Patterns of Trust and Collaboration among Nonprofit Organizations and Health Funds: A Case Study

The importance of collaboration between organizations, especially in the modern world, has been discussed extensively by researchers from different fields. Yet, the importance of the context, trust dynamics, and the employment social environment, such as the interplay among these factors, i.e., trust, individual behavior, and political behavior, has been less studied. This study evaluates the role of trust in and between organizations on successful collaboration processes. Using qualitative methodology, we interviewed 11 senior directors who were involved in a specific case-study of collaboration among four major organizations as well as direct observation, documentation, and archive records. Our findings emphasize the importance of analyzing multilevel trust, interpolitics, and intrapolitics, even when success is at stake. We suggest that managers have to account for emotional involvement at the individual level, even when successful organizationallevel collaboration occurs. Overall, we found that there are two aspects of trust in a collaboration process between organizations: system’s aspect and personal aspect. Each aspect is influenced by various factors, mainly different goals and interest and lack of procedures or regulations (from the system’s aspect) and feelings of vagueness in goals and managerial procedures as well as feelings of exploitation (from the personal aspect). In addition, we found that past acquaintances, mutual experience, and shared visions raise the level of trust, which in turn affects the reciprocal relations and therefore the collaboration process resulting in higher social effectiveness for social services.

In recent years, researchers have examined various elements that affect the success of organizational collaboration, e.g., vision, commitment, power, trust, and partner characteristics (Hatzakis & Searle, 2006;Lundin, 2007;Raymond, 2006). For instance, Jacobson and Choi (2008) found that commitment and shared vision are important for constructing success, as is pairing of open communication and trust (Ekanayke, 2008). Yet the importance of the context (Johns, 2006), trust dynamics (Capell, Tzafrir, Enosh, & Dolan, 2017), and the employment social environment (Tzafrir, Gur, & Blumen, 2015), such as the interplay among these factors, i.e., trust, individual behavior, and political behavior, has been less studied. McDonald, Jayasuriya, and Harris (2012) studied providing or receiving care and found that trust plays a key role in collaboration processes. Yet the important role of the upper administrative echelon (Hambrick & Mason, 1984) and managers (Long, Cunningham, Carswell, & Braithwaite, 2014) in the collaboration process has been neglected. In the field of nonprofit organization studies, several authors have suggested that trust plays a key role in collaboration (Snavely & Tracy, 2002;Vangen & Huxham, 2003), whereas others argued that different constructs play an essential role (Gazley, 2010;Gazley & Brudney, 2007); therefore, a more precise in-depth study is warranted. The present case study sought to determine whether trust among collaborating members (at different levels) has any effect on participants' views of the collaboration process and its outcomes.

Collaboration between Organizations
A strategic alliance is any voluntary interorganizational cooperative agreement (functional, material, or both) involving exchange, sharing, or co-development (Gulati, 1999). The modern marketplace demands that organizations not only respond to customers' demands but also extend commitments to meet potential clients' envisaged needs. Intersystemic collaboration may enhance competitive stability (MacDonald, 2009;Nix, Lusch, Zacharia, & Bridges, 2008). Studies have shown that collaboration merits exceed its cost, defined in financial, valuable, ethical, or social terms (Arya & Lin, 2007;Raymond, 2006;Shaw, 2003;Vigoda, 2002). Gazley and Brudney (2007) found essential issues such as service and quality increased after nonprofit organizational collaboration (Jang & Feiock, 2007;Laville & Nyssens, 2000). For instance, Snavely and Tracy (2002) noted the importance of trust in making collaboration work. Thomson and Perry (2006) mentioned trust and reciprocity as important factors that build social capital norms for successful collaboration processes. The third-sector organization's resources and goals differ from, and thus may complement, those of public organizations (Jang & Feiock, 2007). Perhaps this explains why the role of NPOs in delivering critical services has grown significantly during the past three decades (Gidron, Bar, & Katz, 2004;Jang & Feiock, 2007;Thompson, Tancredi, & Kisil, 2000).
Collaboration between NPOs and public organizations is more than important. Organizations feel threatened by loss of identity and control (Kumar & Ghadially, 1989;Samuel, 2004) and unbalanced growing expenses (Thompson et al., 2000). Likewise, collaboration between Health Funds (HFs) and NPOs may lead to a more fruitful health system with improved vision and better social justice. However, studies have shown difficulties on the way to successful collaboration (Raymond, 2006).

Political Behavior in Organizations
Politics is a fundamental activity in organizations, reflected in political relationships, agenda, and affiliations (Kumar & Ghadially, 1989;Vigoda-Gadot, 2007). Organizational politics is an inseparable part of organizational reality and therefore invites consideration. Organizational political activity may limit communications, control access (to information, individuals, or both), and neutralize change or opposition to change (Kumar & Ghadially, 1989). Alternatively, it may facilitate decision-makers' abilities to promote collaborative processes (Lundin, 2007;Raymond, 2006;Samuel, 2004). Political acts or groups can change and twist the results of a process or influence the strategic power of individuals or organizations guarding their own interests.
Perspective organizational politics may be the main concept for understanding the structure and behavior of organizations (Lundin, 2007;Samuel, 2004). As Vigoda-Gadot (2007) stated, understanding the advantages and disadvantages of different leadership styles may affect organizational politics and eventually affect organizational strategy and performance. Another important element is interpersonal politics. Interpersonal politics play a major role in the success of an organization. Samuel (2004) explained that interpersonal politics between members of the organization and outsiders is also important.
When collaboration processes are envisaged as part of a multiorganizational strategic work plan, the political behavior and the resulting political members' affiliation may determine the outcome of the collaboration process (Jacobson & Choi, 2008).

Trust
Studies on the subject of trust, and its influence on management, date back to the late 1960s. Gamson (1968) defined trust as "a syndrome of attitudes which can be characterized on a leftright dimension" (p. 40). Driscoll (1978) cited Gamson's declaration that trust is the key variable predicting individual behavior. This notion is similar to Rousseau, Sitkin, Burt, and Camerer's (1998) definition of trust as "a psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another" (p. 395). Driscoll (1978) also explored the usefulness of trust and participation in decision-making as contributors to satisfaction in a work organization. Abramov (2009) demonstrated the importance of trust in complex relationships between several organizations-some governmental, others private or public-emphasizing that trust is an essential factor for the success of fragile situations and collaborations (Williams, 2007). Capell et al. (2017) demonstrated that trust levels in an organization and managers are likely to affect each other. Furthermore, based on a socio-constructivist approach (Shamir & Lapidot, 2003), it is likely that, due to various group and social information processes, individuals may develop a collective assessment of trust in their organizational authorities. Hence, managerial actions directed toward a collaborative project could influence trust levels of the collective group and at the individual level. Fulmer and Gelfand (2012) also highlighted the importance of examining trust at different levels.
Further research investigated various aspects of trust as influencing organizational management or behavior (Yener, 2009), such as the impact on the effectiveness of organizations (Dovey, 2009) and their business success (Ekanayke, 2008). Basically, trust is viewed as an essential ingredient for a healthy personality and as a foundation for interpersonal relationships, cooperation, and stability in social institutions and markets (Lewicki, McAllister, & Bies, 1998). Usually, high-trust individuals are found to exhibit more honest behavior (Rotter, 1980) and are more subject to norms of reciprocity.
Over the last few decades, researchers have increasingly examined trust not as a sole factor but in combination with others. For example, Lundin (2007) studied 203 dyadic collaborations between Swedish public employment service offices and municipalities regarding how resource interdependence, goal congruence, and trust, both separately and integrated, affect joint actions in policy implementation. The author's results show that mutual trust is necessary to increase cooperation between agencies; furthermore, the results imply that trust and goal congruence must exist simultaneously to promote joint actions.
More specifically, Sims, Hewitt, and Harris (2015) studied inter-professional teamwork in healthcare and found that trust is an important variable that has an influence on productivity of team collaboration. In a similar line of research, focusing on teamwork in the Canadian primary care setting, Al Sayah, Szafran, Robertson, Bell, and Williams (2014) explored the importance of trust to the efficiency of interdisciplinary teamwork. Moreover, studies concerning collaboration between organizations in the healthcare industry (Proulx, Hager, & Klein, 2014) emphasized the importance of trust in the success of the collaboration process. Palinkas et al. (2014) found that trust was one of the leadership characteristics associated with "increased levels of exchange of information and resources and with larger and more connected networks" (p. 82).
In addition, multidisciplinary studies have identified social, psychological, economic, and structural determinations of trust. For example, Yener (2009) referred to cooperation and trust as a management concept (structural), and Ekanayke (2008) recognized interfirm trust as an important variable of control (psychological). Therefore, researchers need to account for all aspects of trust and the specific context involved (Agranoff, 2006;Johns, 2006). More specifically, our research refers to the question of why organizations decide to cease successful collaboration regarding all the benefits.

Methodology
This study employed the naturalistic approach of qualitative research, the aim being to understand and describe a social phenomenon from the participant's points of view (Lincoln & Guba, 1985;Patton, 1990). Case study methodology is often used in research involving a process (Denzin & Lincoln, 1994) and is especially appropriate in new topic areas (Eisenhardt, 1989). Theory developed from case study research is likely to have important strength like novelty, testability, and empirical validity, which arise from intimate linkage with empirical evidence (Eisenhardt, 1989). Our research is based on a case study, and the nature of the information collected is mainly descriptive, supported by data collected from archives. Therefore, our findings can be a point of departure for collaboration researchers in developing hypotheses in further research.

Procedures and Analysis Methods
To develop a more holistic picture, we collected data from various sources such as documentation (including protocols, contracts, and reports), presentations, archival records, interviews, and direct observations. We used all sources mentioned but focused on two main ones: interviews and documentation analysis.
Exploratory interviews were conducted to learn about actions, beliefs, and personal and social processes as they took place and as interpreted by the key players. We approached the participants by email or phone and scheduled appointments. One-hour semistructured in-depth interviews were conducted at the participants' offices. The interviews were based on an interview guide that was constructed through an iterative process, which involved literature review and consultation with academic experts, managers, and administrative directors. Following the emergence of new subjects, the interview guide was revised. The final version of the interview guide dealt with clinical, service, economic, organizational, and managerial aspects of the collaboration project. 1 Each interview was recorded and transcribed to allow for later cross-sectional analysis. Participants cooperated willingly, after signing a consent letter approved by the affiliated university's ethics committee. Confidentiality at all levels of publication was secured.
Various documents such as protocols, contracts, and presentations were examined and integrated in the analysis of the findings. Because of the paradigmatic difference between quantitative and qualitative research approaches, Lincoln and Guba (1985) suggested different concepts more applicable to qualitative research for reliability and validity: dependability and trustworthiness. Internal validity was obtained by the participant's approval of his or her presentation, combined with documents analysis to validate interview-based data (Yin, 1994). The use of full documentation of the data and formal protocols helped control for interviewbased inaccuracy or misunderstanding (Lincoln & Guba, 1985).

Participants
Of the 11 participants, three were from the HF, another three were from the NPO, two were from the hospital, and three were also from the municipality (participant's characteristics are presented in Table A1 in the appendix). 2 All participants whom we chose to interview were senior directors in their organization. Although additional personnel were involved in the project (workers), the interviewees were chosen based on the fact that they were responsible, among other issues, for determining policy. Most of them were in the highest position of management and were independent, within reasonable boundaries, in their functional responsibilities. 3

Data Analysis
Analyses of the interviews were done according to the different stages demanded by the methodology of qualitative research: original analysis, categorized analysis, and focused analysis (Denzin & Lincoln, 1994). 4 The focused analysis was based on the categorized analysis; this is a process researchers use to arrange the categories toward a central theme. In the process of defining the main category, researchers use categorized analysis to find the main points of interest in each interview. Some of the categorized elements become more dominant in the focused stage, others less, and some may not even be included in the focused analysis. The process results in defining a main category from which other categories branch out. This process reveals detailed and various elements raised by the interviewees. The basic results were platformed to enable such analyses. Identifying, coding, and categorizing results allows themes to be determined in the best unbiased manner (Patton, 1990).
In addition, various documents such as meeting protocols, contracts, declarations, presentations, reports, etc. were examined and analyzed to obtain more adequate and objective information. Information from the documents was integrated with all other findings.

Israel's Health System
The Israeli context provides researchers and stakeholders interested in the health sector with a suitable arena for studying and analyzing health system structures and patterns specifically, and collaboration projects in the health system in general because it is a Maduradam (miniature copy) of developed countries in Western Europe and North America (Harel & Tzafrir, 1999). Israel's health system features three main suppliers: HFs (responsible for supplying nationwide health services based on a lawfully defined therapeutic package), hospitals (supplying ambulatory and hospitalization services), and nonprofit health organizations (providing treatment and prevention services in both mental and physical health care; Gidron et al., 2004). The government's primary responsibility is the control and development of the health system (Bin Nun, Berlovitz, & Shani, 2005).
Although the responsibility for supplying health services to all citizens lies with the HF, its budget is not specifically predetermined-it is influenced by the global budget, which is composed of citizen's taxes and government allowance, and is allocated to the privatized system only after government approval. The need to fill the gap between what the health system ought to supply with their limited resources is the essence for collaboration between organizations; in our study, we examined reasons, especially the role of trust as a central factor for the failure of a specific collaborative process among several organizations.
The complexity of this system, being that it is meant to be fully privatized but is not completely independent, creates difficulties. HFs would like to supply services according to the public's needs as reflected in the lawfully defined package. They are constrained, however, by limited resources and bureaucracy. The fact that (a) citizens as consumers and service providers have different perceptions of the rendering of services (e.g., in terms of quality, availability, and accessibility) and (b) the system is only partly privatized could, and often does, lead to serious gaps in addressing the public's needs (Arya & Lin, 2007;Nix et al., 2008).
The partial privatization of the health system was intended to close the gap by introducing some basic parameters with regard to proper budgeting and quality of services and clinical treatment. One of the major incentives (Bin Nun et al., 2005) expected to result from the privatization process was increased competition among HFs, consequently raising the level of treatment, while making the system more efficient. The issue of funding in a system wherein monetary sources are diminishing ever since 1995 (Central Bureau of Statistics, 2007), yet the law requires good clinical practice leads to an inherent moral, social, and clinical tension (Laville & Nyssens, 2000) and an unavoidable gap between availability and needs.
This moral and material gap invites social players such as the private sector, the third sector, and citizens themselves to create bridges (Thompson et al., 2000;Vigoda, 2002). The unique features of NPOs allow them to deliver important social services as part of their goals. The deficiencies of the system and the potential of NPOs to complement it suggest that collaboration between the two could be advantageous for the beneficiary (Jacobson & Choi, 2008). [contract] [contract] The Hospital HF Director CNPO SPHMS

Director(N) (MH)Director Initiator
The collaborated project MCC staff

Case Study
This case study involves a situation wherein four major organizations involved in the health system collaborated with one another to supply better developmental and mental health services to children. The organizations were: (a) an HF, a public organization funded by taxes according to the number of members (clients) of the fund; (b) an NPO, a private organization funded by donors; (c) a hospital, specifically the staff of its child developmental institute (CDI), a public organization funded by its activities; and (d) a municipality, a public organization funded by its activities and governmental support. In the case study, we examined a scenario in which one of the HFs was lacking a certified CDI of its own and had to comply with a Ministry of Health regulation to render specific services solely through such an institute. 5 This situation led to establishing a project co-led by an NPO and the municipality, thus creating a team that diagnoses and treats children with developmental delays. Lack of HF resources and services, on one hand, and the new prospect of community medical and psychosocial services, on the other, led the parties to collaborate, with the benefit of the public in mind (Grubbs, 2000). The common aim of the project was pooling of professional and infrastructure resources to augment availability and accessibility of services while eliminating bureaucratic challenges and expenses for children and their families. Furthermore, according to the Israeli law, the HF publicly obligated to render services should take responsibility for these services. In doing so, the HF collaborated with hospital's CDI and the NPO to provide holistic services by pooling resources. Nevertheless, the HF's managers refused to either partake in the project's steering committee or initiate a new one. This study sought to understand why a potentially successful collaboration was halted and the role of trust in the process.
The project had several managerial pitfalls. For instance, it featured an intricate hierarchy; the part-time director of the HF institute was subordinate in this capacity to the director of HF, on one hand, and as a hospital employee to the chief of the hospital's CDI, on the other. Other complex hierarchies existed in the project. 6 Moreover, the NPO compensated the HF for lost income due to loss of members' compulsory fees. The vision for the project, including the professional parties involved, is presented in Figure 1.
This new collaboration model involving an HF, NPO, municipality, and CDI began in January. Teams began working together, and the professional consequences were immediate and mostly positive. Nonetheless, after a full year, the HF unilaterally announced cessation of the project.

Findings
Our analysis related to a central theme, trust affecting a collaboration process that focused on the role of trust as a factor affecting political associations and reciprocal relationships between parties and its effect on the collaboration's outcome. Summarizing the interviews, we organized the findings into two major categories: participants' interpersonal relationships and systemic aspects of the project (see Table A2 in the appendix).

Interpersonal Relationships
Most (nine of 11) interviewees referred to the interpersonal category. The interpersonal category of trust can be divided into two subcategories: personality characteristics and managerial style. Personality characteristics consist of the personal feelings, character traits, and personal beliefs of participants. 7 Such characteristics include ego, exploitation, negative suspicion, grudges, control, and transparency and sharing.
Regarding ego, all managers were sensitive to their self-respect: "Everyone has ego at a different level" (Participant 2); 8 "There are very few people who are willing to give up their power for a joint cause" (Participant 4); "When they sat to work together there were egos involved" (Participant 2). It can be difficult to overcome feelings related to ego, but participants suggested that individuals can suppress them to ensure the success of a process: There are people … who are fully minded [to the project], they manage a very big system without a lot of ego. They want to see results-this is … why, in some organizations, initiatives may succeed, while in others they fail. (Participant 1) This element was stipulated by four main interviewees as one of the most important elements affecting collaboration.
Exploitation is defined as deployment of resources, actions, and investments for personal and organizational interests and opportunities (Foss & Klein, 2012). This feeling was expressed mostly by NPO managers, who stated that the NPO had exceeded all other organizations in its input toward the success of the project; they offered help in settling differences of opinions between parties, solutions for crisis situations, and funding. They described doing much more than needed to ensure the success of the project, but that the other parties, especially the HF, disregarded these efforts. As one of the NPO's managers said: There was no intent for such collaboration … The feeling was that the NPO courted the HF, and the HF took advantage of the situation. These feelings led to mistrust in the partnership.
It is interesting to note that even a central manager outside the NPO, but internal to the system and the process, described the possibility of exploitation. "The HF director said, 'Let's take advantage of [the NPO] and then we will leave him-by then, our system will be all set'" (Participant 5). These feelings and the reality they may have reflected led to suspicion among the parties and in the system as a whole.
Negative suspicions refer to the feeling that a stakeholder has a potential hidden motive for engaging in a behavior, typically negative (Hilton, Fein, & Miller, 1993). That feeling was described by all NPO participants. Deterioration in trust was triggered by the HF's defining the NPO as a service provider and declining to collaborate in research and via the steering committee.
The vague process and noncollaborative sectarianism led to a feeling of wariness among both participants and organizations: "[I felt] a kind of suspicion … [a] lack of understanding of what he is leading up to. What does he want to systemically accomplish?" (Participant 5) Suspicion may rise from difficulty understanding the intricate and holistic vision of the project. Complex systems require sophisticated coping abilities among managers. Rather than discussing differences during joint meetings, some of the parties chose to evade meetings and discussions of the complex situation. Hence, a shared formulation of the new situation did not develop and, therefore, could not lead to compromises alleviating suspicion in the new system.
Grudges refer to "hanging on to negative sentiment and judgments" (Struthers, van Monsjou, Ayoub, & Guilfoyle, 2017, p. 2). They often evolve from a previous negative experience between parties. Managers mentioned "hard feelings" toward others in the project. One manager commented: "It was time consuming in an unnecessary way … It took too many resources; lots of anger and hard feelings" (Participant 1). This participant invested minimal resources and tried to achieve only his own goals. This may be a case in which a partner's view of its own high cost of collaboration leads to grudges.
Control involves a high degree of centralization (Crilly & Sloan, 2014). Some leading managers were willing to cooperate; however, each wanted to maintain systemic control: "There are paranoid feelings in collaboration over loss of control" (Participant 6). Control was linked to power: "There were fights over respect and power" (Participant 1). Control was an inseparable aspect of the managerial style of all directors involved. Consequently, vying for systemic control and the basic definition of professional control led to a destructive process.
Transparency and sharing (openness) is the readiness to be mentally open to new ideas and freely give and share information (Butler, 1991). The managers described a lack of planning, defining goals, delineating strategy, and sharing doubts and problems arising during negotiations. The process and relationships were negatively affected. As one manager said: "You cannot ask me to be a manager of this project and not tell me the economic code … They kept it very close to themselves" (Participant 5).
A manager from a different organization described similar feelings. Lack of communication between parties led to distancing and disrespect: "There was no communication; in different situations, it was all blah blah" (Participant 1).
The third manager, to whom the first two referred, described several occasions on which his organization was not included in the planning: "The first sign was the announcement of the project-it wasn't presented as an integrated project of four bodies but as a mere child development institute run by HF" (Participant 6). Regarding another situation, the same participant mentioned: "I came to meetings with [name] and [name] was already there; they were whispering. I had the feeling that things were being preplanned without including me." Without control measures, such as a steering committee, the implemented model reflected a mere reduced version of the model-excluding the municipality and NPO as partners. The system was amorphous and lacking in trust between the partners and in the system as a whole.

Managerial Style
The second category, managerial style, was divided by the interviewees into three issues: (a) personal or hidden interests, (b) professional dignity, and (c) allegiance to political organizations. These issues related directly to trust and the project's atmosphere.
Personal or hidden interests (or secondary benefits) diverge at times from organizational ones. Suspicion of hidden interests was an issue raised by three leading managers, each from a different organization. Distrust among the parties ensued: "There were personal interests … in order to fix the system according to their needs … [name] came [into the project] in order to grow, to be a 'big manager'" (Participant 6). Another manager was described as blocking the process: "He didn't care. He was a functionary who receives lots of monies who did not need to upgrade himself; therefore, he did not help, just threatened all the time" (Participant 6).
Professional dignity refers to values and beliefs managers hold, including self-esteem and achieved professional status. One interviewee described a negative attitude of managers from another organization. He wondered how managers of organizations, legally bound to supply services, could display neglect: "We aimed at a national model … invested great efforts to raise monies… but they, who carried the responsibility … were not willing to allocate" (Participant 6).
Although representatives of the other party were aware of the responsibility involved, they degraded the model offered and insisted on operating the project their way. Another manager referred to professional control from a different perspective: "It was clear to me I was not going to let him be my boss" (Participant 5).
Regarding political organization, one leader raised the issue of politics in and between the organizations, describing how some parties settled issues privately to gain advantage in the process. From the beginning, the project was presented by the HF as its project. The NPO's managers were not involved: "The project was not presented as a joint one … [The NPO] and its participation was ignored" (Participant 7; supported by documentation).
In summary, the primary feeling of all managers was distrust, including between individuals, toward the system (due to vagueness, exploitation, and interests), and between organizations (due to lack of transparency and management style). Distrust toward the project itself ensued.

Systemic Aspect
This category focuses on the system levels and not the individual one. Five subcategories concerning trust existed in this dimension: management, vagueness, complex system and readiness, resources, and climate.
Regarding management, most interviewees described issues concerning project planning and management related to preplanning collaboration and coordinating procedures among parties. All managers emphasized the importance of early alignment and the lack thereof: "We lacked a steering committee; we didn't know where we were heading" (Participant 5). As to lack of preplanning and guidelines, this participant stated: "It was a great idea missed because not enough thought was put into it." Due to poorly defined and agreed-upon coordinating procedures, implemented guidelines proved unacceptable to managers of different organizations. Distrust between the parties ensued: "There was a trust offense. People need to render services. Such difficulties demand we think twice about future engagement and about the agreement's formulation" (Participant 3).
This issue raised distrust and doubts regarding the success of the project specifically and the collaboration process in general.
Regarding vagueness, interviewees referred to vagueness of goals, job descriptions, agenda, hierarchy of management, and authority. The vision was presented and accepted by all participants as most important, yet implementation was poorly planned. Objectives and directives remained unclear: "There must have been an idea, but the construction was very rough … I knew what I was doing but not where things were leading to" (Participant 8).
One of the partners queried whether other parties had a clearer agenda. He felt that he did not understand the setting: "I don't understand; maybe he saw it clearer than we did at the beginning; this may explain his frustration" (Participant 5).
Uncertainty adversely affects trust and decision making concerning engagement in a collaboration process.
In terms of complex system and readiness, all managers identified the project's complexity and importance in advance: "The system is objectively complex. The health system is the most complex system even before complicating it any further" (Participant 2). Despite the complexity, one respondent stated: "We had faith. It was highly time consuming, but there was a real attempt [to make it happen]. Everyone thought it was important" (Participant 5). There were difficulties adjusting to change: "The failure reflected a poor managerial coordinated implementation of a modernistic idea." The municipality and health organizations displayed differences in their readiness and willingness to cooperate: It is a matter of concept. Municipalities are committed to citizens and thus systematically build in pooling of resources through NGOs. Health insurers are committed to patients and work in separation. We have tried to cooperate with them but with little success. (Participant 9) The HF's position could reflect an authoritarian-isolationist worldview that under the conditions of a complex collaboration among organizations may deplete trust and bring the collaboration to a halt.
Regarding resources, use of initially allocated and ongoing resources may affect interorganizational continuous collaboration. All participants reported on expansion of human, budgetary, and infrastructure resources. Nonetheless, all managers displayed uncertainty about the continuity of funding. The HF's managers refused to commit to a holistic model combining child development services with mental health services, because mental health services were not part of the legally mandated "health basket." 8 Furthermore, the project may have deployed the HF's services in line with community needs, but in what HF officials perceived as too available and accessible manner. The HF retracted from its commitment to implement creative ways to cover the extra costs. As one of the field managers said: "The project failed not because of ideological reasons but due to economic distress" (Participant 8).
Regarding climate, managers reported that their workers expressed vulnerability and insecurity. These feelings related to vagueness in presenting goals, defining authority, and stating clear objectives and problematic management. Widespread distrust ensued. Budgetary strictures and uncertainty led to cuts. The economic difficulties became apparent to all and left the remaining workers insecure: "There were two, maybe three circuits of dismissals … People did not believe in the project; it was intolerable" (Participant 8). A high workload left workers unmotivated. The project's unstable sustainability led to an embittered and frustrated climate.

Discussion
Our findings describe two aspects of trust affecting the reciprocal relations among the parties engaged in collaboration: interpersonal relationships and the project's systemic aspects.
Collaboration is a necessary intersystemic managerial tool (MacDonald, 2009) leading to organizational competitive advantages, proximate financial and quality measures (Laville & Nyssens, 2000;Lundin, 2007), client and employee overall satisfaction (Driscoll, 1978), and more secure survival (Abramov, 2009;Nix et al., 2008). In our study, we ventured to formulate how trust affects the success of collaboration.
Trust may promote integrative activity among organizations, sustain structures or systems, and lead to a solid organizational environment (Abramov, 2009;Dovey, 2009;Yener, 2009). The present study adds a new perspective to how the level of trust, through systemic and personal aspects, affects reciprocal relationships among managers, especially in a complex political and interorganizational collaboration (Agranoff, 2006). Trust alone is well known to be an essential factor in interorganizational relationships (Hatzakis & Searle, 2006;Jacobson & Choi, 2008;Lundin, 2007;Tsasis, 2009;Yener, 2009), which affects their success. It contains different elements (such as ego, respect, commitment, etc.), each playing a role in the relationships among the parties. Some of the elements are defined as personal characteristics, whereas others as systemic.
In agreement with previous studies, we found that the construct of trust was composed of personal characteristics such as ego, control, and exploitation and systemic characteristics such as different interests, management style, and lack of clarity in goals and purposes. These elements, among others, define and are important for building trust (Yener, 2009). Two major issues led to feelings of distrust among the organizations; the first was differences in vision and expectations, particularly because the package of medical services did not include the category of mental health. This emphasized the divergence in approaches, interests, and goals of each organization. Lack of a mutual agenda led to differences in attitudes toward the project as a whole, resulting in personal negative feelings toward the leading managers. "Trust and goal congruence must exist simultaneously in order to promote joint actions" (Lundin, 2007, p. 651). Managers' trusting expectations during the initial stages of cooperation have a major impact on the development of these relationships in later stages (Vlaar, Van den Bosch, & Volberda, 2007). Our case differed. At the beginning, the NPO and HF shared a common agenda; as their agendas diverged, the level of trust diminished. Tzafrir's (2005) notion of the dynamic role of "actionreaction" in developing and building the cycle of trust was invoked in our findings.
The need to share and manage cooperation with others in the system is accompanied by many feelings, including the fear of loss of control (Dovey, 2009;Samuel, 2004). Some organizations succeed in adopting collaboration as part of their standard management (in our case, the municipality and NPO), whereas others, if they even collaborate, adopt collaboration on a restricted basis (HF). Abramov (2009) expanded on the need for a structure that enables participation, i.e., protocols, tools, and resources. The author suggested four-phase protocols for forming and maintaining effective partnerships. The complexity of our specific project and its collaborative potential, on one hand, and the leading partner's (HF) lack of experience, on the other, led the managers to lose faith and trust in the ability of the process to take place and succeed.
The second defined dimension was interpersonal characteristics, categorized as the personal aspect (Thomson & Perry, 2006). Although, according to resource dependence theory, the domain of consensus, i.e., complementary goals and shared vision, is an important factor (Tsasis, 2009), we found that interpersonal relations also had a significant effect on the level of trust. Of all the elements we identified and divided into five major categories, two had a more significant role in determining the level of trust among the personnel: vagueness and exploitation. Vagueness may have resulted from managers' lack of clarity and understanding as to the collaborative goals, objectives, motives, and interests. Managers from some organizations described feeling that they were being deceived and there was no transparency in the process, leading participants to become increasingly suspicious of the process as a whole and their managerial counterparts. Transparency among parties is crucial for the success of cooperation (MacDonald, 2009).
Regarding the second principal element constituting the interpersonal characteristics, exploitation is "a relationship in which unearned income results from certain kinds of unequal exchange" (Gouldner, 1960, p. 165). Reciprocal stable relations demand that each party both give and receive; otherwise, negative result may occur (Wu, Zhang, Chiu, Kwan, & He, 2014). Extra efforts made by one of the parties must be compensated to counteract tensions and maintain stability-or else trust can be damaged.
All of these elements emerged in our case study. During the project, several incidents indicated a change in the level of trust among the parties that could have been resolved by a demand for control, exclusion in some cases, concessions by a party, and even holding back responses to situations to maintain the desired collaboration.
Several incidents indicated impending pitfalls in the level of trust. The HF and NPO planned an outcome-based research evaluation to assess the project; monies were set aside for the purpose of writing a grant proposal; but HF withdrew from the plan. The heralding event of the relationship breakdown among the parties was the exclusion of the NPO from the launching conference as an equal leading professional partner. In addition, the HF opposed the establishment of a steering committee consisting of the four partners. Hence, neither joint management nor protocols were agreed upon. Furthermore, the HF may have changed its agenda after a national decision that mother-child clinics not be transferred to the HFs' responsibility, as had been initially planned. Moreover, mental health services were still excluded from the health basket, and the HF found it complicated to assume responsibility for a project committing to the delivery of these services. A withdrawal from the originally agreedupon holistic model of a CDI was followed by a shrinking vision of medical child development services. Evidence of this intention emerged in the HF's final request of the NPO to serve as an infrastructure mediator in obtaining the municipality's permit to use the mother-child clinics for their own service-delivery operation. By then, the HF lost all working relationships and trust with both the municipality and NPO. The cessation of the project was in close sight.
To conclude, a few main stumbling blocks occurred along the way, suggesting difficulties in the collaboration process. They all related to major systemic structural and functional interorganizational lacunae. First, although milestones were determined for the collaboration, there was no established follow-up process or integrated management to discuss implementation problems. Second, there was no joint steering committee or system regulation, which have a critical role in planning and executing the collaboration process.
These incidents led to wariness on the part of the different parties, expressed as deleterious changes in behavior, attitude, relations, and interactions, leading to changes in the political behavior of the parties. The ongoing adjustments made by each party as a response to these incidents changed the balance of power and affected the level of trust in the newly born system.
As described in the literature, level of trust can be determined by objective elements such as mutual goals, reputation, performance, outcomes, etc. (Yener, 2009), and can also evolve from successful past cooperation. Mutual positive interaction among past partners leads to mutual dimensions of trust (Tzafrir, 2005). Similarly, the positive intensity of the relationships among partners raises the level of trust (Adobor, 2006). In this case (see Figure 2), the positive interaction between the municipality and NPO led to positive expectations concerning future collaboration. Although uncertainty is part of the collaboration process, for the process to succeed, effective partners need to act to reduce uncertainty and raise the level of trust (Adobor, 2006). However, Vlaar et al. (2007) stated that a certain combination of trust and distrust may be most productive for interorganizational cooperation. In our case, uncertainty became too high, and the actors were unwilling to act to build trust, even in the face of diverging agendas.
The differences among the parties involved in the project occurred in the domain of commitments and thus their inherent goals. The municipality and NPO traditionally carry a universal commitment, whereas the HF a particularistic one-to their consumers. Furthermore, the municipality and HF run on a limited budget, whereas the NPO can adapt budgets as it goes along, enjoying an economic public infrastructure to support innovative goals. These differences, in themselves, put the various institutions in different positions of short-and longterm control and commitment and thus created different economic and operational interdependencies among the partners.
In hindsight, it is understandable that the HF, required by law to render certain services with fixed budgets, would be concerned about extending its commitment while relying on voluntary organizations. It has very little experience and tradition with this approach. The municipality had developed a long tradition of extending beyond statutory commitments for the benefit of the city's inhabitants. The very existence of the current project is, among others, a reflection of shrinking budgets on the side of the municipality coupled with its unwillingness to withdraw important services and desire to create more innovative ones aided by NPOs.
As described in the introduction, the first phase of collaboration, i.e., between the municipality and its clinics and the NPO was successful-the universal commitment and tradition of collaboration were perceived by the interviewees as a good outcome. The HF's budgetary regulations and its more particularistic commitments (in term of consumers and area of activity) may have led to more aggressive behavior regarding control and exploitation, to which the other organizations might not have been either prepared or ready to learn from and respond to in real time. Trust was replaced by false trust-to help avert awareness-and functional control was replaced by defensive control, increasing political and collaborative conflicts. The system was not functioning as a whole, and the parties never stopped to assess its actual state. Partial systems determined the outcome and thus the general tenor of the collaboration. To conclude, a complex collaborative structure, integrating four different organizations, each operating by different managerial codes, is a special situation wherein the effect of trust on the success of the collaborative process is unique.

Limitations and Future Research
The most significant limitation of this study is its case study design. Can such research legitimately derive generalization from qualitative findings (Denzin & Lincoln, 1994)? We achieved trustworthiness and dependability via several methods: (a) cross-over between deductive and inductive analysis using information collected through interviews, gathering different descriptions of specific incidents from several interviewees, and manipulating personal reflections on the experiences as described by the interviewees; (b) anchoring interviewees' information with documentation such as contracts, protocols, etc., and (3) confirming all quotations with the interviewees themselves. All methods are evidence-based concerning trustworthiness and dependability (Lincoln & Guba, 1985;Patton, 1990;Yin, 1994).
The model described here is unique because the case study itself takes a unique approach, and to extrapolate findings from it to other similar cases is questionable. Yet we provided a deeper understanding of nuances and factors at play that we could not see in a larger study or with a different methodology. Nevertheless, finding similar characteristics that define a certain situation may help researchers in future cases. Further research should elucidate the relationships among organizations (as opposed to key individuals). Such research could support the development of a model to promote trust among individual managers and managerial units.

Contributions
The present study has important conceptual, methodological, and practical implications. From a conceptual perspective, understanding the collaboration process enables managers to predict results and plan their actions in advance. Our findings emphasize the importance of analyzing multilevel trust (Costa, Fulmer, & Anderson, 2017), interpolitics, and intrapolitics, even when

Political behavior in & between the organizations
Trust among decision makers in organizations Collaboration success is at stake (Lawrence, Hardy, & Phillips, 2002). We suggest that managers have to account for emotional involvement at the individual level, even when successful organizationallevel collaboration occurs. Researchers, managers, and consultants must pay attention to different interests, egos, visions, and other topics as previously described to prepare themselves to successfully collaborate and better promote synergistic integration.
From a methodological point of view, this two-year study of collaborative activities using a case study approach demonstrated the utility of qualitative methodology for learning about the initiation of collaborative projects. This study and its findings reveal potential managerial tools. With this knowledge, managers could cope with complex environments requiring trust development to enhance organizational collaboration, particularly in times of systemic shortage.
Results of the collaboration might be anticipated by (a) assessing economic and psychosocial general (e.g., commitment) parameters; and (b) identifying methodological systemic evasion concerning realistic assessment of the process of collaboration. Maintaining trust in the face of signs of exploitation may prove counterproductive. Aired and discussed, mistrust could be superseded, or the collaboration could be responsibly terminated.
Our study is consistent with the work of Grubbs (2000), who stated that the new generation of public administrators will have to find creative and collaborative ways to deal with the new public management model. More specifically, health care systems, confronted with rising public demand for services, could enlist available public and social service systems to comply with their professed duty and better serve their consumers.

Notes
1. Overall, 45 questions dealt with general and specific aspects of the project and the collaboration process. 2. To follow the ethics committee's stipulations, we do not identify them by name. 3. Samples in qualitative research are typically small (Hill, Thompson, & Williams, 1997) and vary across studies, and each participant has great importance in the analysis. Choosing the appropriate interviewees is extremely important. As Morrow (2005) suggested, the "numbers alone have little to do with the quality or adequacy of qualitative data" (p. 255). Hill et al. (1997) recommended including at least eight to 15 participants as a sufficient sample; "using much larger samples is unrealistic" (p. 532), according to these authors. Also, "far more important than sample size are sampling procedures; quality, length, and depth of interview data; and variety of evidence" (Morrow, 2005, p. 255). In our case, all managers of the project were interviewees (N=11). 4. Several methods have been reviewed in the literature concerning technical approaches to data analysis, all of which were accepted by different researchers and sometimes complement one another (Denzin & Lincoln, 1994). 5. Three years earlier, a certain NPO joined social public health municipal services to establish a team for proper evaluation and early intervention with preschool children at risk in some mother-child clinics in communities of low socioeconomic status. The project goal was developmental diagnosis and treatment by health funds, as obligated by law. The municipality brought the integrated project under the purview of a steering committee directed by the chairperson of the municipal health committee. In Israel, the mother-child clinic system is governmentally funded, supplies necessary preventive medical care, and covers health issues for 90% of children in their first year of life. Nurses refer identified children to pediatricians assigned by the municipal hospital.
Further assessment requires referral to pediatricians at HFs and then, if needed, to a CDI. Hence, three pediatric teams are involved in the full developmental and mental assessment, with a complicated process involving waiting lists and paying compulsory out-of-pocket fees. 6. The initiator and former co-director of the project, representing the NPO, also had a separate and different function in the HF that placed him under the subordination of the HF's director. Therefore, he stepped aside from any operational role, and a separate HF child mental health director was assigned to the position. 7. We examined these characteristics in the eye of the beholder within a specific context, while participants were acting and defined according to their organization's culture and agenda. 8. The "health basket" is defined as a set of services, activities, and goods that is considered to be essential and thus within the scope of reimbursement or direct service delivery (Velasco-Garrido, 2006).