Number Eight

Building Trust in Public and Nonprofit Networks Personal, Dyadic, and Third-Party Influences Kristina T. Lambright Pamela A. Mischen Craig B. Laramee Binghamton University, New York

This article provides greater understanding of factors influencing interpersonal trust in networks composed of public and nonprofit service providers. The present theoretical model identifies propensity to trust, the perceived trustworthiness of the trustee, the relationship between the trustee and trustor, and third-party relationships as influencing interpersonal trust. The model is tested using action research data collected from a network of local social service providers. Key findings include the following: (a) Successful past cooperation between a trustor and a trustee and structural equivalence increase the likelihood the trustor will perceive the trustee as trustworthy; (b) the frequency of interactions between the trustor and trustee, trust transferability, and the perceived trustworthiness of the trustee have a direct, positive impact on whether the trustor trusts the trustee; and (c) trust between the trustor and trustee has a positive impact on expected future cooperation.

Keywords: social network analysis; network development; trust; social services

The delivery of public services is becoming more complex, with nonprofit, for-profit,and public organizations all playing a role in the new world of devolved public policy (Milward & Provan, 2000). Reflecting this trend, organizations are increasingly “net- worked” (O’Toole, 1997a, 1997b), and there is a growing scholarly interest in networks in a variety of public policy settings (Agranoff, 2007; Agranoff & McGuire, 1999; Edelenbos & Klijn, 2007; Kapucu, 2006; Milward & Provan, 1998; Musso, Weare, Oztas, & Loges, 2006; O’Toole & Meier, 2004; Provan & Milward, 1995). Drawing on O’Toole (1997a), this article defines networks as “structures of interdependence involving multiple organiza- tions” (p. 445).

Networks typically focus on specific policy or policy area (Agranoff & McGuire, 1999) and involve multiple, reciprocal exchanges (Powell, 1990). Each party in a network is dependent on resources controlled by another party, and pooling resources produces some type of benefit (Powell, 1990). Network relationships often span different sectors and include both formal and informal ties between organizations (Agranoff & McGuire, 1999).

The American Review of Public Administration

Volume XX Number X Month XXXX xx-xx

© 2009 Sage Publications 10.1177/0275074008329426

http://arp.sagepub.com hosted at

http://online.sagepub.com

Authors’ Note: Please address correspondence to Kristina T. Lambright, Department of Public Administration, College of Community and Public Affairs, Binghamton University, P.O. Box 6000, Binghamton, NY 13902; e-mail: [email protected].

Initial submission: March 25, 2008 Accepted: November 10, 2008

The American Review of Public Administration OnlineFirst, published on January 7, 2009 as doi:10.1177/0275074008329426

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

In contrast to hierarchies, there is no single guiding organizational goal in networks (Agranoff & McGuire, 1999). “Wicked problems,” challenges that must be addressed holis- tically rather than through fragmented strategies, have contributed to the emergence of net- works in the public sector (McGuire, 2006; O’Toole, 1997b; O’Toole & Meier, 2004). Although the flexibility of networks is an advantage, they are less stable than markets and require complex coordination and accountability mechanisms (Milward, 1996; Milward & Provan, 2000).

As personal relations and structures are “embedded” in institutions (Granovetter, 1985; Uzzi, 1997), trust has been identified as critical to the functioning of effective networks (Agranoff, 2007; Agranoff & McGuire, 2001; Cross & Parker, 2004; LaPorte, 1996; McGuire, 2006; O’Toole, 1997a; Powell, 1990). Trust is essential because networks replace hierarchal power with cooperative relationships based on interdependence and have fewer superordinate mechanisms for ensuring sustained operations compared with hierarchies and markets composed of competing hierarchies (LaPorte, 1996). There has been a great deal of scholarly interest in what influences the development of interpersonal trust in iso- lated, dyadic relationships (including Bohnet & Huck, 2003; Butler, 1995; Lewicki & Bunker, 1996; McAllister, 1995; Mayer, Davis, & Schoorman, 1995; Malhotra & Murnighan, 2002; Pillai, Schriesheim, & Williams, 1999; Rotter, 1971). However, organizational mem- bers typically interact with multiple individuals and are a part of complex social networks. There is little research that has modeled and directly assessed the influence that these third- party relationships have on the development of interpersonal trust (Ferrin, Dirks, & Shah, 2006). Moreover, the majority of research that examines intraorganizational relationships has focused on the private sector although the emergence of interorganizational networks is fundamentally changing the structure and processes for delivering public goods and services (Isett & Provan, 2005).

To fill this gap, this article provides a greater understanding of the factors influencing interpersonal trust in networks, contributing to the growing literature on trust in interorgani- zational networks (including Agranoff, 2007; Alter & Hage, 1993; Edelenbos & Klijn, 2007; Lane & Bachman, 1998; Uzzi, 1997). Building on Ferrin et al. (2006), this article examines the impact that third-party relationships have on the development of interpersonal trust. However, unlike Ferrin et al. (2006), we examine an interorganizational network composed of public and nonprofit service providers rather than an intraorganizational network in the private sector. The theoretical model developed in this article also considers the impact of other factors identified in the literature on the development of interpersonal trust in dyadic relationships. These other factors include propensity of the trustor to trust, the perceived trustworthiness of the trustee, and the relationship between the trustee and the trustor. In our article, the trustor is the individual in a dyadic relationship who is placing trusting in the other member of the dyad. The other member of the dyad is referred to as the trustee and is the individual in the dyadic relationship who is being trusted. We test our theoretical model on the determinants of interpersonal trust using action research data collected from a net- work of local social service providers collaborating to address the problem of adolescent self-injury by conducting a pseudo-path analysis (Kadushin, 1995). We conclude by consid- ering the implications of our findings for trust building in networks and highlighting areas for further research.

2 The American Review of Public Administration

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

Factors Influencing Interpersonal Trust in Networks

In this section, a theoretical model of the factors influencing interpersonal trust in net- works is developed. There is not a universally accepted definition of trust (Kramer, 1999; Rousseau, Sitkin, Burt, & Camerer, 1998). However, there is a consensus across disciplines that the two following conditions must be met in order for trust to exist: (a) risk and (b) inter- dependence (Rousseau et al., 1998). The dependent variable in our model is trust between the “trustor” and “trustee,” where the trustor is putting herself in a position of vulnerability and taking a risk by placing trust in the trustee.

Although public administration scholars have recognized the importance of trust in net- works (Agranoff & McGuire, 2001; LaPorte, 1996; McGuire, 2006; O’Toole, 1997a), the model presented in this article to explain the development of trust is relatively new for the field of public administration. Much of the research on the development of trust is in other fields such as management and organizational studies and psychology (including Burt & Knez, 1996; Ferrin et al., 2006; Gulati, 1995; Kramer, 1999; Lewicki & Bunker, 1996; Mayer et al., 1995; McAllister, 1995; Rotter, 1971; Vangen & Huxham, 2003; Uzzi, 1997). Reflecting this, the literature used to develop this article’s model is cross-disciplinary. Figure 1 depicts our model of factors influencing interpersonal trust in networks. The remainder of this section describes the relationships shown in this model and details our specific hypotheses. Each of the 10 arrows displayed in Figure 1 correspond to a separate hypothesis.

In our model, the perceived trustworthiness of the trustee directly influences whether the trustor trusts the trustee. Drawing on the influential work by Mayer et al. (1995), three factors shape the extent to which a trustee will be viewed as trustworthy: (a) the perceived ability of the trustee, which refers to the skills and competencies of the trustee in a specific domain; (b) the perceived benevolence of the trustee, which refers to the extent to which the trustor believes the trustee will act in his best interest; and (c) the perceived integrity of

Lambright et al. / Building Trust in Public and Nonprofit Networks 3

Third party relationships

Relationship between trustor and trustee

Network closure

Structural equivalence

Trust transferability

Propensity of trustor to trust

Perceived trustworthiness of trustee

1. Perceived ability 2. Perceived benevolence 3. Perceived integrity

Trustor trusts trustee

Frequency of interactions

Successful past cooperation

Expected future cooperation

Figure 1 Factors Influencing Interpersonal Trust in Networks

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

the trustee, which refers to the extent to which the trustor believes the trustee follows a set of principles that are acceptable to the trustor. In addition to having a direct impact on trust, the perceived trustworthiness of the trustee also acts as a mediator between trust and several other factors included in our model.

Hypothesis 1: The perceived trustworthiness of a trustee positively influences the likelihood a trustor will trust the trustee.

Three general categories of factors are hypothesized to influence the perceived trustwor- thiness of the trustee: (a) the propensity of the trustor to trust, (b) the relationship between the trustor and trustee, and (c) third-party relationships. Individuals vary in their general willingness to be vulnerable and risk placing trust in others (Kramer, 1999; Mayer et al., 1995; Rotter, 1971). Similar to Mayer et al. (1995), we view this propensity to trust as a personality trait that is stable across situations. Depending on their propensity to trust, one would expect some individuals to be predisposed to positively assess others’ ability, benev- olence, and integrity and other individuals to be predisposed to negatively assess others’ trustworthiness. Moreover, one would expect that a trustor with a propensity to trust would be more likely to trust a trustee regardless of who the trustee is and her perceived trustwor- thiness (Mayer et al., 1995). A trustor’s propensity to trust is especially likely to be impor- tant in unfamiliar situations because the trustor will not be able to draw on past experiences to determine the trustworthiness of the trustee (Rotter, 1971).

Hypothesis 2: A trustor’s propensity to trust positively influences a trustee’s perceived trust- worthiness.

Hypothesis 3: A trustor’s propensity to trust positively influences the likelihood the trustor will trust a trustee.

The relationship between the trustee and trustor is also likely to shape whether the trustor perceives the trustee as trustworthy. Interactions between the trustee and trustor provide the trustor with information that can be used to assess the trustee’s disposition, intentions, and motives (Kramer, 1999). In particular, one would expect the frequency of interactions to influence the development of trust (Bohnet & Huck, 2003; Edelenbos & Klijn, 2007; Gulati, 1995). In situations in which the trustor and trustee interact frequently, there is the potential for either individual to retaliate for past opportunistic behavior in future interac- tions. Given this, one might expect the trustor to be more likely to believe that the trustee will behave in a trustworthy way to avoid possible negative consequences. The quality of past interactions is also likely to matter (Mayer et al., 1995; Ring & Van de Ven, 1992; Van Slyke, 2006; Vangen & Huxham, 2003). In cases in which the trustee and trustor have cooperated successfully in the past, one would expect the trustor to have a more favorable assessment of the trustee’s ability, benevolence, and integrity. In addition, our model hypothesizes that the quality of past interactions affects their frequency: individuals are more likely to choose to interact with individuals that they have had cooperated with successfully in the past.

Hypothesis 4: The frequency of interactions between a trustor and a trustee positively influences the trustee’s perceived trustworthiness.

4 The American Review of Public Administration

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

Hypothesis 5: Successful past cooperation between a trustor and a trustee positively influences the trustee’s perceived trustworthiness.

Hypothesis 6: Successful past cooperation between a trustor and a trustee positively influences the frequency of their interactions.

Although trust is an inherently individual-level phenomenon (Zaheer, McEvily, & Perrone, 1998), it is rarely private (Burt & Knez, 1996). In organizational life, social struc- tures are complex, with most individuals participating in several dyadic relationships simultaneously rather than just one isolated dyadic relationship (Ferrin et al., 2006). As a consequence, one would expect the social networks that surround dyadic relationships to influence the development of interpersonal trust in dyadic relationships (Burt & Knez, 1996; Ferrin et al., 2006). Based on Ferrin et al. (2006), our model identifies three third- party relationships likely to indirectly affect the development of interpersonal trust: (a) net- work closure, (b) structural equivalence, and (c) trust transferability.

Our model hypothesizes that network closure increases the probability that the trustor and trustee have successfully cooperated. Network closure is the number of third parties who interact with both the trustor and trustee. In situations where the trustee and trustor only interact with each other, successful cooperation will just enhance the trustee’s reputation with the trustor. However, in situations where there is network closure, there will be an audience for cooperative efforts. Drawing on the field of impression manage- ment (Goffman, 1959), this creates an opportunity for the trustee and trustor to influence the impressions of third parties. When a third party is aware of an instance of successful cooperation between the trustor and trustee, one would expect this to favorably affect the third party’s assessment of both the trustor and trustee. As a result, successful cooperation will not just enhance the trustee’s reputation with the trustor but also will enhance the trustor’s and trustee’s reputation with third parties. As the audience for cooperative acts grow, the incentives for the trustor and trustee to behave cooperatively increases as well (Ferrin et al., 2006).

Hypothesis 7: Network closure positively influences the likelihood that a trustor and a trustee have successfully cooperated in the past.

We also expect structural equivalence to increase the likelihood the trustor and trustee have successfully cooperated. Structural equivalence is the extent that the trustor and trustee are similar in terms of the relationships they have with others as well as in terms of the relationships they do not have with others. In an interorganizational network, both the set of network members that the trustee and trustor interact with and the set of network members that the trustee and trustor do not interact with will be very similar when struc- tural equivalence between the trustee and trustor is high. Structural equivalence increases the likelihood the trustor and trustee will see each other as part of the same subgroup within a network and view their relationship as interdependent. According to Ferrin et al. (2006), the interdependence resulting from having similar relationships creates incentives for the trustor and trustee to behave cooperatively: the trustee and trustor will be more likely to believe that their future outcomes are linked together, to feel a sense of responsibility to each other, and to develop similar attitudes and beliefs.

Lambright et al. / Building Trust in Public and Nonprofit Networks 5

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

Hypothesis 8: Structural equivalence positively influences the likelihood that a trustor and a trustee have successfully cooperated in the past.

In addition to earning a reputation through interactions with a trustor, a trustee can earn a reputation through interactions with third parties. As a final relevant third-party relationship, we theorize that trust transferability will affect the trustor’s perceived trustworthiness of the trustee. Trust transferability is the number of third parties who trust the trustee and are also trusted by the trustor. Even though a trustor will not be able to directly observe many of the interactions between a trustee and trusted third parties, the trustor often will want to take these interactions into account when assessing whether a trustee is trustworthy. As a result, a trustor often does not just base assessments regarding the perceived trustworthiness of a trustee on direct observation but also must rely on the reputation that the trustee has with third parties. The judgments of the third party provide valuable supplemental information on the trustee’s trustworthiness (Ferrin et al., 2006). In situations where a trustee has earned a reputation of being trustworthy with a third party and that third party is trusted by the trustor, the trustor is more likely to believe that the trustee is trustworthy.

Hypothesis 9: Trust transferability positively influences a trustee’s perceived trustworthiness.

As previously discussed, our model theorizes that successful past cooperation between a trustor and a trustee is positively related to the trustee’s perceived trustworthiness and that the trustee’s perceived trustworthiness is positively related to the likelihood the trustor will trust the trustee. Cooperation is likely to not just play an important role in the development of interpersonal trust but also to be a consequence of trusting relationships (Alter & Hage, 1993; Edelenbos & Klijn, 2007; Van Slyke, 2006; Vangen & Huxham, 2003). Trust facili- tates future cooperation by reducing uncertainty in the relationship and concerns about opportunism, thereby minimizing the transaction costs involved in future relationships (Edelenbos & Klijn, 2007).

Hypothesis 10: Trust between the trustor and trustee positively influences expected future cooperation.

Method

To study the impact of trust on network development, we surveyed 37 social workers working for public and nonprofit human service agencies who attended a 2-day training workshop on dialectical behavioral therapy, a treatment for those engaging in self-injury. We also surveyed one of the organizers of the training session, even though he was not in attendance. At the beginning of the first day, all attendees were asked to complete a social network survey. This survey was prepopulated with the names of the other attendees. There were 44 names on the list. However, only 37 individuals attended the workshop. All attendees completed the survey, but one survey was unusable because we could not iden- tify the respondent. Consequently, our sample consisted of 36 attendees plus 1 organizer not in attendance.

6 The American Review of Public Administration

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

To some extent, the boundaries of this network are artificial. There are more social work- ers in the area who may be treating individuals engaging in self-injury than those who attended this training workshop. However, in conjunction with the sponsoring organization, we decided that those attending the workshop would be an appropriate group with whom to develop a “network response” to this issue for two reasons. First, those who participated in the workshop were there because of their interest in the problem of self-injury. Second, many who were there have been concerned about this issue for some time and have been actively engaged around the topic for more than a decade.

Up to this point, however, “engagement” has been limited to bringing speakers to the area from outside the community to conduct training sessions with social workers. Occasionally, this knowledge was passed to teachers and others who come in close contact with adolescents via training sessions conducted by individual social workers. Workshop organizers and an outside funding agency decided that it was time to try for a more con- certed approach to the problem and asked the researchers to help them develop an interor- ganizational network to respond to the problem. Because of this evolution from awareness, to individual action, to the development of local experts, and to a desire to collaborate more fully, we expect that results from this sample can be generalized to other groups at early stages of network formation.

Milward and Provan (2006) identify four types of networks: service implementation, information distribution, problem solving, and community capacity building. Based on the description provided above, the network that this study focused on is primarily an informa- tion distribution network in the formative stage. At a second meeting, which was attended by 12 from the initial group plus an additional 13, we reviewed the results of the initial data collection and formed an action list of items that indicated that the group wanted to move beyond the information distribution role. As action researchers, we play both an insider’s and outsider’s role to this process. We are outsiders in that we are not social workers and do not treat clients. We are insiders in that we are taking an active role in many of the action items identified by taking on the role of network coordinator (a role that is currently lacking). Because we have become part of the network, we will be included in future social network surveys so that our own roles in the process can be examined.

Depending on the interaction, social network data represents either the presence or absence of a relationship or the degree of a relationship between each pair within a net- work. As a result, the observations violate the assumption of normality required by most linear statistical modeling techniques. In addition, the dependent variable, in this case whether respondent i (the trustor) indicates that she has a trusting relationship with respon- dent j (the trustee), is binary. Therefore, to characterize the relationship between the depen- dent (response) variable and a set of independent (predictor) variables, we instead model the probability of a tie using a vectorized form of logistic regression. This analysis was per- formed using the netlogit function in the Social Network Analysis package for the R statis- tical programming language (http://www.R-project.org). Briefly, this function performs a logistic regression of the network response variable on a set of network predictor variables and estimates the model parameters (coefficients). To determine the significance of the resulting fits and coefficient estimates, they are compared to estimates generated by repeated random relabeling of the vertices of the predictor networks (Quadratic Assignment

Lambright et al. / Building Trust in Public and Nonprofit Networks 7

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

Procedure). This procedure yields the proportions of randomly determined coefficients that are greater than or equal to (and less than or equal to) the observed value of the coefficients from the model. The netlogit function requires that all data be in square matrices corre- sponding to each dyadic relationship. The variable names, coding schemes, and survey questions used in this analysis can be found in Table 1.

In addition to the variables directly measured by the survey, three third-party relation- ship variables were constructed from the dependent variable (Trust) and the Interaction variable. Following Ferrin et al. (2006), network closure (Closure) was calculated using the Interaction variable. First, the values were symmetrized so that every cell was the average of Xij and Xji. Because the Interaction variable is valued, we then dichotomized the variable so that every Xij ≥0.5 was coded as 1 and every Xij <0.5 as 0. By multiplying this matrix by itself we were able to determine for every dyad the number of individuals who commu- nicated with both the trustor (i) and the trustee (j).

8 The American Review of Public Administration

Table 1 Model Variable Descriptions

Variable Name Coding Survey Question

Relationship between trustor and trustee

Past 0 = Did not select respondent Indicate with whom you have successfully 1 = Selected respondent worked together to achieve a common

goal in the past. Interaction 0 = Never Estimate how often you communicated

1 = 1 to 5 times (rarely) with the following people over the past 2 = 6 to 12 times (every other month) year. Please include all communication, 3 = 13 to 24 times (monthly) not just interactions related to work. 4 = 25 to 50 times (2 to 3 times per Interactions may include personal

month) contact, e-mail, phone calls, and meetings. 5 = More than 50 times (weekly)

Perceived trustworthiness of trustee

Expertise 0 = Did not select respondent These people have expertise related to 1 = Selected respondent self-injury.

Interests 0 = Did not select respondent Please indicate those who you think will 1 = Selected respondent act in your best interests.

Values 0 = Did not select respondent Please indicate those who you think have 1 = Selected respondent values and beliefs similar to yours.

Dependent variables Trust 0 = Did not select respondent Indicate those individuals with whom you

1 = Selected respondent have a trusting relationship. Please note that by leaving a individual’s name blank, you are not indicating that he or she is “untrustworthy.”

Future 0 = Did not select respondent Please indicate with whom you expect to 1 = Selected respondent work together to achieve a common

goal in the future.

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

Structural equivalence (Equivalence) was also calculated using the Interaction matrix. Using the correlation method (CONCOR) in UCINet, we formed a matrix of correlation coefficients for each dyad. UCINet is one of the most commonly used software packages for conducting social network analysis and can be used to generate centrality measures, subgroup identification, role analysis, elementary graph theory, and permutation-based statistical analysis (Borgatti, Everett, & Freeman, 2002). In our matrix, each cell represents the extent to which i’s level of communication with the other members of the network is correlated with j’s level of communication.

The process used to create trust transferability (Transferability) was similar to network closure. But because we are interested in the directional relationship, we did not sym- metrize the matrix. In addition, because the relation is already dichotomous, we did not need to dichotomize it. By multiplying the Trust matrix by itself, we were able to calculate the number of individuals who trusted j and were trusted by i.

The model suggested by the literature contains a number of intervening relationships. Therefore, we conducted a pseudo-path analysis following Kadushin (1995). The analysis is a “pseodu-path” because the binary nature of the dependent variables prohibits the use of OLS estimators. Likewise, we test for mediating effects following Baron and Kenny (1986). Because there are more variables than the proposed relations between them, model identification is not an issue.1

The probability that i trusts j is a function of the perceived trustworthiness of j (Hypothesis 1), which is a function of the frequency of interaction between i and j (Hypothesis 4), whether i and j cooperated successfully in the past (Hypothesis 5), and trust transferability (Hypothesis 9). The probability that i indicates that he had a successful past cooperation with j is related to the frequency of interaction (Hypothesis 6), network closure (Hypothesis 7), and structural equivalence (Hypothesis 8). To test this model with all of its intervening rela- tionships, a full model for Trust was estimated first with all predictor variables:

Equation 1: Trust = f(Expertise, Interest, Values, Past, Interaction, Transferability)

Then Trust was regressed against only the perceived trustworthiness variables. In a sep- arate set of equations, trust was regressed against the rest of the predictor variables without the perceived trustworthiness variables. Finally, each of the perceived trustworthiness vari- ables was separately regressed against the rest of the predictor variables.

Equation 2: Trust = f(Expertise, Interest, Values) Equation 3: Trust = f(Transferability, Past, Interaction) Equation 4: Expertise = f(Transferability, Past, Interaction) Equation 5: Interest = f(Transferability, Past, Interaction) Equation 6: Values = f(Transferability, Past, Interaction)

If the variables omitted from Equation 2 exert an indirect effect on trust through the per- ceived trustworthiness variables, one would expect (a) the coefficients of the variables omitted from Equation 2 to be significant in Equations 4 to 6, (b) these variables to have a significant impact on Trust without the perceived trustworthiness variables and therefore be significant in Equation 3, and (c) for the significance and/or magnitude of the coefficients to be greater in Equations 4 to 6 than Equation 1 (Baron & Kenny, 1986).

Lambright et al. / Building Trust in Public and Nonprofit Networks 9

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

A similar procedure is followed to test whether past cooperation or the frequency of interaction plays an intervening role.

Equation 7: Expertise = f(Transferability, Past, Interaction, Closure, Equivalence) Equation 8: Interest = f(Transferability, Past, Interaction, Closure, Equivalence) Equation 9: Values = f(Transferability, Past, Interaction, Closure, Equivalence) Equation 10: Expertise = f(Closure, Equivalence) Equation 11: Interest = f(Closure, Equivalence) Equation 12: Values = f(Closure, Equivalence) Equation 13: Past = f(Closure, Equivalence) Equation 14: Interaction = f(Past)

Finally, whether trust between i and j leads to willingness to cooperate in the future (Hypothesis 10) is tested by the following equation:

Equation 15: Future = f(Past, Trust).

Because propensity of i to trust is an attribute of the respondent and not a relational variable, it could not be included in the netlogit model, and Hypotheses 2 and 3 were not directly tested. However, we analyze the difference between the predicted and actual values of the model in the next section. This analysis provides some interesting insights into an individual’s propensity to trust and suggests that an individual’s propensity to trust may be an important omitted variable from our netlogit model.

Findings

This section presents findings from our social network analysis of 37 individuals collabo- rating to address the problem of adolescent self-injury. We first investigated whether a relation- ship exists between perceived trustworthiness and the likelihood that a trustor (i) will trust a trustee (j). Table 2 presents the results from all of the logistic regression equations.

For all three measures of perceived trustworthiness—Expertise, Interest, and Values—a positive and significant relationship exists with Trust.

Once we established the relationship between a trustee’s perceived trustworthiness and trust, we tested a series of hypotheses relating to the impact of the relationship between the trustor and trustee. We specifically examined whether either the frequency of interactions or successful past cooperation affects a trustee’s perceived trustworthiness as well as whether successful past cooperation affects the frequency of interactions.

The frequency of interactions is positively related to all three measures of perceived trust- worthiness. Interestingly, for all three measures, the size of the coefficient is smaller than when the frequency of interactions is used as a direct predictor of Trust, indicating that the frequency of interactions is actually a result of perceived trustworthiness rather than a predic- tor of it. We confirmed that perceived trustworthiness affects the frequency of interactions using logistic regression. On the other hand, successful past cooperation, although also a sig- nificant predictor of all three perceived trustworthiness measures, has a smaller effect on Trust directly than for two of the three perceived trustworthiness variables—Values and Interest.

10 The American Review of Public Administration

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

11

T ab

le 2

M od

el R

es u

lt s

E qu

at io

n E

qu at

io n

E qu

at io

n E

qu at

io n

E qu

at io

n E

qu at

io n

E qu

at io

n E

qu at

io n

E qu

at io

n E

qu at

io n

E qu

at io

n E

qu at

io n

E qu

at io

n E

qu at

io n

E qu

at io

n V

ar ia

bl e

1 2

3 4

5 6

7 8

9 10

11 12

13 14

a 15

T ru

st D

V D

V D

V 2.

35 (

.0 0)

E xp

er ti

se 2.

15 (

.0 0)

2. 83

( .0

0) D

V D

V D

V F

ut ur

e D

V In

te re

st 1.

60 (

.0 5)

2. 61

( .0

0) D

V D

V D

V V

al ue

s 1.

33 (

.1 1)

2. 82

( .0

0) D

V D

V D

V P

as t

2. 17

( .0

0) 3.

35 (

.0 0)

2. 13

( .0

0) 3.

22 (

.0 0)

3. 86

( .0

0) 2.

20 (

.0 0)

3. 33

( .0

0) 4.

02 (

.0 0)

D V

2. 08

( .0

0) 3.

59 (

.0 0)

In te

ra ct

io n

0. 62

( .0

0) 0.

79 (

.0 0)

0. 21

( .1

0) 0.

49 (

.0 0)

0. 53

( .0

0) 0.

08 (

.5 7)

0. 49

( .0

0) 0.

40 (

.0 2)

D V

T ra

ns fe

ra bi

li ty

0. 49

( .0

0) 0.

47 (

.0 0)

0. 14

( .2

1) 0.

05 (

.6 9)

0. 10

( .3

9) 0.

11 (

.4 5)

0. 10

( .5

2) 0.

13 (

.3 3)

E qu

iv al

en ce

0. 87

( .1

3) 0.

27 (

.7 2)

1. 24

( .0

7) 1.

61 (

.0 0)

1. 79

( .0

1) 2.

29 (

.0 0)

1. 09

( .0

1) C

lo su

re 0.

01 (

.9 4)

–0 .0

6 (.

55 )

–. 09

( .3

0) 0.

15 (

.0 4)

0. 17

( .0

5) 0.

16 (

.0 3)

0. 29

( .0

0) In

te rc

ep t

–5 .7

5 (.

00 )

–3 .5

4 (.

00 )

–5 .2

3 (.

00 )

–3 .9

3 (.

00 )

–4 .0

2 (.

00 )

–4 .7

8 (.

00 )

–3 .9

6 (.

00 )

–3 .9

0 (.

00 )

–4 .6

3 (.

00 )

–3 .3

4 (.

00 )

–2 .7

0 (.

00 )

–2 .8

7 (.

00 )

–2 .4

6 (.

00 )

0. 25

( .0

0) –3

.1 8

(. 01

) P

se ud

o R

2 0.

83 0.

74 0.

76 0.

67 0.

67 0.

73 0.

68 0.

67 0.

73 0.

62 0.

46 0.

50 0.

35 0.

41 0.

67 %

C or

re ct

94 .9

% 93

.2 %

92 .8

% 90

.3 %

89 .8

% 91

.1 %

90 .2

% 90

.5 %

91 .9

% 90

.5 %

84 .8

% 85

.7 %

77 .6

% N

A 92

.2 %

N ot

e: V

al ue

s ar

e B

(s ig

.) . D

V =

d ep

en de

nt v

ar ia

bl e.

N A

= no

t ap

pl ic

ab le

. a.

E st

im at

ed u

si ng

t he

n et

lm p

ro ce

du re

i n

R ,w

hi ch

u se

s O

L S

r eg

re ss

io n,

in st

ea d

of t

he n

et lo

gi t

pr oc

ed ur

e be

ca us

e th

e de

pe nd

en t

va ri

ab le

i s

bi na

ry . T

he re

fo re

,t he

c oe

ff ic

ie nt

i n

th is

m od

el i

s no

t co

m pa

ra bl

e to

t ho

se i

n th

e ot

he r

m od

el s.

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

This provides support for Hypothesis 5 and suggests that we learn about others’ values and beliefs through successful past cooperation and form an opinion as to whether they will act in our best interests. In addition, our analysis confirms that the quality of interactions between a trustor and a trustee is positively related to the frequency of their interactions.

To explore the role that social networks play in the development of interpersonal trust, we also tested a series of hypotheses relating to the impact of third-party relationships.

Network closure is both a positive and significant indicator of successful past coopera- tion. Likewise, it is a predictor of all three perceived trustworthiness variables. However, both the magnitude and significance of Closure is greater when used as a predictor of Past than of the perceived trustworthiness variables directly. This suggests that Past is a mediat- ing variable and that network closure influences whether the trustor and trustee have coop- erated successfully in the past, but it is the successful past cooperation that leads to the perceptions of trustworthiness. Structural equivalence also has both a positive and signifi- cant impact on successful past cooperation. But unlike network closure, Equivalence has a stronger direct effect on the three trustworthiness variables than on successful past cooper- ation. Based on these results, past cooperation does not mediate the relationship between structural equivalence and perceived trustworthiness. Instead, structural equivalence appears to have a direct, positive impact on whether a trustor perceives a trustee as trustworthy. This suggests that individuals are more likely to perceive individuals who have similar social networks as trustworthy regardless of whether they have successfully cooperated in the past. Also contrary to our expectations, trust transferability is not a significant predictor of any of the three measures of a trustee’s perceived trustworthiness. However, Transferability is a positive and significant predictor of Trust directly.

Finally as part of our netlogit model, we explored the consequences of trust. We specifi- cally examined whether there was a relationship between trust and future cooperation.

As expected, whether a trustor identifies a trustee as trustworthy has a positive and sig- nificant impact on whether the trustor indicates that she is willing to cooperate with the trustee in the future. This finding provides empirical evidence of an important tangible benefit of trusting relationships.

Because propensity of a trustor to trust is an attribute of the respondent and not a relational variable, it could not be included in the netlogit model. However, to explore the impact of a trustor’s propensity to trust on the development of interpersonal trust, we compared the predicted values of trust from Equation 1 to the actual values.

A careful analysis of the difference between the actual and predicted values of Trust suggests that propensity to trust is an important variable. Figure 2 depicts the errors in pre- dictions for trusting relationships by network member. As illustrated by this figure, there was at least one error in predicting which relationships would be trusting for 29 of the 37 individuals in the sample. The space between each small black hash mark in Figure 2 represents the prediction errors for one network member. The light gray bars indicate the number of false-positive predictions for a particular network member, and the dark gray bars indicate the number of false-negative predictions for a particular network member. As an example, the first network member listed in Figure 2 had six false-negative predictions but no false-positive predictions. For 22 of these individuals, the prediction errors were either false negatives (the equation predicted that individual i would not choose individual j when in fact he did) or false positives (the equation predicted that individual i would

12 The American Review of Public Administration

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

choose individual j when he did not). That is, the equation has a tendency to predict some- one is less trusting or more trusting than he is rather than to make random errors in predic- tion. This result suggests that an individual’s propensity to trust may indeed be an important omitted variable from our netlogit model.

Discussion

This article examines what factors influence interpersonal trust in a network composed of public and nonprofit service providers. The theoretical model developed in this article identifies propensity of the trustor to trust, the perceived trustworthiness of the trustee, the relationship between the trustee and trustor, and third-party relationships as influencing interpersonal trust. Based on our analysis, we find empirical evidence supporting Hypotheses 1, 5, 6, 7, and 10: (a) a trustor is more likely to trust a trustee if the trustor per- ceives the trustee as trustworthy; (b) successful past cooperation between a trustor and a

Lambright et al. / Building Trust in Public and Nonprofit Networks 13

False Positives False Negatives

6 4 2 0 2 4 6 8

Figure 2 Errors in Predictions for Trusting Relationships by Network Member

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

trustee increases the likelihood the trustor will perceive the trustee as trustworthy; (c) a trustor and a trustee are likely to more frequently interact if they have successfully collab- orated in the past; (d) as the number of third parties who interact with both the trustor and trustee increases, the likelihood that the trustee and trustor have successfully collaborated increases; and (e) trust between the trustor and trustee has a positive impact on expected future cooperation. These findings are consistent with the work of a variety of other schol- ars including Edelenbos and Klijn (2007), Ferrin et al. (2006), Mayer et al. (1995), Ring and Van De Ven (1992), Van Slyke (2006), and Vangen and Huxham (2003). The last find- ing supporting Hypothesis 10 is perhaps the most important because it highlights why trust matters in networks: trusting relationships facilitate cooperation in networks. Trust acts as the glue that holds networks together, enabling networks to function effectively even though they lack a hierarchal power structure.

Furthermore, our analysis suggests that although we did not find empirical evidence sup- porting Hypotheses 4, 8, and 9, frequency of interactions, structural equivalence, and trust transferability do influence the development of interpersonal trust. These variables just do not affect the development of interpersonal trust through the pathways that we originally hypothesized. Instead of influencing a trustee’s perceived trustworthiness, frequency of interactions between the trustor and trustee has a direct, positive impact on whether the trustor trusts the trustee. In addition, trust transferability has a direct, positive impact on whether a trustor trusts a trustee. In other words, as the number of third parties who trust the trustee and are also trusted by the trustor increases, the likelihood that the trustor will trust the trustee increases. Finally, contrary to Ferrin et al. (2006), we do not find that structural equivalence encourages cooperative behavior. Rather than increasing the likelihood of past successful cooperation, we find that structural equivalence has a direct positive impact on a trustee’s perceived trustworthiness: as the similarity of their network interactions increases, the likelihood that the trustor perceives the trustee as trustworthy increases. Although we were not able to assess the impact of a trustor’s propensity to trust as part of our netlogit model, a careful analysis of the difference between the actual and predicted val- ues suggests that propensity to trust is another key variable influencing interpersonal trust in networks. These findings are consistent with Mayer et al.’s theoretical model (1995) of interpersonal trust development.

Figure 3 depicts our revised model of the factors influencing interpersonal trust modi- fied based on our empirical findings. This model highlights the important role that relation- ships play in interpersonal trust. Individuals in networks consider their personal experiences and the interactions they have had with a network member when determining whether to trust this person. Third-party relationships also affect trust. Trust transferability has a direct impact on whether an individual will trust another network member; structural equivalence and network closure exert their impact indirectly by influencing perceptions of trustworthiness and successful past cooperation, respectively. Finally, the importance of trust in networks is underscored by the fact that it influences whether one individual would be willing to cooperate with another in the future.

Even though these data were collected at a single period in time, they provide some insight as to how trust evolves in a network and possibly, how that process could be accel- erated through outside intervention. As discussed earlier, this group of social workers has a long history of coming together around the issue of self-injury in adolescents. However,

14 The American Review of Public Administration

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

the agencies for which they work do not exclusively serve clients engaging in self-injuring behavior but rather treat different types of clients with varying problems. Therefore, the social workers have had ample opportunity to communicate and collaborate on many different types of issues, some of which were related to self-injury and some of which were not related to this topic. These interactions created informal networks of individuals that evolved over time.

Suppose Bob, Teresa, Nancy, and Frank are all social workers and members of the adoles- cent self-injury network that was the focus of this study. According to the revised model depicted in Figure 3, the evolution of trust takes the following hypothetical form in our research context. Both Bob and Teresa interact with Nancy and have referred adolescents engaged in self-injuring behavior to Nancy for treatment (network closure). As a result of their mutual link with Nancy, Bob and Teresa are motivated to successfully cooperate by orga- nizing a training session for other network members. Because of their interactions with the same set of individuals (structural equivalence) and their cooperation, Bob and Teresa per- ceive each other to be trustworthy. These perceptions cause them to interact more frequently and develop a trusting relationship. Because Bob now trusts Teresa, this increases the chance that Frank, who already trusts Bob, will also trust Teresa (trust transferability). Because Frank trusts Teresa, he may be willing to cooperate with her in the future to conduct a community needs assessment on the problem of adolescent self-injury. This may bring Frank in contact with Nancy, which allows them to start the same trust-building process.

Over a period of time, these processes allow trust to build within a network. But because both network closure and structural equivalence measure the extent to which there are shared interactions with others, they can tend to reinforce existing bonds without necessar- ily creating new ones. This can lead to the creation of cliques within networks and a slower diffusion of trust as these cliques become more insular. What an outside influence can pro- vide to this process is to create opportunities for cooperation between individuals who do

Lambright et al. / Building Trust in Public and Nonprofit Networks 15

Third party relationships

Relationship between trustor and trustee

Network closure

Structural equivalence

Trust transferability

Propensity of trustor to trust

Perceived trustworthiness of trustee

1. Perceived ability 2. Perceived benevolence 3. Perceived integrity

Trustor trusts trustee

Frequency of interactions Successful past cooperation

Expected future cooperation

Figure 3 Revised Model of Factors Influencing Interpersonal Trust in Networks

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

not normally interact. Ensuring diversity of involvement is particularly important in the context of public sector networks given representativeness is a core value for public admin- istrators (Kaufman, 1969; Mosher, 1968). Moreover, intervening early in network develop- ment may also assist network members in developing more accurate perceptions of trustworthiness based on knowledge of each other gained through cooperation rather than by adopting the perceptions of others.

Although our analysis provides valuable insights into factors influencing interpersonal trust in public and nonprofit networks, there are some limitations to our research. Our data was collected from one network at a single period in time. Also, as detailed in the Methodology section, this study focused on an information distribution network at the for- mative stage. As a result, this study’s findings may not apply to better established networks or networks created for other purposes.

Much more research is needed on the development of interpersonal trust in networks composed of public and nonprofit service providers. Our analysis demonstrates that both relationships between a trustor and trustee and third-party relationships affect interpersonal trust. Models of interpersonal trust that focus on this phenomenon as developing in isolated, dyadic relationships are ignoring the social complexities of organizational life. To have a complete picture of interpersonal trust development, future research should consider the impact of third-party relationships.

It would be interesting to explore in future research whether the revised model that is proposed at the end of this article is applicable to other types of networks such as service implementation, problem-solving, and community capacity-building networks. For example, past cooperation may play a different role in the development of interpersonal trust in dif- ferent types of networks given that the activities involved in past collaborative efforts are likely to be different.

Another area for future research is to examine other ways in which the relationship between the trustor and trustee may influence the development of interpersonal trust. When considering how the relationship between the trustor and trustee may influence the devel- opment of interpersonal trust, this article specifically focused on successful past coopera- tion and the frequency of interactions as our relationship measures. However, the intention of interactions may also be important and should be considered in future research.

In addition, longitudinal research on interpersonal trust development is essential. At present, very few studies have examined networks using longitudinal data. As one of the few exceptions, Provan, Nakama, Veazie, Teufel-Shone, and Huddleston (2003) find that trust in a network of health and human service organizations actually decreased over time. The model we propose in this article is static. However, as Vangen and Huxham (2003) and McGuire (2006) highlight, the development of interpersonal trust is a dynamic phenome- non, with trust promoting collaboration and collaboration promoting trust. In our model, there is likely to be a feedback loop between expected future cooperation and past success- ful cooperation. Longitudinal data would enable the testing of a more dynamic model of interpersonal trust. Future research should also explore whether models of interpersonal trust are different for networks at different stages in their development. The answers provided by this research will further understanding of the role trust plays in networks of public and nonprofit service providers as well as further understanding of mechanisms fostering trust in these networks.

16 The American Review of Public Administration

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

Note 1. Model identification is not an issue regardless of whether propensity to trust is included.

References

Agranoff, R. (2007). Managing within networks: Adding value to public organizations. Washington, DC: Georgetown University Press.

Agranoff, R., & McGuire, M. (1999). Managing in network settings. Policy Studies Review, 16, 18-41. Agranoff, R., & McGuire, M. (2001). Big questions in public network management research. Journal of Public

Administration Research and Theory, 11, 295-326. Alter, C., & Hage, J. (1993). Organizations working together. Newbury Park, CA: Sage. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological

research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182.

Bohnet, I., & Huck, S. (2003). Repetition and reputation: Implications for trust and trustworthiness in the short and in the long run. Social Science Research Network Web site. Retrieved March 4, 2004, from http://papers .ssrn.com/sol3/papers.cfm?abstract_id=474544#PaperDownload

Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). Ucinet for Windows: Software for social network analysis. Harvard, MA: Analytic Technologies.

Burt, R. S., & Knez, M. (1996). Trust and third-party gossip. In R. M. Kramer & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and research (pp. 68-89). Thousand Oaks, CA: Sage.

Butler, J. K., Jr. (1983). Behaviors, trust, and goal achievement in a win-win negotiating role play. Group and Organization Management, 20, 486-501

Cross, R., & Parker, A. (2004). The hidden power of social networks: Understanding how work really gets done in organizations. Boston: Harvard Business School Press.

Edelenbos, J., & Klijn, E. H. (2007). Trust in complex decision-making networks: A theoretical and empirical exploration. Administration & Society, 39, 25-50.

Ferrin, D. L., Dirks, K. T., & Shah, P. P. (2006). Direct and indirect effects of third-party relationships on inter- personal trust. Journal of Applied Psychology, 91, 870-883.

Goffman, E. (1959). The presentation of self in everyday life. Garden City, NY: Doubleday Anchor. Granovetter, M. (1985). Economic action and social structure: The problem of embeddedness. American

Journal of Sociology, 91, 481-510. Gulati, R. (1995). Does familiarity breed trust? The implications of repeated ties for contractual choice in

alliances. Academy of Management Journal, 38, 85-112. Isett, K. R., & Provan, K. G. (2005). The evolution of dyadic interorganizational relationships in a network of

publicly funded nonprofit agencies. Journal of Public Administration Research and Theory, 15, 149-165. Kadushin, C. (1995). Friendship among the French financial elite. American Sociological Review, 60, 202-221. Kapucu, N. (2006). Interagency communication networks during emergencies: Boundary spanners in multia-

gency coordination. The American Review of Public Administration, 36, 207-225. Kaufman, H. (1969). Administrative decentralization and political power. Public Administration Review, 29, 3-15. Kramer, R. M. (1999). Trust and distrust in organizations: Emerging perspectives, enduring questions. Annual

Review of Psychology, 50, 569-598. Lane, C., & Bachman, R. (Eds.). (1998). Trust within and between organizations: Conceptual issues and empirical

applications. Oxford, UK: Oxford University Press. LaPorte, T. R. (1996). Shifting vantage and conceptual puzzles in understanding public organization networks.

Journal of Public Administration Research and Theory, 6, 49-74. Lewicki, R. J., & Bunker, B. B. (1996). Developing and maintaining trust in work relationships. In R. M. Kramer

& T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and research (pp. 114-139). Thousand Oaks, CA: Sage.

Lambright et al. / Building Trust in Public and Nonprofit Networks 17

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

McAllister, D. J. (1995). Affect- and cognition-based trust as foundations for interpersonal cooperation in orga- nizations. The Academy of Management Journal, 38, 24-59.

McGuire, M. (2006). Collaborative public management: Assessing what we know and how we know it. Public Administration Review, 66, 33-42.

Malhotra, D., & Murnighan, J. K. (2002). The effects of contracts on interpersonal trust. Administrative Science Quarterly, 47, 534-559.

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. The Academy of Management Review, 20, 709-734.

Milward, H. B. (1996). Symposium on the Hollow State: Capacity, control, and performance in interorgani- zational settings. Journal of Public Administration Research and Theory, 6, 193-195.

Milward, H. B., & Provan, K. G. (1998). Principals for controlling agents: The political economy of network structure. Journal of Public Administration Research and Theory, 8, 203-221.

Milward, H. B., & Provan, K. G. (2000). Governing the hollow state. Journal of Public Administration Research and Theory, 10, 359-379.

Milward, H. B., & Provan, K. G. (2006). A manager’s guide to choosing and using collaborative networks. Washington, DC: IBM Center for the Business of Government.

Mosher, F. (1968). Democracy and the public service. New York: Oxford University Press. Musso, J. A., Weare, C., Oztas, N., & Loges, W. E. (2006). Neighborhood governance reform and networks of

community power in Los Angeles. The American Review of Public Administration, 36, 79-97. O’Toole, L. J. (1997a). The implications for democracy in a networked bureaucratic world. Journal of Public

Administration Research and Theory, 7, 443-459. O’Toole, L. J. (1997b). Treating networks seriously: Practical and research-based agendas in public administra-

tion. Public Administration Review, 57, 45-52. O’Toole, L. J., & Meier, K. J. (2004). Desperately seeking Selznick: Cooptation and the dark side of public

management in networks. Public Administration Review, 64, 681-693. Pillai, R., Schriescheim, C. A., & Williams, E. S. (1999). Fairness perceptions and trust as mediators for trans-

formational and transactional leadership: A two-sample study. Journal of Management, 25, 897-933. Powell, W. W. (1990). Neither market nor hierarchy: Network forms of organization. In B. Staw & L. Summings

(Eds.), Research in organizational behavior (pp. 295-336). Greenwich, CT: JAI Press. Provan, K. G., Nakama, L., Veazie, M. A., Teufel-Shone, N. I., & Huddleston, C. (2003). Building community

capacity around chronic disease services through a collaborative interorganizational network. Health Education & Behavior, 30, 646-662.

Provan, K. G., & Milward, H. B. (1995). A preliminary theory of interorganizational network effectiveness: A comparative study of four community mental health systems. Administrative Science Quarterly, 40, 1-33.

Ring, P. S., & Van de Ven, A. H. (1992). Structuring Cooperative Relationships between Organizations. Strategic Management Journal, 13, 483-498.

Rotter, J. B. (1971). Generalized expectancies for interpersonal trust. American Psychologist, 26, 443-452. Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline

view of trust. The Academy of Management Review, 23, 393-404. Uzzi, B. (1997). Social structure and competition in interfirm networks: The paradox of embeddedness.

Administrative Science Quarterly, 42, 35-67. Van Slyke, D. M. (2006, September). The role of collaboration in relational contracting. Paper presented at the

Maxwell Conference on Collaborative Public Management, Washington, DC. Vangen, S., & Huxham, C. (2003). Nurturing collaborative relations: Building trust in interorganizational col-

laboration. Journal of Applied Behavioral Science, 39, 5-31. Zaheer, A., McEvily, B., & Perrone, V. (1998). Does trust matter? Exploring the effects of interorganizational

and interpersonal trust on performance. Organization Science, 9, 141-159.

Kristina T. Lambright is an assistant professor of public administration at Binghamton University's College of Community and Public Affairs. Her research interests include service delivery structure, privatization and contracting, networks, organizational ownership, and service learning.

18 The American Review of Public Administration

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from

Pamela A. Mischen is an assistant professor of public administration and co-director of the Center for Applied Community Research and Development at Binghamton University. Her primary research interest is the inter- section of organization theory and policy implementation. Recent work includes the application of action research, network analysis, and complexity theory to implementation research.

Craig B. Laramee is an assistant professor of bioengineering in the Thomas J. Watson School of Engineering and Applied Science at Binghamton University. His research interests include bioelectromagnetics and the analysis of social and biological networks.

Lambright et al. / Building Trust in Public and Nonprofit Networks 19

at TEMPLE UNIV on May 30, 2016arp.sagepub.comDownloaded from