectations of futureexchanges develops obligations between group participants (Putnam 1993).
Shared understanding, empathy, and respect for others’ viewpoints develop overtime as individuals in a group have the opportunity to share stories and hear others’perspectives. Understanding individuals’ perspectives, in turn, is a prerequisite forrespecting them. Increased trust and strengthened relationships may develop froma shared appreciation of multiple viewpoints. When asked what it takes to buildstronger relationships, a participant commented:
‘‘I think it’s just, nothing specific, it’s just being with each other, andnoticing that that’s not a bad person, it’s just his or her job, that’s justwhat they do. I think that is really, what it is, is starting to see peoplefor their own personalities and being an actual person instead of a forceout there that sues me or whatever else.’’
Transparency, being open about motives, objectives, and actions, encourages trustby feeding understanding. Cohen and Prusak (2001) comment: ‘‘Knowing who peopleare and what they are doing builds social connections and trust, just as secrecy buildssuspicion’’ (46). Regarding transparency, an agency employee notes:
‘‘The transparency has been really helpful in building the trust level forthe [agency]. They can see what we do and we explain why we didn’tdo certain things. It is such an open process that it is hard to hide any-thing from them, I think that has really helped the level of trust withthe community and the [agency] . . . once we start sharing informationand being transparent and letting them right in on the process.’’
Dependability and predictability can encourage trust and reciprocity by facilitat-ing the formation of a positive reputation for being trustworthy (Ahn 2002). Oneparticipant commented:
‘‘I think that trust is developed in the atmosphere of dependability.Saying, I know where you are coming from and because I know whereyou are going to come from I can trust your position.’’
340 C. L. Wagner and M. E. Fernandez-Gimenez
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
In effect, people are more trusting when they are not surprised and people’s actionsare predictable. They may disagree with others’ positions, but at least their behavioris expected and transparent. Ostrom (1998, 12) notes, ‘‘A reputation for being trust-worthy, or for using retribution against those who do not keep their agreements orkeep up their fair share, becomes a valuable asset.’’ She continues that collectiveaction and social outcomes are produced by the interaction of three core variables,‘‘links between the trust that individuals have in others, the investment others makein trustworthy reputations, and the probability that participants will use reciprocitynorms.’’
Building social capital is a process that depends on the quality of communi-cation and interaction. Repeated interactions by committed individuals who engagein transparent, predictable, and respectful communication will likely build socialcapital among group participants. However, the relationship between the qualityof communication and the building of social capital in collaborative groups requiresfurther inquiry.
Conclusion
De Souza Briggs (2004 155) noted, ‘‘most of what we know about social capital isbased on studies of social capital in use . . . we clearly need to better understand socialcapital under development.’’ Our study has attempted to fill this gap by evaluatingsocial capital development in one context, community-based collaborative groups.Our findings have important implications for collaborative groups, communities,natural resource management agencies, and researchers.
We found that collaboration often builds social capital, particularly betweenagencies and other CBCRM participants. In conducting this research we found sup-port for portions of social capital theory, namely, that collective action may buildsocial capital. However, our research also illustrates the fragile quality of social capi-tal, which can be damaged by divisive collaborative efforts, and which must beactively maintained. Our findings suggest that some aspects of social capital, parti-cularly respect and considering all input equally, may be the foundations of socialcapital development. Finally, this research expands on existing theory by providinga description of how social capital develops in collaborative groups. Key mechan-isms for building social capital in collaborative groups emerged, including commit-ment and continuity; understanding, empathy, and respect; transparency; anddependability and predictability. This research provides groups and agenciesengaged in collaboration with a practical survey instrument with which to measuresocial capital within their group, as well as recommendations for how such groupsand agencies can actively encourage social capital development.
This research increases the understanding of social capital development incollaborative groups, and suggests that collaborative approaches may significantlybenefit natural resource management agencies by increasing their social capital.However, many questions remain unanswered. First, the measurement of social capi-tal at the group level remains challenging. Using a similar survey instrument in alongitudinal design with true pre- and posttest surveys of the same population wouldhelp to increase the reliability and validity of results, especially if coupled withadditional qualitative research. Second, further research is needed to understandthe extent to which characteristics of community-based collaborative groups affectsocial capital development and how social capital developed in groups translates into
Does Collaboration Increase Social Capital? 341
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
benefits for communities. Finally, research is needed to thoroughly understand howincreased social capital between agencies and other collaborative group participantsaffects agency efficacy and management.
Notes
1. Following our definition of community-based collaborative resource management, wecharacterize all of our study groups as community-based in that all focus on planning ormanagement for a particular location and all include participation by local residents andstakeholders, as well as agency staff, and in some cases local or regional nongovernmentalorganizations (NGOs) or advocacy organizations.
2. In this study, we treated each group as a unit of analysis. However, due to the relativelysparse population in this area, many individuals participated in more than one of the stud-ied groups, creating inter-group linkages. This study focuses on social capital built throughinteractions among the participants within each group, rather than between-group linkages.However, a network analysis of intergroup connections would be an excellent topic forfuture study.
3. Walden, the county seat of Jackson County, is suffering from rural decline due to sawmilland coal mine closures in the late 1980s and early 1990s. More recently, several majorranches in North Park, the large area of high-elevation rangelands that surrounds Walden,have sold and are now owned by wealthy absentee owners. It is not yet clear to what extentthis change in ownership is affecting social or economic relationships within this histori-cally agricultural community. Steamboat Springs, the seat of Routt County, is a bustlingresort community where demographic change and soaring land values threaten the viabilityof traditional land-based livelihoods. Here the impacts of demographic change on thearea’s rural agricultural character and on the increasingly diverse values of local residentsare clearly evident. Moffat County covers a vast area of northwest Colorado and its econ-omy continues to be dominated by natural resource-based industries, including ranching,hunting, and a rapidly expanding energy development sector. Growing attention to the sce-nic and environmental values of the area has increased conflicts over the use and manage-ment of public lands in the county.
4. The first pilot study involved a survey of a community-based collaborative group very simi-lar to the groups included in the study. We presented our survey at a monthly group meet-ing and sent it by email to all participants on the group’s mailing list. In addition to fillingout the survey, participants were asked to fill out a supplemental question form about thesurvey addressing the content, wording and relevance of the survey questions. We received10 completed surveys from this initial pilot and revised the survey based on this feedback.We then conducted a second pilot test using an abbreviated version of the revised survey ona group of representatives of clubs with a common agricultural interest. We presented thesurvey at a monthly meeting and received 42 completed surveys. The primary purpose ofthe second pilot test was to test the statistical analyses that would be applied to the finalsurvey. After analyzing the pilot survey data, the survey was revised again before it wasimplemented on the study population.
References
Adger, W. 2003. Social capital, collective action, and adaptation to climate change. Econ.Geogr. 79(4):387–404.
Ahn, T. 2002. Trust and collective action: concepts and causalities. Paper read at Annual Meet-ing of the American Political Science Association, August 28–1 September, Boston.
Ahn, T. and E. Ostrom. 2002. Social capital and the second-generation theories of collectiveaction: An analytical approach to the forms of social capital. Paper read at the AnnualMeeting of the American Political Science Association, August 29–1 September, Boston.
Bourdieu, P. 1986. The forms of capital. In Handbook of theory and research for the sociologyof education, ed. J. G. Richardson, 241–258. New York: Greenwood Press.
342 C. L. Wagner and M. E. Fernandez-Gimenez
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Brick, P., D. Snow, and S. Van De Wetering (Eds.). 2001. Crossing the great divide: Explora-tions in collaborative conservation in the American West. Washington, DC: Island Press.
Carr, D., S. Selin, and M. Schuett. 1998. Managing public forests: Understanding the role ofcollaborative planning. Environ. Manage. 22(5):767–776.
Coffey, A. and P. Atkinson. 1996. Making sense of qualitative data: Complementary researchstrategies. Thousand Oaks, CA: Sage.
Cohen, D. and L. Prusak. 2001. In good company. Boston: Harvard Business School Press.Cohen, J. 1988. Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Lawrence
Erlbaum Associates.Coleman, J. 1990. Foundations of social theory. Cambridge, MA: Harvard University Press.De Souza Briggs, X. 2004. Social capital: Easy beauty or meaningful resource? J. Am. Plan.
Assoc. 70(2):151–158.Dillman, D. 2000. Mail and Internet surveys: The tailored design method. New York: John
Wiley & Sons.Eisenhower, D., N. Mathiowetz, and D. Morganstein. 2004. Recall error: Sources and bias
reduction techniques. In Measurement errors in surveys, ed. P. Biemer, R. Groves, L.Lyberg, N. Mathiowetz, and S. Sudman, 127–144. Hoboken, NJ: John Wiley and Sons.
Granovetter, M. 1973. The strength of weak ties. Am. J. Sociol. 78(6):1360–1380.Gray, G. J., M. J. Enzer, and J. Kusel (Eds.). 2001. Understanding community-based forest
ecosystem management. Binghamton, NY: Food Products Press.Krishna, A. and E. Shrader. 1999. Social capital assessment tool. Paper read at Conference on
Social Capital and Poverty Reduction, June 22–24, Washington, D.C.Leach, W. 2002. Surveying diverse stakeholder groups. Society Nat. Resources 15:641–649.Leach, W. and P. Sabatier. 2005. Are trust and social capital the keys to success? Watershed
partnerships in California and Washington. In Swimming upstream: Collaborative approachesto watershed management, ed. P. Sabatier, W. Focht, M. Lubell, Z. Trachtenberg, A. Vedlitzand M. Matlock, 233–260. Cambridge, MA: MIT Press.
Leana, C. R. and H. J. Van Buren III. 1999. Organizational social capital and employmentpractices. Acad. Manage. Rev. 24(3):538–555.
Mullen, M. and B. Allinson. 1999. Stakeholder involvement and social capital: Keys to water-shed management success in Alabama. J. Am. Water Resources Assoc. 35(3):655–662.
O’Brien, M., C. Burdsal, and C. Molgaard. 2004. Further development of an Australian-basedmeasure of social capital in a U.S. sample. Social Sci. Med. 59:1207–1217.
Olsson, P., C. Folke, and T. Hahn. 2004. Social-ecological transformation for ecosystemmanagement: The development of adaptive co-management of a wetland landscape insouthern Sweden. Ecol. Society 9(4). http://www.ecologyandsociety.org/vol9/iss4/art2[Accessed January 1, 2005].
Onyx, J. and P. Bullen. 2001. The different faces of social capital in NSW Australia. In Socialcapital and participation in everyday life, ed. P. Dekker and E. Uslaner, 88–111. London:Routledge.
Ostrom, E. 1990. Governing the commons: The evolution of institutions of collective action.New York: Cambridge University Press.
Ostrom, E. 1997. Investing in capital, institutions, and incentives. In Institutions and economicdevelopment: Growth and governance in less-developed and post-socialist countries, ed. C.Clague, 153–181. Baltimore, MD: Johns Hopkins University Press.
Ostrom, E. 1998. A behavioral approach to the rational choice theory of collective action. Am.Polit. Sci. Rev. 92(1):1–22.
Pace, R. and D. Faules. 1994. Organizational communication, 3rd ed. Upper Saddle River,NJ: Prentice Hall.
Paulson, D. 1998. Collaborative management of public rangeland in Wyoming. Prof. Geogr.50(3):301–315.
Pretty, J. and H. Ward. 2001. Social capital and the environment. World Dev. 29(2):209–227.
Does Collaboration Increase Social Capital? 343
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14
Putnam, R. 1993. Making democracy work: Civic traditions in modern Italy. Princeton, NJ:Princeton University Press.
QSR International Pty. 1999–2000. NVivo 1.2 Software QSR revision 1.2.,Victoria, Australia.Schuett, M. and S. Selin. 2002. Profiling collaborative natural resource initiatives and active
participants. No. Jl. of Appl. Forestry 19(4):155–160.Selin, S., M. Schuett, and D. Carr. 2000. Modeling stakeholder perceptions of collaborative
initiative effectiveness. Society Nat. Resources 13:735–745.Singleton, S. 2002. Collaborative environmental planning in the American West: The good,
the bad, and the ugly. Environ. Polit. 11(3):54–75.Stone, W. and J. Hughes. 2002. Social capital: Empirical meaning and measurement validity.
Research Paper No. 27. Melbourne, Australia: Australian Institute of Family Studies.Strauss, A. and J. Corbin. 1998. Basics of qualitative research: Techniques and procedures for
developing grounded theory. Thousand Oaks, CA: Sage.Sturtevant, V. and R. Horton. 2000. Revisiting community capacity: Five years after FEMAT:
Insights from case studies of Rogue River National Forest communities. Portland, OR:USDA-Forest Service Pacific Northwest Research Station.
Walker, P. and P. Hurley. 2004. Collaboration derailed: The politics of ‘‘community-based’’resource management in Nevada County. Society Nat. Resources 17:735–751.
Wall, E., G. Ferrazzi, and F. Schryer. 1998. Getting the goods on social capital. Rural Sociol.63:300–322.
Weber, E. 2003. Bringing society back in: Grassroots ecosystem management, accountability,and sustainable communities. Cambridge, MA: MIT Press.
Wondolleck, J. M. and S. L. Yaffee. 2000. Making collaboration work: Lessons from innovationin natural resource management. Washington, DC: Island Press.
Woolcock, M. 1998. Social capital and economic development: Toward a theoretical synthesisand policy framework. Theory Society 27:151–208.
Woolcock, M. and D. Narayan. 2000. Social capital: Implications for development theory,research, and policy. World Bank Res. Observer 15(2):225–249.
Wright, S. 1992. Adjusted p-values for simultaneous inference. Biometrics 48(4):1005–1013.
344 C. L. Wagner and M. E. Fernandez-Gimenez
Dow
nloa
ded
by [
Ast
on U
nive
rsity
] at
10:
43 2
4 Ja
nuar
y 20
14